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
|
|---|---|---|---|---|---|
Post-translational modifications of histones play a key role in DNA-based processes , like transcription , by modulating chromatin structure . N-terminal acetylation is unique among the numerous histone modifications because it is deposited on the N-alpha amino group of the first residue instead of the side-chain of amino acids . The function of this modification and its interplay with other internal histone marks has not been previously addressed . Here , we identified N-terminal acetylation of H4 ( N-acH4 ) as a novel regulator of arginine methylation and chromatin silencing in Saccharomyces cerevisiae . Lack of the H4 N-alpha acetyltransferase ( Nat4 ) activity results specifically in increased deposition of asymmetric dimethylation of histone H4 arginine 3 ( H4R3me2a ) and in enhanced ribosomal-DNA silencing . Consistent with this , H4 N-terminal acetylation impairs the activity of the Hmt1 methyltransferase towards H4R3 in vitro . Furthermore , combinatorial loss of N-acH4 with internal histone acetylation at lysines 5 , 8 and 12 has a synergistic induction of H4R3me2a deposition and rDNA silencing that leads to a severe growth defect . This defect is completely rescued by mutating arginine 3 to lysine ( H4R3K ) , suggesting that abnormal deposition of a single histone modification , H4R3me2a , can impact on cell growth . Notably , the cross-talk between N-acH4 and H4R3me2a , which regulates rDNA silencing , is induced under calorie restriction conditions . Collectively , these findings unveil a molecular and biological function for H4 N-terminal acetylation , identify its interplay with internal histone modifications , and provide general mechanistic implications for N-alpha-terminal acetylation , one of the most common protein modifications in eukaryotes .
The nucleosome is the basic unit of chromatin and comprises 147 base pairs of DNA wrapped around a histone octamer , which contains two copies of each of the four histones H2A , H2B , H3 and H4 . These histones are subjected to a variety of post-translational modifications , such as methylation , acetylation and phosphorylation , mediated by specific modifying enzymes [1] . Histone modifications often function by recruiting effector molecules to alter the structure of chromatin in order to regulate DNA-based processes such as transcription , replication and DNA repair [1] . An additional property of these modifications is the fact that they cross-talk to each other , whereby one modification influences the establishment or maintenance of a second modification [2] . N-alpha-terminal acetylation is a type of modification occurring on histones . In fact , it is one of the most common protein modifications , present on 80–90% of soluble mammalian proteins and 50–70% of yeast proteins [3] , [4] . This mark , which is deposited on the first amino acid residue of the protein has a range of molecular and biological roles , including regulation of protein degradation , protein translocation , protein complex formation , membrane attachment , apoptosis and cellular metabolism [5] . All four core histones [6]–[8] and the linker histone H1 [9] possess N-terminal acetylation , but this modification is more abundant on histones H2A and H4 [8] . N-terminal acetylation of these two histones is mediated by the N-alpha terminal acetyltransferase Nat4 ( also known as NatD or Naa40 ) . This enzyme was originally identified in the budding yeast S . cerevisiae [7] , but also the human ortholog hNaa40 ( also designated as hNatD , Nat11 or Patt1 ) has been recently characterized [10] , [11] . Both yeast Nat4 and hNaa40 target only histones H2A and H4 , and this specificity differentiates them from all other described N-alpha acetyltransferases , which can target numerous substrates [5] . Previous studies have attempted to determine the biological role of yeast and human Nat4 . Deletion of NAT4 in yeast showed growth sensitivity when cells were cultured in media containing various chemicals such as 3-aminotriazole ( 3-AT ) , an inhibitor of transcription [12] . This sensitivity is enhanced when the NAT4 deletion is combined with mutations in histone H4 where lysines 5 , 8 and 12 have been replaced by arginines ( K5 , 8 , 12R ) [12] , suggesting that modifications at these residues and N-terminal acetylation are linked through a mechanism that remains elusive . In humans , hNaa40 has been identified as a pro-apoptotic factor and has been implicated in hepatocellular carcinogenesis [11] . Furthermore , a recent study demonstrated that in mice this N-terminal acetyltransferase plays a role in hepatic lipid metabolism [13] . Although some studies have already uncovered phenotypes related to the loss of Nat4 and have provided insights about the biological role of its human ortholog , the molecular function of histone N-terminal acetylation still remains unknown . Arginine methylation is another histone modification that has attracted much attention in recent years . This is because of its involvement in various cellular processes and the identification of a family of enzymes that catalyze it . These enzymes are called protein arginine methyltransferases ( PRMTs ) and have already been associated with cancer pathogenesis [14] . PRMTs deposit one or two methyl groups to the guanidino groups of arginine residues resulting in monomethylated ( Rme1 ) , asymmetrically dimethylated ( Rme2a ) or symmetrically dimethylated ( Rme2s ) states . Others and we have previously shown that arginine methylation cross-talks with adjacent histone modifications by controlling their deposition [15]–[21] . It is , however , also important to discover the mechanisms that regulate PRMT activity and the deposition of histone arginine methylation . Histone H4 arginine 3 ( H4R3 ) is one of the residues that can possess any of the methylation states [20] . In particular , its asymmetrically dimethylated form is mediated by PRMT1 and it is associated with active transcription in mammals [22] , [23] . In yeast , however , the functional homolog of PRMT1 ( known as Hmt1 ) that catalyzes H4R3me2a in vitro [24] , has been linked to transcriptional repression . Specifically , Hmt1 and its associated H4R3me2a modification have been implicated in the formation of silent chromatin at yeast heterochromatin-like loci , including the rDNA repeat region [25] . This region contains an array of approximately 150 tandem repeats covering approximately 1–2 megabases of chromosome 12 [26] . Within each 9 . 1 kb rDNA repeat there are two transcriptional units known as RDN5 and RDN37 which , respectively , encode for the 5S and 35S rRNAs . The 35S transcript is quickly processed after transcription to generate the 18S , 5 . 8S and 25S rRNAs [27] , [28] , which together with 5S are components of a eukaryotic ribosome . Whether the link between H4R3me2a and rDNA silencing relate to the transcriptional levels of these rRNAs is unclear [25] . A previous study has shown that neighboring histone acetylation at lysines 5 , 8 and 12 regulates the activity of PRMT1 towards H4R3 in vitro [29] . A similar crosstalk among these adjacent modifications has also been proposed to occur in yeast histones [30] , but in general the regulation of H4R3me2a in vivo remains largely unexplored . Here , we sought to identify factors that control the occurrence of this mark by employing a GPS ( Global proteomic screen in S . cerevisiae ) approach [31] . Using an antibody that specifically detects H4R3me2a we identified Nat4 as an inhibitor of this modification in yeast . Consistent with a role of H4R3me2a in promoting silencing at the rDNA region [25] , we find that deletion or inactivation of Nat4 results in enhanced silencing of ribosomal DNA genes . Importantly , we demonstrate that this regulation is mediated through N-terminal acetylation of H4 ( N-acH4 ) , but not of H2A . Additionally , we show by using in vitro methylation assays that H4 N-terminal acetylation inhibits the activity of the Hmt1 arginine methyltransferase towards H4R3 . Interestingly , we find that combinatorial loss of H4 N-terminal and internal K5 , 8 , and 12 acetylation can induce H4R3me2a deposition even more . Excessive H4R3me2a leads to a severe growth defect , which is rescued by preventing arginine 3 methylation by mutating this residue to lysine . Finally , we provide evidence that the interplay between N-acH4 and H4R3me2a functions under conditions of calorie restriction , which induce rDNA silencing . Altogether , our results reveal the function of H4 N-terminal acetylation in gene regulation , and elucidate the underlying molecular mechanism that links this N-terminal acetylation to other internal histone modifications .
We sought to identify proteins that regulate the deposition of asymmetrically dimethylated arginine 3 on histone H4 ( H4R3me2a ) . To do this we developed an antibody that recognizes specifically methylated H4R3 ( Figures S1A and S1B ) when it is asymmetrically dimethylated ( Figure S1C ) and performed a GPS screen using the yeast deletion collection . We found that deletion of the N-alpha acetyltransferase 4 ( nat4Δ ) results in robust induction of the H4R3me2a levels ( Figure 1A , lane 6 ) . None of the other four yeast N-terminal acetyltransferases ( NatA , NatB , NatC or NatE ) showed an effect on H4R3me2a when they were deleted ( Figure S2A ) . This effect was specific to the asymmetrically dimethylated form at H4R3 , as specific antibodies ( Figure S1C ) towards monomethylated ( H4R3me1 ) or symmetrically dimethylated ( H4R3me2s ) states of this residue detected similar levels for these marks between wild-type and nat4Δ strains ( Figure 1B , compare lane 1 to 2 ) . To determine whether Nat4 regulation towards H4R3me2a was dependent on its N-terminal acetyltransferase activity we constructed a catalytically inactive version of this enzyme . Mutation of four highly conserved residues ( Figure 1C ) found within the two motifs of its acetyltransferase domain [7] result in increased signal of H4R3me2a , phenocopying nat4Δ ( Figure 1D , compare lane 2 and lane 6 ) . Notably , the increase of H4R3me2a in Nat4 deficient cells is not due to epitope preference of the H4R3me2a antibody , as it recognizes equally well H4R3me2a peptides that are either N-terminally acetylated or unacetylated ( Figure S1C , compare rows 5 and 6 ) . Together , these findings show that Nat4 regulates the levels of H4R3me2a through its N-terminal acetyltransferase activity . Since a previous study has linked H4R3me2a with the establishment of silencing at the four heterochromatin-like regions in yeast ( rDNA , HML , HMR and telomeres ) [25] , we sought to determine whether deficiency of Nat4 activity will affect expression at these loci . We found that deletion of NAT4 did not affect greatly the expression at HMR , HML and TEL-VII-L ( Figure S3 ) , but strongly enhances silencing at the rDNA locus ( Figure 2A ) . Due to recent concerns in using FOA-sensitivity assays to assess chromatin silencing [32] , [33] , we tried to validate the above result by testing the expression of the endogenous rDNA transcripts ( Figure 2B ) . Examining the levels of the different ribosomal RNAs ( 5S , 25S , 5 . 8S and 18S , as well as their precursor 35S ) by real time-PCR we confirmed the above result , as deletion of NAT4 significantly reduced the amount of all rRNAs ( Figure 2B ) . It is worth mentioning that because the 35S primary transcript is quickly processed [27] , [28] , the observed changes in the levels of rRNAs are most likely caused by a decrease in transcription . Interestingly , the deletion of NAT4 does not affect the mRNA levels of the ribosomal protein Rpp0 ( Figure 2B , rightmost panel ) . A similar result was obtained when the Nat4 catalytic mutant strain ( nat4cmAB-HA ) was used to examine the levels of 25S ( Figure S4A ) , suggesting that rDNA expression is dependent on the Nat4 acetyltransferase activity . According to these findings , we anticipated that the reduced expression of all rRNAs would correlate with increased deposition of H4R3me2a at the rDNA genes . Indeed , ChIP analysis confirmed that there is higher nucleosomal deposition of H4R3me2a across the entire rDNA locus when Nat4 is absent ( Figure 2C , top panel ) . Consistent with this , an induction of H4R3me2a deposition was also observed at RDN25 in the strain expressing a catalytically inactivated Nat4 ( Figure S4B ) . As expected , based on the results in figure 1B we did not see changes in the occupancy of H4R3me1 at RDN25 in the nat4Δ strain ( Figure S4C ) . Importantly , the lack of N-terminal acetyltransferase activity in the nat4Δ and Nat4 catalytic mutant ( nat4cmAB-HA ) strains was confirmed by an antibody against N-terminally acetylated H4 ( N-acH4 ) , which showed that this modification was reduced throughout the rDNA region ( Figure 2C , bottom panel and figures S4B–C ) . Although our data above demonstrate that Nat4 suppresses rDNA silencing and prevents H4R3me2a deposition , they do not show which one of its two targets , histone H4 or H2A , is implicated in this regulation . To determine this , we constructed yeast strains in which either H4 or H2A were compromised for N-terminal acetylation . Endogenous H4 or H2A were expressed with an alanine instead of a serine ( H4S1A or H2AS1A ) at the first residue because for both histones , the sequence of their first 30 amino acids is absolutely required for efficient acetylation by Nat4 [7] , [12] . Figure 3A shows that mutation of H4 serine 1 to alanine induces H4R3me2a deposition ( compare lanes 1 and 2 ) , but the same mutation in H2A has no effect on this methylation ( compare lanes 3 and 4 ) . Furthermore , in the H4S1A mutant strain we detected higher amounts of H4R3me2a deposited at the RDN25 gene compared to an isogenic wild-type ( H4WT ) strain ( Figure 3B ) . On the other hand , the H2AS1A mutant did not show significant difference in H4R3me2a levels compared to H2AWT strain ( Figure 3B ) . We also like to note that the results of the H4S1A haploid mutant strain are not affected by the fact that H4S1 is also phosphorylated because this modification is only induced under sporulation conditions in diploid cells [34] . To validate that Nat4 regulation of rDNA silencing is mediated through H4 , we then examined the expression of this locus in the H4S1A strain ( Figure 3C and 3D ) . Both , analysis of 25S rRNA expression levels and silencing spot assays demonstrate that the H4S1A mutation enhances repression of this locus similarly to nat4Δ ( compare Figure 3C and 3D to Figure 2B and 2A , respectively ) , albeit to a lesser extent . This result is in agreement with the increased H4R3me2a levels shown above ( Figure 3B ) . In contrast , H2AS1A mutation does not alter the levels of 25S rRNA ( Figure 3C ) . Additional evidence that rDNA silencing is mediated through N-terminal acetylation of H4 comes from using a strain expressing a H4S1P mutant . The presence of proline at position 1 blocks N-terminal acetylation completely as shown by mass-spectrometry analysis of proteins extracted from yeast , Drosophila melanogaster and human cells [3] , [35] . As expected , we observed a significant decrease in the levels of 25S rRNA in the H4S1P mutant strain , similarly to the effect observed in the nat4Δ strain ( Figure S5 ) . Altogether , these results suggest that Nat4 regulates rDNA silencing and H4R3me2a deposition via H4 N-terminal acetylation , but not through N-acH2A . Finally , to verify the link of N-acH4 within this mechanism , we have also monitored the levels of H4R3me2a in a nat4Δ strain that expresses ectopically the human ortholog of Nat4 ( hNaa40 ) . It was previously shown that expression of hNaa40 in yeast results in N-terminal acetylation of H4 but not of H2A [10] . In agreement with our data above , we found by western blotting that expression of hNaa40 in a nat4Δ strain reduces H4R3me2a back to wild-type levels ( Figure 3E , compare lane 3 to lane 1 ) . ChIP analysis also showed that expression of hNaa40 in a nat4Δ strain fully restores the N-acH4 levels at RDN25 ( Figure 3F ) , confirming that histone H4 is the main substrate through which Nat4 regulates H4R3me2a . The above results suggest that N-acH4 regulates the deposition of H4R3me2a . To explore this further , we wanted to determine whether the activity of the yeast arginine methyltransferase Hmt1 , which was previously shown to target H4R3 in vitro [24] , is inhibited by N-acH4 . We performed methyltransferase assays using Hmt1 purified from yeast cells ( Figure S6 ) and synthetic peptides corresponding to the first twenty amino acids of H4 . Immunoblotting for H4R3me2a showed that Hmt1 dimethylates much more efficiently H4R3me1 peptides that are not N-terminally acetylated as opposed to those that possess N-acH4 ( Figure 4A , compare lanes 13 and 14 ) . Overall , these findings show that N-acH4 represses the deposition of H4R3me2a by blocking the activity of the associated arginine methyltransferase . The previous results link Nat4 with rDNA silencing and H4R3me2a . However , they do not demonstrate whether methylation at H4R3 is necessary and sufficient for the effect of Nat4 towards rDNA expression . To determine this , we investigated the effect of NAT4 deletion on 25S rRNA expression when arginine 3 was mutated to lysine ( H4R3K ) in order to prevent its methylation ( Figure S1A ) . Despite the loss of N-acH4 in a H4R3K nat4Δ double mutant strain ( Figure 4B ) , the expression levels of 25S rRNA are not reduced compared to the nat4Δ only strain ( Figure 4C ) . ChIP analysis at RDN25 confirms that H4R3me2a is induced in the nat4Δ strain , and is undetected in the H4R3K nat4Δ strain ( Figure 4B ) . This finding indicates that H4R3 and most likely its methylation are absolutely required for the control of rDNA silencing by Nat4 and N-acH4 . Evidence from two previous studies have raised the hypothesis that N-acH4 works together with acetylation of H4K5 , H4K8 and H4K12 to control the deposition of H4R3me2a . The first study showed that asymmetric dimethylation of H4R3 mediated by PRMT1 is inhibited in vitro by acetylation of lysines 5 , 8 and 12 of H4 [29] . The second one demonstrated a synthetic defect in yeast containing nat4Δ and a triple lysine to arginine mutant ( H4K5 , 8 , 12R ) [12] . Hence , to explore this hypothesis , we combined nat4Δ with the H4K5 , 8 , 12R mutant because deletion of Esa1 , that acetylates these three lysines is inviable [36] . Interestingly , we found that concurrent loss of N-acH4 and acetylation of H4K5 , 8 , 12 ( H4K5 , 8 , 12R nat4Δ ) results in robust induction of H4R3me2a ( Figure 5A , compare lanes 2 and 4 ) , suggesting that these H4 residues collaborate to regulate H4R3me2a . This result is not due to an antibody artifact , as the H4R3me2a antibody recognizes slightly better methylated peptides in which positions 5 , 8 , and 12 are lysines than when these residues are arginines ( Figure S7 , compare rows 2 and 3 ) . Because it was recently shown that H4K5 , K8 and K12 could also be methylated by Set5 [37] , we wanted to investigate the possibility that methylation of these lysines could act synergistically with N-acH4 to control H4R3me2a . Double nat4Δ set5Δ deletion did not enhance H4R3me2a levels compared to the nat4Δ single mutant ( Figure S8 , compare lanes 2 and 4 ) , indicating that it is acetylation , and not methylation of H4K5 , 8 , 12 that cooperates with N-acH4 . Notably , N-acH4 is the major regulator of H4R3me2a , as the H4K5 , 8 , 12R mutant alone does not increase the levels of H4R3me2a to the same extent as nat4Δ ( Figure 5A , compare lanes 2 and 3 ) . To further validate the above results , we also examined the effect of the combination of nat4Δ with H4K5 , 8 , 12R on the deposition of H4R3me2a and 25S rRNA expression . Consistent with the previous findings we observed a significant enrichment in H4R3me2a at the RDN25 gene when NAT4 is deleted together with the H4K5 , 8 , 12R mutant as opposed to the nat4Δ single mutant ( Figure 5B ) . Moreover , the higher presence of H4R3me2a in the H4K5 , 8 , 12R nat4Δ double mutant strain results in further reduction of 25S rRNA levels when compared to the nat4Δ alone ( Figure 5C ) . Taken together , these results indicate that , both N-acH4 ( Figure 4A ) and internal lysine acetylation [29] can impede on the methylase activity that targets H4R3 , but according to our findings N-acH4 is the predominant regulator of H4R3me2a and rDNA silencing ( Figure 5A–C ) . Considering that the double mutant of nat4Δ with H4K5 , 8 , 12R results in robust reduction of 25S rRNA levels ( Figure 5C ) , we then examined the growth rate of this strain using serial dilution spotting assays ( Figure 5D ) and by measuring its doubling time ( Figure S9 ) . Notably , the double mutant strain ( H4K5 , 8 , 12R nat4Δ ) has a severe growth defect in comparison to the corresponding single mutants ( Figures 5D and S9 , left panels ) . This growth defect becomes lethal when cells are grown at a higher ( 37°C ) temperature ( Figures 5D and S9 , right panels ) . Moreover , when H4S1A is combined with the H4K5 , 8 , 12R mutant a growth defect is also observed , albeit less severe ( Figures 5D and S9 , left panels ) , consistent with the milder deregulation of H4R3me2a and rDNA expression in the H4S1A mutant as opposed to nat4Δ ( compare Figures 2 and 3 ) . This synthetic defect supports the synergistic effect between N-acH4 and internal H4 lysine acetylation in controlling H4R3me2a and rDNA expression . Based on the previous experiments which showed that arginine 3 is necessary and sufficient for the regulation of rDNA silencing by Nat4 ( Figure 4 ) , we then examined whether H4R3K can rescue the growth defect caused by the combination of nat4Δ and the H4K5 , 8 , 12R mutant . Interestingly , H4R3K rescues entirely the growth defect of the double H4K5 , 8 , 12R nat4Δ mutant grown at an ambient ( 30°C ) or even at a higher ( 37°C ) temperature ( Figures 5D and S9 ) . Additionally , H4R3K restores the rRNA expression levels to almost near wild-type in the double H4K5 , 8 , 12R-nat4Δ mutant strain ( Figure 5C ) . All together , these results reveal that excessive H4R3 asymmetric dimethylation caused by lack of N-acH4 and internal lysine acetylation impairs cell growth . The expression of the rRNA transcripts is modulated by various environmental and intracellular stress conditions . One such condition is calorie restriction , which is studied in yeast by diminishing the levels of glucose in the media . Previous studies have shown that reduction of glucose levels from 2% to 0 . 5% can enhance rDNA silencing [38] , [39] . Hence , we sought to determine whether the crosstalk of N-acH4 and H4R3me2a is induced under these conditions in a wild-type yeast strain . In agreement with previous studies , we found that lowering the glucose availability decreases the levels of 25S rRNA , and this reduction is greater under severe ( 0 . 1% and 0 . 05% glucose ) calorie restriction ( Figure 6A ) . Most importantly , the decrease in the amount of 25S rRNA correlates with an increase in the H4R3me2a∶N-acH4 enrichment ratio at the RDN25 gene . The increase in the enrichment of H4R3me2a against N-acH4 is evident under severe calorie restriction , in line with the lower levels of 25S rRNA ( Figure 6B , see 0 . 1% and 0 . 05% glucose ) . These findings suggest that the interplay between H4 N-terminal acetylation and H4R3me2a controls rDNA silencing in response to environmental stimuli such as nutrient deficiency .
The molecular function of histone H4 N-terminal acetylation was unknown until now , even though this is an abundant and conserved modification that was reported several decades ago [40] . In this study , we describe an important role of N-acH4 in the regulation of histone arginine methylation and rDNA silencing . Taken together , our data support a model in which N-acH4 mediated by Nat4 strongly inhibits the activity of the Hmt1 methyltransferase towards H4R3 . This inhibition leads to activation of the rDNA loci . Removal of N-acH4 by a yet unknown mechanism , allows deposition of H4R3me2a and results in repression of rRNA transcription ( Figure 7 ) . This mechanism is activated during calorie restriction in order to reduce the expression of the rDNA region in response to the limited source of energy . In the absence of N-acH4 , internal lysine acetylation at K5 , K8 and K12 catalysed by Esa1 [36] or Hat1 [41] remain unaffected ( Figure 5A , compare lanes 1–2 and Figure S10 ) . These acetyl marks can fine-tune the levels of H4R3me2a because otherwise excessive methylation of H4R3 will result in a severe growth defect ( Figures 5D and S9 ) . The proposed mechanism also provides an explanation for the previously observed synthetic defect of the double H4K5 , 8 , 12R nat4Δ mutant strain [12] . Whether the growth defect observed in our experiments is due to deregulation of the rDNA region only or whether other genomic loci whose expression is influenced by H4R3me2a also contribute to this phenotype is still unclear . There are two possible scenarios , which are not mutually exclusive , on how H4R3me2a then mediates rDNA silencing in yeast . First , it was proposed earlier that H4R3me2a facilitates recruitment of Sir2 to the rDNA region [25] . Second , based on previous findings that arginine methylation occludes recruitment of effectors to adjacent modifications [16]–[18] , [20] , [42] , it is possible that H4R3me2a prevents the binding of an activator at the neighboring N-terminal or lysine acetylation marks . Previous studies suggested that Nat4 and hNaa40 acetylate H4 co-translationally as they were found associated with the ribosomes [10] , [12] . However , it remains possible that Nat4 and hNaa40 target H4 post-translationally because a significant amount of hNaa40 localizes to the nucleus [10] , [11] . Similarly to other acetyl marks such as H4K5ac and H4K12ac [43] , N-acH4 might be catalyzed on soluble nuclear histones that are subsequently incorporated into chromatin . In support of this , N-alpha-terminal acetylation has been proposed to occur post-translationally on other proteins [44] , [45] . How N-acH4 is then removed from histones is another pending question . One possibility is through histone exchange by which unacetylated H4 replaces N-terminally acetylated H4 found in chromatin . Another scenario is through active deacetylation mediated by a deacetylase , an activity that has not been demonstrated yet for any protein N-terminal acetylation mark [5] . Interestingly , Nat4 is not the only Nat that has been implicated in the regulation of heterochromatic regions in yeast . NatA has also an active role in chromatin silencing but possibly functions through a mechanism that is distinct from that of Nat4 for three main reasons . Firstly , NatA establishes telomeric and HML silencing by acetylating Orc1 and Sir3 in order to stimulate their recruitment onto chromatin [46]–[48] . However , silencing at the rDNA region does not involve these proteins [49] . Secondly , in our experiments the absence of Ard1 ( the catalytic subunit of NatA ) has no effect on the levels of H4R3me2a ( Figure S2A ) . Finally , in the ard1Δ strain , the levels of 25S rRNA are not significantly altered compared to a wild-type strain ( Figure S2B ) . Therefore , we believe that Nat4 and NatA impact on chromatin silencing through different pathways . A link between calorie restriction and increased lifespan in yeast and other organisms has already been established [50] . Considering that changes in the levels of N-acH4 and H4R3me2a are associated with calorie restriction ( Figure 6 ) , it would be interesting to determine in future studies whether these modifications and their respective enzymes are part of a mechanism that extends cellular lifespan . Another histone H4 acetylation ( H4K16ac ) has already been implicated in the regulation of lifespan in yeast through a mechanism that maintains telomeric chromatin intact [51] . In contrast , we anticipate that N-acH4 and H4R3me2a , if involved in lifespan regulation , would be part of a pathway that controls rDNA silencing [38] , [39] , since our data show that deletion of NAT4 does not affect telomeric silencing ( Figure S3 ) . Alternatively , N-acH4 , H4R3me2a and their associated enzymes could influence longevity by regulating rDNA recombination , given that Hmt1 activity represses this process [25] . Two recent studies support this idea because they show that calorie restriction suppresses rDNA recombination independently of rDNA silencing in order to extend lifespan [52] , [53] . Although this study was performed entirely in yeast , there is evidence suggesting that the cross-talk among N-acH4 , internal lysine acetylation and H4R3 methylation may be conserved in mammals . For instance , the activity of Nat4 towards H4 is conserved in humans [10] , and its ortholog hNaa40 can re-establish normal levels of H4R3me2a in the absence of Nat4 ( Figure 3E–F ) . Furthermore , mass spectrometry analysis of mouse histone H4 revealed that N-terminal acetylation co-exists with K5 , K8 and K12 acetylation and has an inverse relationship with H4R3 methylation [54] . Interestingly , this anticorrelation in mouse cells does not involve asymmetric dimethylation but rather a trimethylated form of H4R3 [54] , whose existence is still under debate . The mutual exclusive pattern between N-terminal acetylation and H4R3 methylation becomes even more apparent on H2A peptides [54] , suggesting that in mammals this modification crosstalk could also occur on histone H2A . This is consistent with the fact that mammalian H2A ( Ser-Gly-Arg-Gly-Lys ) has an arginine at position 3 and its N-terminal sequence is identical to H4 , in contrast to yeast H2A ( Ser-Gly-Gly-Lys-Gly ) whose third residue is a glycine . Determining whether Naa40 utilizes a similar mechanism to control gene activation in mammalian cells is intriguing , considering that this enzyme has a pro-apoptotic function and was found significantly downregulated in hepatocellular carcinomas [11] . In summary , this study provides a novel link between protein N-terminal acetylation and the regulation of gene expression . This regulation employs a unique mechanism by which histone N-terminal acetylation influences the deposition of another in cis modification . Since N-terminal acetylation occurs on the majority of soluble eukaryotic proteins [3] , [4] , we propose that its crosstalk with internal post-translational modifications might be a common mechanism for controlling protein function .
All strains used in this study are listed in Table S1 and described in Protocol S1 . Rabbit polyclonal antibodies were raised against H4R3me2a and N-acH4 by Eurogentec ( Belgium ) . Additional details are provided in Protocol S1 . Other antibodies used were: H4K5ac ( ab51997; Abcam ) , H4K12ac ( ab46983; Abcam ) , H4K8ac ( ab15823; Abcam ) , H4R3me1 ( ab17339; Abcam ) , H4R3me2s ( ab5823; Abcam ) , H3 ( ab1791; Abcam ) , H4 ( 62-141-13; Millipore ) , Naa40 ( ab106408; Abcam ) , b-Actin ( ab8226; Abcam ) and His-tag ( 2365; Cell Signalling ) . Overnight cultures were diluted to OD ∼0 . 1 and grown to mid-log phase . Approximately 1 . 2×104 cells were serially diluted 10-fold , and spotted onto the right media plates ( YPAD , SC or SC+5′-Fluoroorotic acid ) . The plates were incubated at 30°C or 37°C for 2 days . Doubling time of cell growth was measured as indicated on http://www . doubling-time . com/compute . php . Total RNA from logarithmically grown ( OD 0 . 8 ) yeast cells was isolated using the hot phenol extraction method [55] and was then treated with the TURBO DNA-free DNase kit ( Ambion ) . Isolated total RNA ( 0 . 5 µg ) from each sample was mixed with 1 µl dNTP mix ( 10 mM ) and 1 µl of primer cocktail that consists of 0 . 5 µl oligo- ( dT ) 20 primer ( 50 µM ) and 0 . 5 µl random hexamers ( 50 µM ) ( Invitrogen ) . DNase RNase-free water was added up to a final volume of 13 µl . The mixture was incubated at 65°C for 5 min for first strand cDNA synthesis . After addition of 4 µl 1× first strand buffer , 1 µl DTT ( 0 . 1M ) , 1 µl RNase inhibitor ( RNaseOut 40 U/ µl ) and 1 µl Superscript III reverse transcriptase ( 200 U/µl ) ( all Invitrogen ) , the mixture was incubated for 5 min at 25°C , 60 min at 50°C and 15 min at 70°C . A negative control reaction was carried out with 1 µl of DNase RNase-free water instead of the SSIII enzyme . 50 µl of DNase RNAse-free water was added to the final cDNA before analyzing with real-time PCR . SYBR Green ( Kapa SYBR Fast Master Mix # KK4602 ) was used to quantify the level of expression . Relative quantification took place using the reference gene RPP0 for normalization . Real-time PCR ( 10 µl reactions ) included 1 µl of cDNA , 0 . 2 µl of forward primer ( 50 µM ) , 0 . 2 µl of reverse primer ( 50 µM ) , 5 µl of SYBR Green and 3 . 6 µl DNase RNase free water . Reactions were incubated in a Biorad CFX96 Real-Time PCR system in 96-well plates using the primers listed in table S2 . ChIP assays were performed as described previously [18] . Purified yeast Hmt1 ( 5 µg ) and 22 . 5 µg of biotinylated histone H4 peptides ( Cambridge Peptides , UK ) were incubated with 40 µl of MyOne Dynal Streptavidine beads T1 ( Invitrogen , #65601 ) in 100 µl total Reaction Buffer ( 20 mM Tris-Hcl pH 8 , 50 mM NaCl , 1 mM EDTA pH 8 , 5% Glycerol , 1 mM DTT , 2 mM S-adenosylmethionine and protease inhibitors ) for 20 hours at 30°C with shaking . The beads were then precipitated using a magnetic rack ( Invitrogen ) and resuspended in 10 µl SDS-loading buffer . The peptides were eluted by alternately boiling , cooling and vortexing the beads three times . The eluted samples were then analyzed by Western blotting and ponceau staining . Yeast cells were grown to mid-exponential phase in a 30°C shaker . Total yeast extracts were prepared by first resuspending cell pellets in a tenfold volume of SDS loading buffer ( 50 mM Tris-HCl pH 6 . 8 , 2% SDS , 10% glycerol , 1% β-mercaptoethanol , 12 . 5 mM EDTA and 0 . 02% bromophenol blue ) . The samples were then alternately boiled and chilled three times to rupture cell membranes . Proteins were separated in a 7 cm long , 17% SDS-PAGE ( Laemmli 1970 ) at 200 V for 1 h . The proteins were wet transferred into a PVDF membrane ( GE Healthcare life sciences ) with 20% Methanol transfer buffer ( 25 mM Tris , 192 mM glycine , pH 8 . 3 ) , at 100 V for 1 h . Before incubation with the appropriate antibody , the membrane was blocked in 5% BSA , 0 . 1% Tween-20 TBS buffer ( 25 mM Tris , 150 mM NaCl , 2 mM KCl , pH 8 ) . Synthesized peptides with at least 90% purity ( Cambridge Peptides , UK ) were dissolved in water , and drops containing 250 , 50 , or 10 pmol were deposited on a PVDF membrane , and allowed to air-dry for 1 h . The membrane was then submerged in 100% Methanol for a minute , water for another minute and then stained with Ponceau S or blocked as described above before probing with the appropriate antibody .
|
The genome of eukaryotic cells is packaged into nucleosomes consisting of an octamer of histone proteins that is wrapped around by DNA . Histone proteins are often modified with chemical groups that can influence the arrangement of nucleosomes and thereby affect DNA-based processes like transcription . Histone N-terminal acetylation , which comprises the addition of a chemical group at the tip of the histone tail , is an abundant modification whose function is unknown . In this work , we show that N-terminal acetylation of histone H4 can strongly inhibit the occurrence of a neighboring modification , namely dimethylation at the third arginine . To do this , N-terminal acetylation cooperates with other internal lysine acetylation marks . We find that the communication amongst these histone modifications is necessary for controlling the expression of ribosomal RNA genes that are required for protein synthesis and cell growth . Our experiments show that in the absence of both N-terminal acetylation and lysine acetylation there is a strong increase in H4 arginine 3 dimethylation levels leading to cell lethality . This growth defect can be rescued by a point mutation on H4 that blocks methylation at position 3 . Together , our results unveil a molecular and biological function for the previously uncharacterized N-terminal acetylation of histones .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
N-alpha-terminal Acetylation of Histone H4 Regulates Arginine Methylation and Ribosomal DNA Silencing
|
The epidemiology of Chagas disease , until recently confined to areas of continental Latin America , has undergone considerable changes in recent decades due to migration to other parts of the world , including Spain . We studied the prevalence of Chagas disease in Latin American patients treated at a health center in Barcelona and evaluated its clinical phase . We make some recommendations for screening for the disease . We performed an observational , cross-sectional prevalence study by means of an immunochromatographic test screening of all continental Latin American patients over the age of 14 years visiting the health centre from October 2007 to October 2009 . The diagnosis was confirmed by serological methods: conventional in-house ELISA ( cELISA ) , a commercial kit ( rELISA ) and ELISA using T cruzi lysate ( Ortho-Clinical Diagnostics ) ( oELISA ) . Of 766 patients studied , 22 were diagnosed with T . cruzi infection , showing a prevalence of 2 . 87% ( 95% CI , 1 . 6–4 . 12% ) . Of the infected patients , 45 . 45% men and 54 . 55% women , 21 were from Bolivia , showing a prevalence in the Bolivian subgroup ( n = 127 ) of 16 . 53% ( 95% CI , 9 . 6–23 . 39% ) . All the infected patients were in a chronic phase of Chagas disease: 81% with the indeterminate form , 9 . 5% with the cardiac form and 9 . 5% with the cardiodigestive form . All patients infected with T . cruzi had heard of Chagas disease in their country of origin , 82% knew someone affected , and 77% had a significant history of living in adobe houses in rural areas . We found a high prevalence of T . cruzi infection in immigrants from Bolivia . Detection of T . cruzi–infected persons by screening programs in non-endemic countries would control non-vectorial transmission and would benefit the persons affected , public health and national health systems .
Trypanosoma cruzi ( T . cruzi ) is a flagellate protozoan that causes Chagas disease ( CD ) . It is traditionally linked to rural areas of continental Latin America , where it is transmitted by a variety of bug vectors . In recent decades , the epidemiological pattern of this disease has undergone considerable changes [1] . In the endemic countries of Latin America , the regional Chagas programs are working to interrupt vector-borne and transfusional transmission , to control congenital Chagas disease and to support initiatives aimed at improving diagnosis , management and surveillance of the disease [2] . In non-endemic countries that receive immigrants from Latin America or send tourists to endemic areas , CD is an emerging disease and has become a public health problem because it can be transmitted by non-vectorial mechanisms [3] , [4] . Spain is a major European host country for people from Latin America . According to the Spanish National Institute of Statistics , in 2009 more than 1 . 8 million immigrants from Latin America were registered , accounting for 3 . 85% of the total population [5] . In recent years several studies of CD in non-endemic countries [6]–[9] have focused in particular on non-vectorial transmission mechanisms such as pregnancy and childbirth [10] , [11] , blood transfusion [12] , [13] and organ transplantation [14] , [15] . However , when reviewing the literature we found little information on imported CD in non-endemic countries at the primary care level [16] , which is ideal for screening the general population [17] . The clinical manifestations of chronic T . cruzi infection include the latent form ( the indeterminate chronic form ) , which occurs in 60% of cases [18] , the cardiac form [19] , the digestive or cardiodigestive form , and sudden death [20] . Therefore , many diagnoses of CD are based on epidemiological suspicion rather than clinical signs and symptoms . The objectives of the present study were ( 1 ) to assess the prevalence of Trypanosoma cruzi infection in the adult Latin American population treated at a health center in Barcelona , Spain; ( 2 ) to analyze the clinical phase of the disease; and ( 3 ) to determine whether screening for imported CD in primary care should be recommended .
We performed an observational , cross-sectional prevalence study at the health center of the Clot district , Barcelona . This center serves a population of 25442 people , with a total foreign population of 13 . 5% and a Latin American population of 6 . 3% ( according to the 2008 census of the Barcelona City Council ) [21] . The staffs participating in the study were 14 general practitioners , 13 nurses , 1 gynecologist and 1 midwife . The study protocol was approved by the Ethical Committee of the Jordi Gol Institute for Research in Primary Care of the Institut Català de la Salut ( Catalan Health Institute ) . Written informed consent was requested from all participants . When participants were children , their parents/guardians provided informed consent . During the period October 2007 to October 2009 , all patients from continental Latin America under 14 years of age who presented at the health center for any health reason were invited to participate in the study . After obtaining informed consent , we collected clinical and epidemiological data . We ascertained the reasons why the patients visited their doctor/nurse by reviewing the electronic patient charts . On a patient's first visit to a primary care centre the Preventive Activities and Health Promotion Program ( PAPPS ) is initiated [22] , and at the Clot health center this program included CD screening of all persons originating from continental Latin America . Serological screening was performed with an immunochromatographic test ( ICT ) that uses recombinant antigens of T . cruzi ( TcD , TcE , PEP-2 and SAPA ) on whole blood collected by finger prick . If the screening was positive , a venous blood sample was collected to confirm the diagnosis at the Parasitology Laboratory of the Faculty of Pharmacy , University of Barcelona . We used 2 enzyme-linked immunosorbent assay ( ELISA ) methods: a conventional , in-house ELISA ( cELISA ) with whole T . cruzi epimastigote antigens [23] and a commercial kit with the recombinant antigens TcD , TcE , PEP-2 and TcLo1 . 2 ( rELISA ) . In accordance with international criteria established by the World Health Organization , sera that were reactive in two serological methods were considered positive [24] . Positive results were confirmed by a third ELISA using T cruzi lysate ( Ortho-Clinical Diagnostics ) ( oELISA ) . In a subsample of 101 patients we performed the ELISA serologies ( cELISA and rELISA ) regardless of the result of the ICT , in order to test the usefulness of this test in screening for CD in primary care [25] . All patients infected with T . cruzi were referred to the Tropical Medicine Unit of Hospital Clínic de Barcelona and clinically evaluated by a complete review of the epidemiologic history and consistent symptoms/signs , a general physical examination and an electrocardiogram . If the electrocardiogram was pathologic , it was assessed with an echocardiogram or 24-hour Holter according to the disease detected . If symptoms consistent with gastrointestinal involvement were detected [26] , an esophageal , gastric and duodenal transit assessment or a barium enema was performed . Benznidazole ( 5 mg/kg/day for 60 days ) was offered to all patients aged 18–50 years without advanced Chagas cardiomyopathy and no other contraindication for start benznidazole ( pregnancy , severe renal or hepatic insufficiency [27] . None of them refused to start it . The sample size was calculated for an alpha level of 0 . 05 and a precision of ±0 . 05% in a bilateral comparison , assuming maximum uncertainty ( 50% prevalence ) ; for a population of 1516 subjects [28] a random sample of 758 was necessary . The programs SPSS version 17 . 0 and Epidat version 3 . 1 were used for the statistical analysis . The χ2 test was used to compare hypotheses of independence between two categorical variables and the Student t test for continuous variables . The confidence interval for all hypothesis comparisons was 95% and the tests were 2-tailed .
A total of 766 persons from continental Latin America were included in the study . The epidemiological data are presented in Table 1 and the countries of origin in Table 2 . Of the 766 patients analyzed , 27 were reactive to the ICT and 20 of these were reactive in cELISA , rELISA and oELISA . Also , 2 patients of the 101 tested by ICT , cELISA , and rELISa regardless of the result of the first were reactive in cELISA and rELISA . Both were also positive in oELISA , so they were considered positive . A total of 22 patients were diagnosed with CD , corresponding to a prevalence of 2 . 87% ( 95% CI , 1 . 6–4 . 12% ) in the sample studied . Of these , 21 were from Bolivia; the prevalence of CD in the subgroup of Bolivian patients studied ( n = 127 ) was 16 . 53% ( 95% CI , 9 . 6–23 . 39% ) . The remaining patient was from Paraguay . All the patients infected by T . cruzi were in the chronic phase of CD . The clinical form and the reasons why they visited the health center are presented in Table 3 . Four patients ( 18 . 2% ) had been previously diagnosed in the country of origin , but none of them mentioned it in the primary care visit because they thought it was a health problem proper to their country that would be unknown to the Spanish health staff ( this information was obtained when they were asked for informed consent to participate in the study ) . None of them were aware of their clinical phase and 2 patients had received incomplete treatment .
The prevalence of T . cruzi infection in the sample studied was 2 . 87% and in the subgroup of Bolivian patients it was 16 . 53% . In the medical literature we found few studies of similar characteristics to ours ( involving screening of the adult Latin American population in primary care ) and their results varied [29] , [12] due to the heterogeneity of the populations analyzed and the distribution of CD in Latin America . The laboratory confirmation of a clinical suspicion of CD is based on consistent results of at least 2 different immunological tests [24] . ICTs are attractive in primary care because they are easy to use in routine clinical practice and do not require sophisticated facilities or specialized staff . In the substudy that we performed in 101 patients [25] , for the ICT used we found a sensitivity of 92 . 5% and a specificity of 96 . 8% . Other studies have evaluated the sensitivity and specificity of ICTs [30] , [31] with similar results . The current sensitivity of ICTs must be increased so that they can be used as effective screening tests . Meanwhile , they should be combined with other methods that offer greater sensitivity [25] , [31] . The highest prevalence was found among Bolivian patients , in agreement with other studies performed in Spain [9] , [16] and other non-endemic countries [12] , [32] . No cases were diagnosed among the Peruvian or Ecuadorian patients , who formed 45% of the sample , probably due to the heterogeneous distribution of CD in endemic countries and the lower seroprevalence of T . cruzi estimated in Peru ( 0 . 69% ) and Ecuador ( 1 . 74% ) [33] . An epidemiologic history of having lived in rural areas and/or adobe houses showed a significant relationship with T . cruzi infection , consistent with the dominant vector-borne transmission mechanism in the countries of origin . All patients with T . cruzi infection had heard of CD in their countries of origin and approximately 82% knew someone who was affected . These data should be taken into account for establishing CD screening criteria in immigrants from endemic zones , because mere knowledge of the disease may be considered as an indirect indicator of its presence in the region of origin . In non-endemic countries CD screening programs have been aimed at particularly susceptible groups: in blood banks ( France [34] , the USA [35] and Spain [36] ) , and in pregnant Latin American women and their neonates ( the Spanish autonomous communities of Catalonia [37] and Valencia [38] ) . In our study only 9 . 5% of the patients with T . cruzi infection had visited the health center due to clinical symptoms suggestive of CD . As it is a silent disease that has recently appeared in non-endemic countries , we stress the importance of establishing in these countries health screening programs based on compatible epidemiologic history among the general immigrant population from endemic areas . These programs should be multidisciplinary [3] , supported by the best scientific evidence possible , and promoted by the health authorities . In non-endemic countries , detecting persons infected by T . cruzi is important in order to control the transmission ( vertical , by transfusion , or by organ transplant ) , reduce reactivations in immunodepressed persons , and delay the onset of the chronic cardiac form through antiparasite treatment [39] , all of which have a great impact on the persons affected , on public health , and on health systems . Nevertheless , the best solution for CD is a combination of treatment and prevention in endemic countries [40] , [41] , where many programs and initiatives are underway [2] , [42] . The data obtained in this study and the experiences described elsewhere [4] , [12] , [14] , [16] , [20] , [27] suggest that it is advisable to perform CD screening in non-endemic countries on all patients from continental Latin America who: ( 1 ) have a suggestive epidemiologic history ( having lived in a rural area , in adobe houses or having knowledge of CD in the country of origin ) , ( 2 ) are pregnant , ( 3 ) are immunosuppressed , ( 4 ) have symptoms suggestive of CD , or ( 5 ) request screening .
|
Chagas disease is a parasitic infection caused by the protozoan Trypanosoma cruzi , and is becoming an emerging health problem in non-endemic areas because of growing population movements . The clinical manifestations of chronic T . cruzi infection include the latent form ( the indeterminate chronic form ) , the cardiac form , the digestive or cardiodigestive form , and sudden death . Therefore , many diagnoses of Chagas disease are based on epidemiological suspicion rather than on clinical signs and symptoms . This study showed that the prevalence of Chagas disease in Latin American patients attending at a health center in Barcelona is 2 , 87% and the highest prevalence was found among Bolivian patients ( 16 , 53% ) . All the infected patients were in a chronic phase of Chagas disease . Detection of T . cruzi–infected persons by screening programs in non-endemic countries would control non-vectorial transmission and would benefit the persons affected , public health and national health systems . The data obtained in this study and the experiences described elsewhere suggest that it is advisable to perform Chagas disease screening in non-endemic countries on all patients from continental Latin America who: ( 1 ) have a suggestive epidemiologic history , ( 2 ) are pregnant , ( 3 ) are immunosuppressed , ( 4 ) have symptoms suggestive of Chagas disease , or ( 5 ) request screening .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"health",
"screening",
"public",
"health",
"and",
"epidemiology",
"chagas",
"disease",
"primary",
"care",
"neglected",
"tropical",
"diseases",
"public",
"health"
] |
2011
|
Chagas Disease among the Latin American Adult Population Attending in
a Primary Care Center in Barcelona, Spain
|
During the pachytene stage of meiosis in male mammals , the X and Y chromosomes are transcriptionally silenced by Meiotic Sex Chromosome Inactivation ( MSCI ) . MSCI is conserved in therian mammals and is essential for normal male fertility . Transcriptomics approaches have demonstrated that in mice , most or all protein-coding genes on the X chromosome are subject to MSCI . However , it is unclear whether X-linked non-coding RNAs behave in a similar manner . The X chromosome is enriched in microRNA ( miRNA ) genes , with many exhibiting testis-biased expression . Importantly , high expression levels of X-linked miRNAs ( X-miRNAs ) have been reported in pachytene spermatocytes , indicating that these genes may escape MSCI , and perhaps play a role in the XY-silencing process . Here we use RNA FISH to examine X-miRNA expression in the male germ line . We find that , like protein-coding X-genes , X-miRNAs are expressed prior to prophase I and are thereafter silenced during pachynema . X-miRNA silencing does not occur in mouse models with defective MSCI . Furthermore , X-miRNAs are expressed at pachynema when present as autosomally integrated transgenes . Thus , we conclude that silencing of X-miRNAs during pachynema in wild type males is MSCI-dependent . Importantly , misexpression of X-miRNAs during pachynema causes spermatogenic defects . We propose that MSCI represents a chromosomal mechanism by which X-miRNAs , and other potential X-encoded repressors , can be silenced , thereby regulating genes with critical late spermatogenic functions .
Meiotic sex chromosome inactivation ( MSCI ) describes the transcriptional silencing of the unsynapsed X and Y chromosomes at the onset of pachynema in mammalian male germ cells [1–5] . Inactivation of the sex chromosome results in the formation of a heterochromatic domain called the sex body [6] . MSCI is one example of a general mechanism , meiotic silencing , which inactivates any chromosome that is unsynapsed during male or female meiosis [7 , 8] . MSCI imposes a repressive chromatin signature on the X and Y chromosomes that is retained later , during spermiogenesis [9–12] . MSCI and its maintenance are regulated by a broad array of DNA double-strand break ( DSB ) repair proteins and chromatin modifications [2 , 5 , 13] . Male mice with chromosome abnormalities , e . g . XYY , or targeted mutations in meiotic synapsis or recombination genes , e . g . Spo11-/- and Brca1-/- frequently exhibit defective MSCI , and this results in misexpression of toxic sex-linked genes and midpachytene arrest [14–18] . Microarray [9 , 19] , RNA-sequencing [20] and RNA FISH [16] studies have concluded that in mice MSCI is robust , and no example of an X-linked protein-coding gene that is actively transcribed during pachynema has yet been identified . This situation contrasts with that later in spermatogenesis , when expression of some X-linked genes from the repressed X chromosome is facilitated by various mechanisms including gene amplification [16] and establishment of active chromatin marks by the ubiquitin ligase RNF8 [21] . However , the activity of X-derived non-coding RNAs , and especially small RNAs , during pachynema is less well understood . Interestingly , the X mouse chromosome is enriched in miRNA-encoding genes , and many of these are expressed in a testis-biased manner [22–24] . Song et al . conducted an extensive study of X- miRNA expression patterns , in which miRNA levels were evaluated by RT-qPCR in purified spermatogenic cell populations . 86% of X-linked miRNA transcripts were detected at high levels in pachytene spermatocytes , and it was been suggested that these genes escape MSCI [25] . High pachytene levels of X-miRNA transcripts have been confirmed by RT-qPCR [22 , 25] , RNA-sequencing [23 , 26 , 27] and in situ hybridisation [25 , 27] approaches . Non-coding RNAs have a prominent role in gene silencing , e . g . in X chromosome inactivation [28] , and repression of transposable elements and centromeric repeats [29] , and it is therefore possible that X-linked miRNAs contribute to the process of MSCI itself . However , definitive proof that X-miRNA genes escape MSCI requires that nascent precursors of miRNAs , so-called pri-miRNAs , are generated during pachynema , and ideally , that these can be visualised as nascent transcripts originating from the otherwise inactive X chromosome , e . g . by techniques such as RNA FISH . We therefore sought to reappraise X-miRNA expression in the male germ line focusing on nascent transcripts .
In order to establish whether X-miRNAs are subject to MSCI , we examined their expression during mouse spermatogenesis using RNA FISH ( Fig 1A–1G ) . We focused on spermatogonia , the early diploid germ cell progenitors in which the X chromosome is active , and pachytene spermatocytes , in which MSCI has taken place . There are currently 167 annotated miRNA genes on the X chromosome ( source: miRBase version 21 ) , the majority of which fall into clusters . We focused on six clusters , located at different sites on the X chromosome , and expressed in the testis ( S1 Fig ) [23] . Together these comprise 78 X-miRNAs , and 83% of them have been reported to escape MSCI [25] . We used a combination of antibody staining for the MSCI marker phosphorylated histone H2AFX ( γH2AFX ) [30] , as well as DAPI nuclear staining , in order to accurately substage germ cells . X-miRNA clusters 1 and 6 reside within introns of the genes Clcn5 and Ftx , respectively , and are transcribed from the same strand as the host genes ( Fig 1A and 1F ) . At miRNA cluster 1 and cluster 6 loci , no putative promoter other than the host gene promoters can be detected upstream the miRNA genes , as assessed by H3K4me3 signal ( S2 Fig ) , and expression of the miRNAs was shown to be dependent on the transcription of Clcn5 and Ftx parental RNAs [31 , 32] . Clcn5 and Ftx primary transcripts therefore represent the X-miRNA precursor transcripts . We used fluorescently-labelled , denatured fosmid DNA probes spanning the intronic Clcn5 and Ftx X-miRNA containing regions in order to detect X-miRNA precursor transcripts ( X-pri-miRNA ) . Cluster 1 and 6 pri-miRNA FISH signals were observed in spermatogonia ( 53% and 75% expressing , n = 32 and 51 cells , respectively ) . However , no pri-miRNA expression could be detected in pachytene spermatocytes ( 0% expressing , n = 100 cells; Fig 1A , 1F and 1G ) . Next , we used fosmid probes to examine expression of the remaining X-miRNA clusters 2 , 3 , 4 and 5 . X-miRNAs located within these clusters do not lie within host genes . For all four clusters , we observed pri-miRNA FISH signals in spermatogonia ( 23% , 8% , 26% and 14% expressing , n = 43 , 46 , 42 and 41 cells , respectively; Fig 1B–1E and 1G ) . In contrast , RNA FISH signals were not observed in pachytene spermatocytes for any of the four clusters ( 0% expressing for each cluster , n = 100 cells each in each case; Fig 1B–1E and 1G ) . The fosmids that we used for our RNA FISH experiments have an average size of 39kb . These probes will detect X-miRNA transcription , but could potentially also detect unannotated transcripts residing in the same locus . To exclude this possibility , we carried out two experiments . For cluster 5 , we used recombineering to excise a 7kb segment containing the X-miRNA genes from the fosmid probe . When the resulting , modified fosmid was used for RNA FISH , no signals were observed in spermatogonia ( 0% displaying signals , n = 41 cells; S3 Fig ) . Secondly , we designed an RNA FISH protocol to assess transcription of specific X-pri-miRNAs . In this approach , we used ~40 nucleotide-long probes matching sequences present in the pri-miRNA , but not the pre-miRNA or the mature miRNA , at the base of the miRNA-containing stem-loop sequence ( S4 Fig ) . We targeted the X-miRNA miR-465 , present in six copies in cluster 4 ( Fig 1D ) . Pri-miRNA signals were observed in spermatogonia but not in pachytene cells ( 0% expressing , n = 53 cells; S4 Fig ) . Thus , in conclusion , we observed transcription of all six X-miRNA clusters ( total 78 X-miRNAs ) in spermatogonia . However , we could not detect expression for any of these X-miRNAs during pachynema . The absence of cluster 1 to 6 X-pri-miRNA FISH signals in pachytene spermatocytes suggests that these genes are subject to MSCI . To test this possibility , we repeated our RNA FISH analysis on a mouse model in which MSCI is defective . In Spo11 null male mice , a domain of γH2AFX is formed at pachynema , but this rarely encompasses the X and Y chromosomes , and it is therefore termed the “pseudo sex body” [15 , 33] . The failure to execute H2AFX phosphorylation on the XY bivalent causes misexpression of sex-linked genes during pachynema in this mutant [16] . We performed pri-miRNA FISH for four representative X-miRNA clusters: 1 , 3 , 4 and 6 ( Fig 2 ) . Pachytene spermatocytes were identified in Spo11 null males by the presence of the γH2AFX-labelled pseudo sex body . Notably , in Spo11 null pachytene spermatocytes we observed pri-miRNA FISH signals for all four gene clusters studied . Expression was observed in 14% , 96% , 98% and 28% of pachytene cells for clusters 1 , 3 , 4 and 6 , respectively ( n = 100 cells in each case; Fig 2A ) . We subsequently repeated the analysis of cluster 4 X-miRNAs using our oligonucleotide RNA FISH approach that specifically detects pri-miRNAs for the six copies of miR-465 . We observed FISH pri-miRNA signals in 93% of Spo11 null pachytene cells ( n = 69 cells; Fig 2B ) . In addition , we performed miR-465-specific pri-miRNA FISH in a second MSCI mutant , the Brca1 Δ11/Δ11 model [17 , 18] . We detected misexpression of this miRNA in 87% of pachytene spermatocytes . ( n = 15 cells; Fig 2B ) . We conclude that defective MSCI leads to X-miRNA misexpression during pachynema . Although MSCI is defective in most Spo11-/- and Brca1 Δ11/Δ11 pachytene cells , in both models domains of γH2AFX are occasionally seen covering sub-regions of the X chromosome [16 , 17] . We predicted that in these rare spermatocytes , X-miRNAs encompassed within γH2AFX regions should be normally silenced . This proved to be the case: in the few Spo11 null and Brca1 pachytene cells in which the miR-465 locus , identified using DNA FISH , lay within a γH2AFX domain ( Spo11 null: n = 2 out of 54 cells; Brca1Δ11/Δ11: n = 2 out of 15 cells ) , no pri-miR-465 expression could be observed ( Fig 2C ) . Thus , the expression status of these X-miRNAs is tightly linked to the presence of the meiotic silencing marker γH2AFX . We conclude that in wild type pachytene spermatocytes , the X-linked miRNAs studied herein are silenced during pachynema as a result of MSCI . To corroborate our X-miRNA FISH data , we next compared expression levels of individual pri-miRNAs in wild type and Spo11 null sibling testes by RT-qPCR at 15 . 5days post-partum ( dpp; Fig 2D ) . At this age , most spermatocytes are in pachynema , and genes subject to MSCI are expected to be overexpressed in Spo11 null relative to wild type males . We examined transcript levels for a number of X-linked and autosomal pri-miRNAs , and expressed these as a Spo11 null / wild type ratio . Experiments were performed in triplicate , each time using a different Spo11 null and wild type sibling . In each case , Spo11 null / wild type ratios for autosomal miRNAs averaged ca . 1 , indicating no difference in pachytene expression levels between the two genotypes ( Fig 2D ) . Conversely , the ratio for X-linked pri-miRNAs significantly exceeded one ( p = 3 . 361e-08; Fig 2D ) , thereby confirming that X-linked miRNAs are upregulated in the absence of MSCI . Finally , we used transgenesis to further investigate whether silencing of X-miRNAs in pachynema is due to MSCI . Previous experiments have demonstrated that X-genes present as transgenes on autosomes continue to be expressed during pachynema [34] . This is because unlike the X chromosome , autosomes are synapsed during pachynema and therefore escape the effects of meiotic silencing . To establish whether pachytene silencing of X-linked miRNA genes was due to their location on the X chromosome , we generated a single copy transgenic line in which X-linked miRNA gene clusters 3 and 4 were located together on an autosome by random BAC integration ( X-miRBAC line 1; Fig 3A ) . We chose a BAC that includes a region of local H3K4me3 enrichment upstream of the miRNA gene cluster , indicative of a putative promoter ( Fig 3A ) . Using pri-miRNA microarrays , we confirmed that cluster 3 / 4 X-miRNAs were overexpressed in X-miRBAC line 1 testes relative to non-transgenic siblings ( Fig 3B ) . We then performed pri-miRNA FISH in pachytene spermatocytes from X-miRBAC line 1 transgenics using BAC probes covering clusters 3 and 4 . We observed expression of cluster 3/4 miRNAs from both the X chromosome and the autosomal transgene prior to pachynema ( Fig 3C ) . However , during pachynema , while the X-located 3/4 miRNAs were silenced , those located on the transgene continued to express ( 100% , n = 50 ) . We observed the same results using our miR-465-specific pri-miRNA FISH protocol on X-miRBAC line 1 ( S5 Fig ) ( 100% , n = 9 ) . Thus , silencing of X-integrated miRNAs during pachynema is due to MSCI . Defects in MSCI cause pachytene arrest , due to misexpression of toxic sex-linked genes , e . g . the Y chromosome genes Zfy1 and Zfy2 [14] . Our analyses indicated that X-miRNAs are subject to MSCI . We therefore wondered whether misexpression of these genes during pachynema would give rise to spermatogenic defects . Interestingly , in our X-miRBAC line 1 males , which carry the autosomally-integrated single copy X-miRNA 3 and 4 cluster transgene , we observed reduced testis weights relative to non-transgenic brothers from as early as five weeks post-partum ( Fig 4A ) . Importantly , histological and TUNEL analysis of X-miRBAC line 1 testis sections revealed spermatogenic defects , principally germ cell apoptosis at stage IV , corresponding to midpachynema , and stage XII , corresponding to the meiotic divisions ( Fig 4B , S7 Fig ) . In order to exclude the possibility that the spermatogenic defects observed in X-mirBAC line 1 males resulted from a transgene integration effects , we subsequently generated two more cluster 3/4 X-miRNA autosomal transgenic lines , with three ( X-miRBAC line 2 ) and eleven ( X-miRBAC line 3 ) transgene copies ( S6 Fig ) . X-miRBAC line 2 showed predominant apoptosis at stage XII ( Fig 4C , S7 Fig ) , while X-miRBAC line 3 exhibited marked apoptosis at mid and late pachynema ( Fig 4D , S7 Fig ) . We conclude that inappropriate expression of X-miRNAs from the X-linked clusters 3 and 4 at pachynema induces spermatogenic defects .
MSCI is a robust silencing process , affecting most or all protein-coding genes on the mouse X chromosome . However , it is unclear whether silencing also affects X-linked miRNAs . Here , using RNA FISH and other transcriptional assays , we find that X-linked miRNA genes are expressed before prophase I but are silent during pachynema . We therefore conclude that X-miRNAs are subject to MSCI . It is important to highlight that we did not study all miRNAs on the X chromosome . It is therefore formally possible that X-miRNAs omitted in our analyses behave differently with respect to MSCI . We find this unlikely , because we chose miRNAs from multiple , distinct clusters on the X chromosome , and we included many X-miRNAs that were previously reported to escape silencing [25] . A priori , one could also argue that our inability to detect X-miRNA FISH signals during pachynema is because our RNA FISH experiments lack the sensitivity required to detect gene expression during this stage of prophase I , rather than because these genes are subject to MSCI . We doubt that this is the case , because we were able to detect expression of these species during pachynema both in MSCI mutants , and in mice carrying autosomally-located X-miRNA transgenes . Finally , our conclusion that X-miRNAs are subject to MSCI was corroborated by RT-qPCR analysis in MSCI mutants versus controls . Taken together , our data support a model in which X-miRNAs behave like X-linked protein-coding genes with respect to X silencing . How can our findings accommodate earlier work ? Several independent reports have documented high levels of X-miRNA expression in pachytene spermatocytes [22 , 23 , 25 , 26] , and there can be little doubt that this contrasts with the generally low level expression detected for protein-coding X-genes . However , in most existing studies X-miRNAs were assayed at the level of mature miRNAs . Notably , with the exception of some rare cases [35 , 36] , miRNAs have an unusually long half-life , on average 5 days , which exceeds that of protein-coding RNAs by ten-fold [37] . The abundant expression of miRNAs during pachynema might therefore be due to their high transcript stability , rather than ongoing generation of nascent miRNA precursors . Interestingly , our work shows not only that X-miRNAs are subject to MSCI , but also that failure to silence them can result in spermatogenic defects which are manifest in the case of the cluster 3/4 X-miRNAs as arrest predominantly at stages IV and XII . Thus , X-miRNAs join the Y-encoded Zfy1/2 genes as being male “pachytene-lethal” genes . Our findings show that MSCI must be extensive , silencing genes not only on the Y chromosome but also on the X chromosome . Given that miRNAs act as gene repressors , the phenotypes resulting from their ongoing expression in our miRNA transgenic males are presumably due to inappropriate target downregulation as a consequence of miRNA overexpression , or to an inability to appropriately upregulate target genes with meiotic and/or post-meiotic functions . In this model , MSCI could function as a chromosome-based mechanism for regulating expression of repressors . From a broader perspective , the large-scale silencing of genes across the X chromosome by MSCI is likely to influence myriad transcriptomic networks within germ cells . As a consequence , MSCI could regulate multiple facets of the mammalian germ cell development program .
All animal procedures were in accordance with the United Kingdom Animal Scientific Procedures Act 1986 and were subject to local ethical review . All mice were maintained on an MF1 background . The miRBAC transgenic lines were produced by microinjection of purified BAC BMQ-333E20 into fertilized eggs from CBA/Ca x C57Bl/10 F1s . Spo11 null and Brca1Δ11/Δ11 mice have been described previously [18 , 38] . Browser tracks were generated in R using the Givz package . The UCSC mouse genome assembly mm10 was used as a basis for analysis . Annotation of transcripts was obtained from the UCSC knownGene database . Genomic coordinates of fosmid and BAC probes were obtained from the CHORI BACPAC resource center . A small RNA testis library was downloaded from GEO under accession identifier GSE40499 ( Meunier et al . ) . Adapter sequences were removed from the reads ( ATCTCGTATGCCGTCTTCTGCTTG ) , and 15 to 23nt-long reads were selected for analysis . Reads were aligned to the mouse mm10 genome using Bowtie ( version 1 . 6 . 0 ) with the parameters -m 50—best—strata -v 2 . MiRNA gene coordinates were obtained from miRBAse ( version 21 ) . MiRNA duplicates sharing a copy on the X chromosome and a copy on an autosome were removed from the analysis . MiRNA read counts were generated in R using the QuasR package as documented in the reference manual [39] . Read counts are expressed as read counts per million reads mapping to miRNA genes . RNA and DNA FISH was carried out with digoxigenin- and biotin-labelled probes respectively , using fosmid and BAC genomic clones ( cluster 1: WI1-603H11; cluster 2: WI1-1995I23; cluster 3: WI1-1646F11; cluster 4: WI1-2045C16; cluster 5: WI1-2828J23; cluster 6: WI1-2859G17; miRBAC: BMQ-333E20 ) . The technique was described previously [40] . For miR-465-specific pri-miRNA FISH , a mix of nine amino-allyl-modified oligonucleotides labeled with fluorolink Cy3 were used as probes ( S1 Table ) . We used the anti-γH2AFX antibody ( Upstate , 16–193; dilution 1/100 ) for immunofluorescence post-RNA FISH . Total RNA was extracted from frozen testis tissues with Trizol ( Invitrogen ) , treated with DNAse I ( Invitrogen ) , and reverse transcribed with random hexamers ( Invitrogen ) and Superscript II ( Invitrogen ) according to the manufacturer’s instructions . Quantitative PCR was performed with pre-designed Taqman pri-miRNA and U6 snRNA Taqman assays according to the manufacturer's instructions . Relative expression was calculated with the ΔCt method using U6 as a normaliser . Pri-miRNA expression ratio for individual pri-miRNAs are provided in S2 Table . A library of H3K4me3 ChIP in adult germline precursor cells and corresponding input controls were downloaded from GEO under accession identifier GSE49624 [41] . Reads were aligned to the mouse mm10 genome using Bowtie ( version 1 . 6 . 0 ) with the parameters Bowtie -m 1—best –strata . Affymetrix Mouse miRNA 2 . 0 microarrays were performed to measure miRNA expression in testis of three wild type and two homozygous transgenic siblings at 37dpp . Total RNA was extracted with Trizol ( Invitrogen ) , treated with DNAse I ( Invitrogen ) , and column-purified ( Ambion ) . Microarray hybridizations were performed according to the manufacturer's instructions . Microarray signal intensity was extracted and normalized using Affymetrix' miRNA QC Tool using default parameters . Statistical analyses were performed using R and the limma package . Fold-changes and FDR-adjusted p-values were computed by fitting a linear model for each microRNA . Standard errors were smoothed using empirical Bayes ( eBayes function of the limma package ) . The transgene copy number of the miRNA BAC transgenic line was estimated by qPCR on genomic DNA , with a technique adapted from the one described in . The data was normalised with Atr PCR for ΔCt calculations , and quantification of Jun copy number was used as a quality control . Primer sequences are provided in S3 Table . Recombination-mediated genetic engineering of fosmid WI1-2828J23 was performed to delete a 7kb fragment encompassing miRNA genes miR-201 and miR-547 using standard procedures . The primers used for recombineering are
|
During male germ cell formation , the X and the Y chromosomes are inactivated . This process is conserved and it is essential for germ cell generation . It is believed that X/Y silencing affects all protein-coding genes , but the status of miRNAs and other non-coding genes needs further investigation . MicroRNAs from the X-chromosome ( X-miRNAs ) have been reported as potential silencing escapers , and they have been proposed to play a role in the inactivation mechanism itself . By looking at the individual cell level , we show unambiguously that X-miRNAs are subject to X/Y silencing , a finding that contradicts the current literature . Moreover , we generated mouse mutants in which we forced expression of X-miRNAs during X/Y silencing , and this lead to germ cell death . We propose that X/Y silencing can influence transcription of essential germ cell genes by regulating X-repressors .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Silencing of X-Linked MicroRNAs by Meiotic Sex Chromosome Inactivation
|
An important goal in evolutionary biology is to understand the genetic changes underlying novel morphological structures . We investigated the origins of a complex wing pattern found among Amazonian Heliconius butterflies . Genome sequence data from 142 individuals across 17 species identified narrow regions associated with two distinct red colour pattern elements , dennis and ray . We hypothesise that these modules in non-coding sequence represent distinct cis-regulatory loci that control expression of the transcription factor optix , which in turn controls red pattern variation across Heliconius . Phylogenetic analysis of the two elements demonstrated that they have distinct evolutionary histories and that novel adaptive morphological variation was created by shuffling these cis-regulatory modules through recombination between divergent lineages . In addition , recombination of modules into different combinations within species further contributes to diversity . Analysis of the timing of diversification in these two regions supports the hypothesis of introgression moving regulatory modules between species , rather than shared ancestral variation . The dennis phenotype introgressed into Heliconius melpomene at about the same time that ray originated in this group , while ray introgressed back into H . elevatus much more recently . We show that shuffling of existing enhancer elements both within and between species provides a mechanism for rapid diversification and generation of novel morphological combinations during adaptive radiation .
One of the major impediments to evolutionary innovation is the constraint on genetic change imposed by existing function [1] . Mutations that confer advantageous phenotypic effects in a novel trait will often result in negative pleiotropic effects in other traits influenced by the same gene . Several mechanisms have been proposed by which evolution can circumvent such constraints , resulting in phenotypic diversification . In particular , the modularity of cis-regulatory elements [2–6] means that novel modules can encode new expression domains and functions without disrupting existing expression patterns [6 , 7] . This modularity underlying gene regulation has led to the assertion that much of morphological diversity has arisen through regulatory evolution [6] . Much of our understanding of modularity in regulatory evolution comes from Drosophila , in which the loss of trichomes on the larval cuticle [5] , the gain of melanic wing spots [8–10] , or changes in abdominal pigmentation [3 , 11] have been shown to involve evolutionary changes in cis-regulatory elements . These elegant developmental studies demonstrate the underlying logic of regulatory modularity , whereby novel expression domains can arise without disrupting existing function . These studies have also established a paradigm in which small effect mutations alter transcription factor binding sites in these regulatory modules and in combination produce large effect alleles [5] . Similar conclusions come from recent work in other taxa , including mice and jewel wasps [2 , 12] . This might seem to imply that the evolution of novel regulatory alleles is relatively gradual , requiring the evolution of many small effect substitutions , but recent adaptive radiations can show extremely rapid rates of morphological change . The role of regulatory modularity therefore remains to be tested in adaptive radiations in which morphological variation evolves very rapidly . Here we explore the origins of adaptive novelty among the wing patterns of Heliconius butterflies . These wing patterns are under strong natural selection for mimicry and warning colour , as well as being important mating signals [13] . The rapid radiation in Heliconius is accompanied by an even more rapid diversification in mimicry patterns as well as convergence among species found in a given locality [14] , both through independent convergent evolution and via introgression of gene regions between races and species [15 , 16] . Mimetic convergence reaches its peak among red dennis-ray pattern phenotypes in the Amazon ( Fig 1 ) , where 11 or more Heliconius species , as well as pierine butterflies and pericopine moths , share the same pattern . In addition to near perfect convergence in wing patterns in a given locality , there is also often striking divergence of patterns between localities as populations adapt to the many different mimicry complexes spread across the Neotropics [17] . This diversity provides an opportunity to study the genetic and developmental basis of evolutionary novelty . Generally , the patterns on butterfly wings are a good system in which to link genetic changes to the developmental processes that generate diversity [18 , 19] . Wing colour patterns are mosaics of scales , each with a single colour , produced by a combination of pigment and ultrastructure . The relative positions of differently coloured scales are established during larval and pupal wing development [20] . Wing development is thought to be broadly conserved in insects , with wing developmental genes showing similar expression patterns between flies and butterflies [21–23] . This therefore raises the question: how is this conserved landscape of wing development translated into the diversity of butterfly wing patterns ? In Heliconius , pattern diversity is controlled by a surprisingly small number of genomic regions with large effect sizes [24 , 25] . In particular , genetic mapping and gene expression studies have shown that red elements are associated with expression of the transcription factor optix across all Heliconius species [26 , 27] . In the absence of fixed coding sequence changes between wing pattern forms , this implies that red pattern variation is controlled by differential regulatory control of optix [27] . Population genomic studies have identified a region of non-coding sequence downstream of optix that is associated with phenotypic change [15 , 28] . Previous work suggests that there may be several distinct elements within this region . Occasional hybrid phenotypes possess only the “dennis” patch on the base of the forewing or the “ray” elements on the hindwing and have been hypothesised to be rare recombinants , although this has never been tested genetically [29 , 30] . Similarly , there are also established forms that exhibit only dennis or ray patterns ( H . melpomene meriana and H . timareta timareta f . contigua , respectively , Fig 2 ) . This suggests that the broad genomic interval already identified might contain discrete regulatory loci that vary the spatial expression of optix in different wing regions , a hypothesis that we can now test with genetic data . Here we focus on the H . melpomene lineage , in which the Amazonian dennis-ray phenotype has evolved recently from a red-banded ancestor [31] . We carry out a population genomic analysis on H . melpomene and its relatives , H . elevatus and H . timareta , to identify putative regulatory modules associated with distinct red pattern elements . Previously , population genetic evidence has suggested that mimicry among H . melpomene , H . elevatus , and H . timareta has evolved through sharing of the dennis-ray allele by repeated adaptive introgression at the optix locus [16] . This is especially surprising in the case of H . elevatus , which forms part of the “silvaniform” clade that diverged from H . melpomene around 4 million years ago [14] . Our analysis here indicates that the origin of the red pattern elements is considerably more complex than has been previously supposed , with the dennis and ray elements of the widespread dennis-ray pattern having distinct evolutionary origins in different clades within the genus .
We took advantage of natural phenotypic variants in which the two red elements , dennis and ray , occur separately to identify putative functional regulatory regions controlling red pattern within the H . melpomene clade . Genomic analysis of 96 individuals from the melpomene-timareta clade revealed two distinct regions that showed strong association with the dennis and ray pattern elements , respectively . Our analysis included a race of H . melpomene , H . m . meriana , from the Guiana shield , which possesses the forewing dennis patch but not ray , as well as H . t . timareta f . contigua from Ecuador , which possesses ray but not dennis , plus a recombinant individual from an H . melpomene hybrid zone in Ecuador with dennis but not ray ( Fig 2A ) . Across all 96 individuals , there were significant genotype-by-phenotype associations across all genome regions surveyed . This “background” signal of genotype-by-phenotype association is likely due to the presence of genetically divergent species in our dataset that are to some degree confounded with phenotype . Nonetheless , our analysis identified a peak of genotype-by-phenotype association spanning roughly 50 kb and located from 60–110 kb 3ʹ of the optix gene , similar to what has been observed previously ( Fig 2B ) [28] . This region also corresponds closely to that identified recently in the mimetic species H . erato [15] , implying convergence in the regulatory architecture controlling wing pattern mimicry at a finer scale than has been previously demonstrated [28 , 31] . Furthermore , within this region in our data , distinct adjacent peaks of association were observed for the dennis and ray elements . Focusing specifically on fixed single nucleotide polymorphism ( SNP ) differences between alternative red phenotypes revealed two distinct peaks of association ( Fig 2C ) . One , approximately 10 kb in length , contained SNPs perfectly associated with the red dennis patch . The other adjacent region was broader , roughly 25 kb , and contained SNPs perfectly associated with red hindwing rays . We next used broader taxonomic sampling to further refine these intervals and identify exact sequence haplotypes associated with each of the two phenotypic elements ( Fig 2D ) . To identify recombination breakpoints around dennis and ray haplotypes , we generated a high-quality sequence alignment by de novo assembly of each individual genome and then identified contigs across the associated region using the Basic Local Alignment Search Tool ( BLAST ) . For the dennis region , alignment was assisted by a sequenced fosmid clone from H . m . aglaope ( dennis phenotype ) to complement the reference genome ( derived from a non-dennis butterfly ) . The final alignment included the 96 melpomene-timareta individuals used for association analysis and a further 46 individuals that included species with no red ( H . cydno ) and species from the more distantly related silvaniform clade including H . elevatus , which has the dennis-ray pattern . The distal end of the dennis region , relative to optix , was delineated by a rapid loss of phenotype-associated variants across all species sampled , whilst the proximal end was determined by a single fixed recombination event in the race H . m . meriana ( dennis but no ray phenotype ) , generating a region of ~7 kb fully associated with dennis . For ray , a breakpoint in the ray-only H . t . timareta f . contigua defined the distal end , whilst a recombination in H . m . meriana defined the proximal end , resulting in a larger ~37 kb region ( Fig 2D ) . Each haplotype group was characterised by diagnostic SNPs as well as a fixed architecture of indel variation ( Fig 3 ) . These analyses therefore support the hypothesis derived from phenotypic evidence , that dennis and ray phenotypes are controlled by adjacent distinct genetic elements . In combination with previous work showing differential expression of optix across a wide diversity of Heliconius species and races , this provides clear genetic evidence for modularity in the cis-regulatory control of optix . We have previously hypothesised that the dennis-ray mimicry pattern introgressed as a single genomic block between H . melpomene and H . timareta , as well as more distantly between H . melpomene and H . elevatus [16] . Our new data suggest a much more complex history than previously recognised , with dennis and ray having quite distinct origins . As expected , a maximum likelihood ( ML ) phylogeny shows that the ray alleles fall within the H . melpomene clade , indicating an origin derived from an ancestral H . melpomene phenotype . In contrast , however , alleles producing the dennis phenotype originated within the silvaniform clade , which diverged from H . melpomene around 4 million years ago ( Fig 4 ) [14] . Members of this clade have mottled orange/red , black , and yellow “tiger” patterns and are mostly co-mimics of butterflies in the tribe Ithomiini , whereas the melpomene-cydno clade are all co-mimics of other Heliconius species . Nonetheless , the silvaniforms commonly have orange patches on the base of the forewing , which in some cases are remarkably similar to the dennis patch of H . melpomene . In particular , the form H . hecale metellus has a dennis-like phenotype ( Fig 4A ) , which suggests a plausible ancestral phenotype that might have provided the source of the dennis allele in H . melpomene . Various scenarios might explain this complex history . Sharing of variation between species can be explained by either retention of ancestral polymorphism or introgression through hybridisation . We can directly test these alternative scenarios using dated trees inferred from our alignments . In order to provide comparable trees , we used the divergence date between the silvaniform and melpomene-cydno clades derived from a recent species tree for the Heliconinii [14] to calibrate the dennis and ray region phylogenies . These trees support introgression and rule out ancestral polymorphism because the dates of coalescence of the H . melpomene and H . elevatus dennis and ray alleles are significantly more recent than the divergence of these two species . These species last shared a common ancestor at around 3 . 96 Ma ( 95% highest posterior density [HPD] interval 3 . 18–4 . 81 Ma ) [14] . In contrast , the dennis allele shared between H . elevatus and H . melpomene/timareta diverged around 1 . 95 Ma ( 2 . 79–1 . 25 Ma HPD ) . The divergence of the ray allele is even more recent and shared a common ancestor between H . elevatus and H . melpomene/timareta around 0 . 66 Ma ( 0 . 93–0 . 43 Ma HPD ) . The recent origin of these alleles is also supported by low levels of genetic diversity within these clades , with the average pairwise sequence divergence among the dennis alleles only 1 . 5% , including those from H . melpomene , H . timareta , and the more distantly related H . elevatus ( Table 1 , top ) . This is less than that found among the same individuals in flanking sequence ( 2 . 4% ) and comparable to that among the red-forewing-banded “postman” group for the same locus ( 1 . 6% ) , which includes only more closely related melpomene and timareta individuals . The ray alleles also show only 1 . 1% average pairwise sequence divergence at the ray locus , similarly less than in the postman group at the same locus . Although sequence diversity is likely to be reduced in these regions because of functional constraint , it seems likely that such constraint is similar across different clades in the phylogeny , so the relatively low levels of diversity within the ray and dennis clade support their recent origin . Our dated trees can also be used to infer the relative timing of introgression events . Here the data indicate that the ray allele originated within H . melpomene at around the same time as dennis introgressed into the H . melpomene clade from an ancestor of H . elevatus , sometime around 1 . 85 Ma ( 2 . 53–1 . 25 Ma HPD; Fig 5 and S3 Fig ) . This suggests that the characteristic mimetic dennis-ray phenotype first came together within H . melpomene at that time . In contrast , H . elevatus did not acquire the ray allele until about a million years later and perhaps persisted during this time as part of the Guiana Shield dennis-only mimicry ring . The dennis and ray alleles of H . timareta are each nested within those of H . melpomene and are more recent than the divergence of these species , implying introgression from H . melpomene into H . timareta and consistent with previous analyses [33] . Nonetheless , dennis and ray events also differ in timing , as most of the H . timareta ray alleles diverge from their H . melpomene relatives around 1 Ma , but the dennis alleles diverged only around 0 . 45 Ma ( Fig 5 and S3 Fig ) . H . timareta ray alleles are polyphyletic with respect to H . melpomene , also supporting multiple introgression events and recombination between the regions . Further sampling will be needed to resolve more clearly the timing and number of introgression events between these species . In addition to recombination between lineages , there is also shuffling of alternate alleles at these regulatory modules within species . Across most of their range , H . melpomene and H . timareta have either postman or dennis-ray haplotypes across the entire studied region . However , the race H . m . meriana has dennis alleles but shows recombination across the adjacent ray locus that removes the ray phenotype . Similarly , a single recombination event in H . t . timareta f . contigua produced a phenotype with ray but not dennis ( Fig 2D; Fig 4 ) . H . elevatus , like H . melpomene , also has a dennis-only race found in the Guiana Shield , which likely represents another case of enhancer shuffling within species , although we have not sampled this species here ( Fig 4 ) . Hence , although alleles within these regulatory modules are now highly divergent and presumably arose through accumulation of a number of mutations of small effect , novel phenotypes could have arisen rapidly through recombination between modules both within and between species . We have demonstrated several aspects of the genetic architecture of wing pattern that have contributed to evolutionary innovation in the Heliconius radiation . First , distinct genetic elements are associated with different patches of red on the butterfly wing . This supports the hypothesis of regulatory modularity , which should facilitate evolutionary innovation . Second , the origin of the dennis-ray phenotype in H . melpomene involved a combination of evolutionary tinkering of existing patterns and introgression between species . Finally , we show that diversity within three lineages ( H . elevatus , H . melpomene , and H . timareta ) has been generated by shuffling of these distinct regulatory modules among populations and species . Within all three lineages , some populations possess one or other of these elements , providing further flexibility in pattern evolution . Our data imply that recombination between lineages can generate novel phenotypic combinations and demonstrate how modularity in the cis-regulatory control of key genes can drive the rapid evolution of novel morphologies . Although the evolution of novel regulatory modules may involve many mutational steps [5] , these can subsequently be exchanged between lineages and shuffled into new combinations enabling rapid adaptive evolution . Recent studies showing that adaptation can proceed via gene flow of preadapted genetic modules between nearby populations or species suggest that similar mechanisms may be important in other radiations . In sticklebacks , adaptation to freshwater involves movement of alleles through the marine landscape [34] . Mosquitoes , Darwin’s finches , and even humans also show evidence for introgression of alleles between species that facilitate adaptation [35–37] . The extent to which recombination between regulatory alleles can contribute to morphological novelty in these other groups of organisms remains to be seen .
The paired-end sequencing reads for the 96 individuals used above plus an additional 43 individuals mainly representing outgroup species ( S3 Table ) were imported into CLC Genomics Workbench v5 . 5 and de novo assembled into contigs using default parameters ( mismatch: 2; insertion: 3; deletion: 3; length fraction: 0 . 5; similarity fraction: 0 . 8 ) . The resulting contigs were imported into Geneious v6 . 1 as FASTA files . These were used to construct BLAST databases in Geneious for each individual . Two further individuals were sampled in the form of the fully assembled reference sequences for dennis and ray regions , which were acquired from the H . melpomene reference genome v . 1 . 1[16] and the fosmid 1048-3N15 from H . m . aglaope , which provided a reference sequence for the dennis haplotype . In Geneious , these sequences were then used as references to conduct BLASTn searches against the contig sequence databases for each of the 139 de novo assemblies . The BLAST results were then mapped back onto the reference sequence . Using the reference as a template , matching contigs were concatenated into a single FASTA file for each individual , with Ns filling regions between contigs spaced according to the reference ( or closest relative ) . Unresolvable genomic repeats were detected as regions in which more than two haplotypes matched by BLAST and were replaced with Ns . Heterozygous haplotypes , in which just two contigs aligned , were assigned to two distinct FASTA versions of the region . Phasing of adjacent haplotypes was arbitrarily assigned , except for individuals heterozygous for phenotype , such as hybrid H . melpomene , in which heterozygous contigs could be clearly assigned to either dennis-ray or banded clades . In total , the final alignment included sequence derived from 142 individuals , including 139 de novo assembled genomes , two fosmid clones , a reference BAC sequence , and the reference genome .
|
Butterflies show an amazing diversity of patterns on their wings . In fact , most of the 18 , 000 species of butterfly can be distinguished on the basis of their wing pattern . Much of this diversity is thought to arise through novel switches in the genome that turn genes on in new contexts during wing development , thereby producing new patterns . Here we study a set of switches that control the expression of optix , a gene that places red patches onto the wings of Heliconius butterflies . We show that two patterning switches—one that produces red rays on the hindwing and the other a red patch on the base of the forewing—are located adjacent to one another in the genome . These switches have each evolved just once among a group of 16 species but have then been repeatedly shared between species by hybridisation and introgression . Despite the fact that they are now part of a common pattern in the Amazon basin , these two pattern components actually arose in completely different species before being brought together through hybridisation . In addition , recombination among these switches has produced new combinations of patterns within species . Such sharing of genetic variation is one way in which mimicry can evolve , whereby patterns are shared between species to send a common signal to predators . Our work suggests a new mechanism for generating evolutionary novelty , by shuffling these genetic switches among lineages and within species .
|
[
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2016
|
Evolutionary Novelty in a Butterfly Wing Pattern through Enhancer Shuffling
|
Neonates , including mice and humans , are highly susceptible to cytomegalovirus ( CMV ) infection . However , many aspects of neonatal CMV infections such as viral cell tropism , spatio-temporal distribution of the pathogen as well as genesis of antiviral immunity are unknown . With the use of reporter mutants of the murine cytomegalovirus ( MCMV ) we identified the lung as a primary target of mucosal infection in neonatal mice . Comparative analysis of neonatal and adult mice revealed a delayed control of virus replication in the neonatal lung mucosa explaining the pronounced systemic infection and disease in neonates . This phenomenon was supplemented by a delayed expansion of CD8+ T cell clones recognizing the viral protein M45 in neonates . We detected viral infection at the single-cell level and observed myeloid cells forming “nodular inflammatory foci” ( NIF ) in the neonatal lung . Co-localization of infected cells within NIFs was associated with their disruption and clearance of the infection . By 2-photon microscopy , we characterized how neonatal antigen-presenting cells ( APC ) interacted with T cells and induced mature adaptive immune responses within such NIFs . We thus define NIFs of the neonatal lung as niches for prolonged MCMV replication and T cell priming but also as sites of infection control .
CMV infection shows an extraordinary high prevalence worldwide which increases with age [1] , [2] , but the majority of infected humans stays asymptomatic . Clinical symptoms dominate in neonates who suffered from congenital infection , postnatal infection of preterm low birth-weight infants , or in immuno-compromised adults [1] , [3] , [4] , [5] , [6] . The prevalence of CMV infection is already high in the very young who seem to be carriers of high viral loads and participate in the shedding of virus [7] . These observations imply that CMV infection is not sufficiently controlled by the immune system at the very early life . Accordingly , neonatal mice are more susceptible to infections with MCMV than adult mice [8] , [9] , [10] , [11] , [12] , [13] , [14] . Similar findings have been reported for other pathogens including Respiratory Syncitial Virus , Listeria monocytogenes , Herpes Simplex Virus type 1 , Influenza Virus , and Pneumocystis [15] , [16] , [17] , [18] suggesting that neonatal mice in general are more vulnerable to infections . The mechanisms behind this phenomenon as well as the differences in antiviral immunity between the very young and adults remain largely undefined [19] , [20] . To understand and predict the outcome of a virus infection , it is of great importance to know where the infection is localized and what types of antiviral immune responses are initiated locally . Human cytomegalovirus ( HCMV ) DNA has been detected in several body fluids like blood , breast milk , saliva , urine , and bronchoalveolar fluid [3] , [5] , [7] . Thus , mucosal surfaces are most likely a primary target of postnatal CMV infection and indeed several routes of virus transmission have been suggested in neonates and children . Oral infection by contaminated breast milk and droplet infection of the lung by infectious saliva have been proposed in several studies [21] , [22] , [23] , [24] , [25] . MCMV has been widely used to investigate CMV infection in vivo in the mouse model [24] , [26] . Infections have been extensively studied in adult mice after systemic administration of the pathogen while the natural infection routes of MCMV , including transmission to newborns , remains a matter of debate [27] . Therefore , it is still unknown which mucosal tissues are targets for viral entry and which cell types become infected to such a challenge . Studying MCMV infection in adult mice has unmasked many aspects of the complex interplay between this pathogen and the immune system . Whereas CD8+ T cells are supposed to be major effectors of the host to control of MCMV infection there is also strong evidence that NK cells as well as CD4+ T cells contribute to keep the virus from undisturbed replication [24] , [28] . However , the composition of the neonatal immune system seems to differ in many ways if compared to that of adults [29] , [30] . This phenomenon is supplemented by the fact that there are enormous changes of immunity during the very first steps of life , especially in mucosal tissue [31] . However , the features of antiviral immunity in neonates infected with MCMV have not been investigated in detail . Accordingly , it is currently not known how the neonatal immune system responses to MCMV infection and why neonates subsequently suffer from increased morbidity and mortality . Here , we investigated the primary virus tropism in mucosal surfaces of neonatal mice with the use of recombinant viruses expressing suitable reporter proteins [32] . Comparative analysis of lung infection in neonatal and adult mice reflected characteristics of HCMV infection in terms of virus tropism and histopathology . Accordingly , primary mucosal MCMV infection in neonatal mice led to a pronounced systemic viral spread and simultaneously caused disease , whereas adults rapidly coped with the infection . The clonal expansion of MCMV-specific CD8+ T cells in both adults and neonates was paralleled by virus control although the time course differed between the two groups . Interestingly , MCMV infection attracted myeloid cells to form morphological unique nodular inflammatory foci ( NIF ) in the neonatal lung . Within these structures MCMV-infected cells were destroyed and subsequently engulfed by local APCs . Notably , using in situ 2-photon microscopy , we visualized priming of naïve CD8+ T cells in NIFs of the neonatal lung suggesting that the neonatal organism allows the local differentiation of myeloid cells into APCs that directly cross-present antigen within NIFs . Thus , this study provides fundamental new insights in early antiviral immune responses during mucosal infections of neonatal mice .
To investigate which mucosal surfaces of neonatal mice can be infected we applied MCMV via different routes . Since virus transmission by contaminated breast milk has been reported in humans [22] , [25] , we firstly analyzed whether the mucosa of the gastrointestinal tract is susceptible to infection . Following oral application , fluorescent latex microspheres ( 0 . 5 µm in diameter , applied in PBS ) were detected in the distal colon within 24 h , confirming sufficient ingestion of the inoculum ( Figure 1A ) . When we fed neonates with high doses ( 106 PFU ) of the recombinant MCMV-3D that encodes the fluorescent protein mCherry as well as Gaussia luciferase [32] we could neither detect mCherry+ infected cells in the oral cavity ( Figure 1B ) , esophagus , stomach , small or large bowel ( Figure 1C+D ) nor luciferase activity in organ homogenates of the gastrointestinal tract ( Figure 1E ) . We therefore concluded that carrier-free MCMV does not infect the neonatal intestine via the oral route . Pneumonitis is a frequently observed manifestation of HCMV infection [5] , [33] , [34] . It has been previously reported that MCMV does infect the lung of adult mice after intranasal and subcutaneous application as well as the adult and neonatal lung after intraperitoneal ( i . p . ) injection [21] , [35] , [36] , [37] . To investigate whether the lung mucosa could be a direct target for virus infection in neonatal mice we established a procedure to infect the lung by inoculation of virus into the laryngopharynx ( designated as “laryngopharyngeal ( l . p . ) infection” throughout the article ) . High numbers of mCherry+ infected cells and dose-dependent expression of luciferase in organ homogenates were detected in the lung within 1 day after virus application ( Figure 1F–H ) . Histological analysis showed solitary infected cells mainly in the epithelium of distal respiratory ducts and terminal bronchioles but not in the epithelium of the trachea ( Figure 1I ) . Approximately 50% of the infected cells were pro-surfactant protein C-expressing type 2 alveolar epithelial cells [38] ( Figure 1J , Figure S1A ) . No CD45+ cells of hematopoietic origin were found to be infected ( Figure 1K , Figure S1B ) . Together , the data illustrate that the neonatal lung epithelium is highly susceptible to MCMV infection . Human CMV infection has not only been described in the lung but also in numerous organs and tissues such as the liver , brain , spleen , vascular endothelium and the kidney [3] , [5] , [39] . Likewise , in models for systemic MCMV infection multiple organs have been reported to become infected [10] , [14] , [35] , [40] , [41] , [42] . We analyzed various organs after lung infection and found that by day 8 post infection ( p . i . ) mCherry+ cells could be detected in all investigated neonatal organs ( Figure 2A ) . As described above for the lung , most infected cells were CD45− demonstrating their non-hematopoietic origin . Various parenchymal cells were found to be infected and the spatial proximity to CD31+ vascular endothelial cells suggested hematogenous virus dissemination ( Figure 2A ) . Surprisingly , we also found infected cells next to CD31+ vascular endothelial cells in the intestine ( Figure S2A ) indicating that hematogenous viral spread can lead to infection of the intestine , while oral application of the virus failed to infect the gut ( see Figure 1 ) . To gain insight into the dynamics of viral dissemination we determined luciferase activity to screen for virus spread to various organs following primary infection of the lung . All screened organs of neonatal mice including the salivary glands , liver , brain , spleen , and kidney possessed luciferase activity after primary lung infection ( Figure 2B ) . However , whereas infection of the lung was already detectable one day p . i . ( Figure 1H ) viral activity in these organs was delayed by approximately six days ( Figure 2B ) , an observation that also supports the idea of a barrier function of the lung preventing massive viral systemic exposure [43] . A comparative analysis with weight-adapted virus doses revealed that in adult mice luciferase was detected only in organ homogenates of the salivary glands and the spleen after intranasal infection with MCMV-3D ( Figure 2C ) . Apart from the salivary glands , which have been described to be a place of ongoing virus persistence [44] , [45] , none of the tissues examined , neither neonatal nor adult , showed luciferase activity three weeks after infection ( Figure 2B+C ) . Therefore , neonates as well as adults are able to cope with respiratory cytomegalovirus infection . However , when monitoring the body weight as a parameter of the health status , differences between adult and neonatal mice could be observed . While in the latter both a low dose ( 5×104 PFU ) as well as a high dose ( 106 PFU ) MCMV-3D infection temporally coincided with the diminished increase in body weight when compared to mock infected animals ( Figure 2D ) , there was no effect in this respect in MCMV-3D-infected ( 106 PFU ) adult mice ( Figure 2E ) . The pronounced vulnerability of neonatal mice to MCMV infection can be explained by productive virus infection in all organs that is accompanied by cachexia . We continued with low dose ( 5×104 PFU ) lung infections in further experiments in neonates to avoid excessive virus exposure to the neonatal organism . We speculated that an impaired antiviral immune response in the neonatal lung might be the cause of the massive systemic viral spread and disease . Therefore , we investigated early cellular immune responses in neonates via histological analysis of the lung 5 days after low dose infection ( MCMV-3D; 5×104 PFU ) , prior to the onset of viral dissemination . At this time-point numerous areas containing multiple infected cells with dense infiltration of CD45+ cells could be detected ( Figure 3A , framed areas ) . These infiltrates were exclusively found around foci of infected cells , apparently sheathing them and causing a nodular appearance of the lung . Similar histopathology has been described in pulmonary HCMV-infection of immuno-compromised adults and termed “nodular inflammatory foci” [46] . Therefore , we equally termed these areas of the neonatal lung “nodular inflammatory foci” ( NIF ) , defined as multiple juxtapositioned MCMV-infected cells and associated immune cell infiltrate . Besides the presence of NIFs we found no evidence for further profound pathological alterations of the lung ( Figure 3B ) . However , we frequently observed some solitary infected cells within the lung that were not contacted by CD45+ cells suggesting that they may escape immune surveillance ( Figure 3A , arrow heads ) . Alternatively , these cells could also reflect earliest stages of NIF development . Most of the infected cells within the NIFs were neither pro-surfactant protein C-expressing type 2 alveolar epithelial cells nor CD45+ hematopoietic cells . Instead , cell morphology and position matched that of stromal cells such as fibroblasts ( Figure 3C ) . Hence , in addition to type 2 alveolar epithelial cells also other cell types become infected during the course of infection . Further analysis of the infiltrated cells identified them as a variety of CD11b+ , CD11c+ and/or F4/80+ myeloid cells and the abundance of these cells account for the “nodular” morphology of the inflammatory foci ( Figure 3D+E ) . In contrast , only few T and B cells were present ( Figure 3E , Figure S3A–C ) . Natural killer cells were also present as illustrated by analysis of Ncr1gfp/+ transgenic mice ( Figure 3F ) and NK1 . 1 cell surface expression ( Figure S3D ) . Further high resolution analysis revealed heterogeneous patterns of mCherry signals within the NIFs . The mCherry fluorescence intensity varied between infected cells suggesting that lung parenchymal cells were either differently permissive to infection or had been infected at different time-points and therefore were in different phases of viral replication ( Figure 3G; arrows ) . In addition to multiple infected cells we found some cellular mCherry+ remnants that most likely were derived from infected apoptotic cells as described previously for MCMV-infected cells in the salivary glands [47] ( Figure 3G; arrow heads ) . Interestingly , we found some of these remnants to be situated within CD45+ cells suggesting engulfment of mCherry-containing cell debris by myeloid cells ( Figure 3G ) . We found remnants in CD11c+ cells and F4/80+ macrophages ( Figure 3H+I ) . CD169+ macrophages were also present in NIFs at high frequencies and similarly contained mCherry+ remnants ( Figure 3J ) . In contrast , few CD103+CD11c+ DCs were found to be present in NIFs but occasionally formed close contacts with infected cells ( Figure S3E ) . Thus , NIFs are clearly different from bronchus-associated lymphoid tissue ( BALT ) , tertiary lymphoid structures of the lung that are localized next to bronchi and characterized by the presence of B cell follicles with separated T cell areas [48] , [49] . Instead , NIFs appear to be areas of viral replication although myeloid cells present in NIFs can engulf remnants of infected cells and probably function as APCs . To gain comprehensive insight into the genesis of the NIFs and their role in viral clearance , we performed a comparative analysis of neonatal and adult lung sections at different time-points after infection with MCMV-3D . Over an observation period of three weeks we constantly found solitary infected cells in the neonatal lung that were not in contact with any hematopoietic cell ( Figure 4A+B ) . However , multiple infected cells could be detected side by side 3 days p . i . suggesting cell-to-cell spread of MCMV in vivo ( Figure 4A+B ) . NIFs could be detected in the neonatal lung from day 3 until at least day 8 p . i . but disappeared , to a large extent , by day 12 p . i . ( Figure 4A+B ) . Accordingly , luciferase activity as well as the number of infected cells per lung slice did not decrease within the first 8 days p . i . ( Figure 4C+D ) . Instead , NIFs appeared to be niches of ongoing virus replication possibly by recruiting susceptible fibroblasts [50] . The disappearance of the NIFs was associated with declining numbers of infected cells ( Figure 4A , B+D ) . Interestingly , solitary infected cells that were not targeted by immune cell infiltrates were still present three weeks after infection ( Figure 4A+B ) . In addition , appearance of NIFs coincided with the presence of mCherry+ cell remnants indicating immune cell-mediated destruction of infected cells within these structures ( Figure 4B+E ) . To compare these findings to those in the adult lung , we intranasally infected 6–8 week old mice with a weight-adapted dose of MCMV-3D . As observed for neonates , the virus also infected type 2 alveolar epithelial cells ( unpublished data ) and a comparable infection pattern to that observed in neonates was evident in adult mice within the first 3 days p . i . ( Figure 4A ) . Likewise , a localized inflammation with immune cell infiltration was found around foci of infected cells . Although the morphology , composition and localization of these infiltrates may differ from the ones found in neonates we also named these structures “NIFs” of the adult lung . However , in contrast to the situation in neonates , the number of remnants of infected cells peaked already at day 5 p . i . in adult NIFs and only few morphologically intact mCherry+ infected cells could still be identified ( Figure 4A , D–F ) . By day 8 p . i . luciferase activity had decreased from a peak at day 3 p . i . ( Figure 4C ) . Accordingly , only residuals of cell infiltrates remained but few solitary infected cells were still detectable ( Figure 4A+D ) . Luciferase activity as well as a comparative quantification of infected cells and mCherry+ remnants per lung slice showed a clear delay of virus control in the neonatal lung as compared to adults ( Figure 4 C–F ) . These data demonstrate that neonatal mice can cope with most of the infected cells in the lung . However , although neonates induce a cellular immune response and form NIFs , they suffer from a persistent lung infection for up to three weeks . During the first 8 days infiltrating immune cells in the neonatal lung tolerate ongoing infection and fail to prevent spread of the infection to neighboring cells whereas adult mice start to contain viral replication in the lung already within the first 4 days . CD8+ T cells have been implicated as major contributors to MCMV infection control in adult mice [28] . To test the hypothesis that the ongoing MCMV infection in the lung of neonates from day 1 until day 8 p . i . could be due to a limitation in the CD8+ T cell response we analyzed the presence of CD8+ T-cells which recognize the immunodominant MCMV epitope M45 [51] . M45-specific CD8+ T cells showed a massive expansion at day 8 p . i . in adult mice and already turned to the contraction phase at day 12 p . i . ( Figure 5A–C ) . In contrast , M45-specific CD8+ T cells in neonates were hardly detectable before day 12 p . i . in lung or lung draining lymph nodes ( Figure 5A–C ) . As the expansion of M45-specific CD8+ T cells showed a temporal coincidence with the initiation of virus control in neonates at ∼day 12 p . i . ( Figure 4A–E ) we depleted CD8+ T cells to investigate their importance in the clearance of infection ( Figure 4D and Figure S4A ) . After depletion , we found higher luciferase activity in the lung , liver and all other organs analyzed ( Figure 5E , Figure S4B ) . In addition , unlike the control group , CD8+ T cell-depleted neonatal mice showed NIFs which contained increased numbers of viable infected cells ( Figure 5F ) confirming the contribution of CD8+ T cells to controlling MCMV infection in neonatal mice . Next , we performed adoptive transfers of CD8+ T cells into MCMV-infected neonatal mice to determine whether this treatment could abrogate MCMV infection . For this , we took advantage of lymphocytes from transgenic mice ( OTI ) , which express a recombinant , Kb-restricted T cell receptor that exhibits high affinity to MHC class I bound SIINFEKL peptide [52] . To that end , we infected neonatal mice either with MCMV-3D or MCMV-3DΔvRAP virus mutants [32] . Both viruses encode the SIINFEKL peptide but MCMV-3DΔvRAP lacks the “viral regulator of antigen presentation” genes encoding for the MHC class I evasion proteins gp48/m06 and gp40/m152 [53] , [54] . In contrast to MCMV-3D , this mutant is therefore expected to lack the ability to interfere with MHC class I peptide surface expression as was recently shown for a related virus mutant [55] . We speculated that MHC class I bound SIINFEKL peptide presentation on MCMV-3DΔvRAP-infected cells would allow direct recognition of infected cells by OTI T cells and decreased luciferase activity in organs of these animals would be an in vivo indicator for the activity of cytotoxic T lymphocytes ( CTL ) . Two days p . i . and at the day of infection we adoptively i . p . transferred various numbers of naïve CD8+ T cells from OTIxGFP mice and analyzed the animals six and ten days after transfer , respectively ( Figure 5G and Figure S5A ) . In animals infected with MCMV-3DΔvRAP the reduction in luciferase activity in the lung and liver depended on the number of OTI T cells transferred ( Figure 5H ) . Moreover , MCMV-3DΔvRAP-infected animals which received high numbers of OTI T cells showed no NIFs and only very few solitary infected cells could be found in the neonatal lung ( Figure 5I and Figure S6C ) . The antiviral effect was also observed in the spleen , kidney , and brain of these mice ( Figure S6B ) . In contrast , we did not observe a robust impact on luciferase activity within the neonatal lung and most organs tested 6 or 10 days after adoptively transferring OTI T cells into MCMV-3D-infected mice ( Figure 5H , Figure S5A+B ) . Even in the presence of adoptively transferred OTI T cells these mice still possessed NIFs that harbored multiple infected cells ( Figure 5I and Figure S6C ) . However , in MCMV-3D-infected neonatal animals , the liver significantly benefited from the transferred cytotoxic T lymphocytes ( Figure 5H and Figure S5B ) . Previous reports have shown that already the deletion of one gene ( m152 ) for the MHCI immune evasion leads to virus attenuation even in BALB/c neonatal mice [56] . In line with this report , in our model with infection of C57BL/6 neonatal mice we saw a trend to lower luciferase activity in the lungs of MCMV-3DΔvRAP-infected mice if compared to MCMV-3D-infected neonates , but the difference was not significant ( Figure 5H , “MCMV-3D no OTI” vs . “MCMV-3DΔvRAP no OTI” , p = 0 . 0675 , unpaired t-test ) . In summary , the CD8+ T cell response to MCMV infection in neonates is strikingly different from the response in adults and likely contributes to delayed virus control in neonates . Priming of naïve CD8+ T cells is supposed to take place in secondary lymphoid tissue [57] , and we have recently shown that T cells can also be primed in tertiary lymphoid tissues such as BALT [49] . Since we found many APCs in the NIFs of the neonatal lung , we wondered whether they could assist in priming naïve T cells already at the site of infection . To test this hypothesis , we adoptively i . p . transferred purified naïve CD8+ T cells from OTIxGFP mice into neonates infected with the reporter viruses MCMV-3D or MCMV-2D ( Figure 6A ) . The latter lacks the sequence encoding the SIINFEKL peptide [32] . Within one day of transfer we found in lung draining lymph nodes of MCMV-3D , but not MCMV-2D-infected neonates , a considerable proportion of OTI T cells to express CD69 , indicating T cell activation and arguing that neonatal lymph nodes are able to prime CD8+ T-cells ( Figure S7A+B ) . To further investigate the priming capability of cells in the NIFs we performed in situ 2-photon microscopy of lung explants from MCMV-infected neonatal mice . Surprisingly , naïve OTI T cells accumulated already within 1 day after i . p . transfer in NIFs of MCMV-3D , but not of MCMV-2D-infected mice ( Figure 6B+C , Movie S1 ) . These T cells in MCMV-3D-infected neonates showed a slowed migration behavior in NIFs , similar to that of naïve T cells that are primed in the lymph node as reported earlier [58] . In contrast , peribronchial T cells which were not next to infected cells were not confined 1 day after transfer ( Movie S2 - Scene 1 ) . Within 2 days of transfer , a high proportion of OTI T cells in NIFs showed a lymphoblastic appearance and enlarged nuclei in MCMV-3D but not MCMV-2D-infected mice ( Figure 6B–D , Movie S3 ) . Additionally , after we subcutaneously treated MCMV-3D-infected neonates with a pulse of the nucleoside analog 5-ethynyl-2′-deoxyuridine ( EdU ) 2 days after T cell transfer and sacrificed the animals within 4 hours , immunohistology revealed a high frequency of proliferating EdU+ OTI T cells ( Figure 6E ) . Furthermore , these T cells within NIFs became highly motile within 4 days of transfer ( Movie S2 - Scene 2 ) . Together , these data indicate that neonatal APCs in NIFs can induce OTI T cells to pass the classical priming program directly at the site of infection , including confined migration behavior after antigen-recognition , subsequent lymphoblastic appearance , cell proliferation and increased cell migration after the differentiation into CTLs . Activated OTI T cells were detected as early as 48 hours after transfer within NIFs indicating that T cells were also primed in these structures . To formally exclude the possibility that activated T cells present in NIFs were initially primed in the lung-draining lymph node we blocked the egress of T cells from lymph nodes by treating neonates from the time of adoptive transfer of OTI T cells with the functional sphingosine 1-phosphate receptor antagonist FTY720 ( Figure 7A ) [59] . Four days after cell transfer , the frequency of OTI T cells in MCMV-2D-infected mice was extremely low in all compartments analyzed ( Figure 7B ) . These cells did not proliferate and did not express the effector/memory marker CD44 ( Figure S8A ) . In contrast , most of OTI T cells in lung draining lymph nodes of MCMV-3D-infected neonates had started to proliferate and expressed CD44 , indicating that they experienced antigen ( Figure 7C ) . Furthermore , we found significantly more OTI T cells in lung-draining lymph nodes of FTY720-treated MCMV-3D-infected neonates than in control animals and hardly detected OTI T cells in the blood of neonates which received FTY720 , confirming the blockade of T cell egress from lymph nodes by this drug ( Figure 7B ) . Despite the inhibition of T cell egress from lymph nodes , the frequency of OTI T cells in the lungs of FTY720-treated neonates was comparable to that in the control group ( Figure 7B ) . Furthermore , OTI T cells in the lung of FTY720-treated animals showed proliferation and CD44 expression that were similar to those in the control group ( Figure 7C ) . Conclusively , these data confirm the hypothesis that activated T cells present in NIFs have also been activated in these structures . As these data suggested that neonatal APCs can potently prime T cells in lymph nodes and in NIFs we wondered if a small T cell receptor repertoire in neonates and therefore a low precursor frequency for MCMV-specific T cells accounted for the delay in clonal expansion of M45-specific CD8+ T cells ( Figure 5A ) . Consequently , we adoptively transferred 107 polyclonal CD45 . 1+CD8+ T cells from adults into neonates at the time of MCMV-3D infection and treated these animals with FTY720 to prevent egress of lymph node-primed T cells ( Figure 7D ) . We speculated that the adult T cell repertoire contains T cells with M45-reactive TCRs and that this adoptive transfer would substitute for the missing MCMV-reactive CD8+ T cells in neonates . As clonal expansion of M45-specific T cells peaked at ∼8 days p . i . in adults ( Figure 5A–C ) we also analyzed neonates at day 8 p . i . ( Figure 7D ) . Of interest , we found a considerable frequency of M45-specific T cells in the neonatal lung within the CD45 . 1+CD8+ T cell fraction ( Figure 7E ) . In addition , the transferred CD45 . 1+ cells ( with ∼90% of CD45 . 1+ cells being CD8+ T cells; unpublished data ) were situated within NIFs suggesting the accumulation of MCMV-specific CTLs in NIFs ( Figure 7F ) . In summary , these data support the hypothesis that APCs in neonatal NIFs are capable of presenting MCMV peptides ( including M45 ) to naïve CD8+ T cells to directly prime these cells at the site of infection . Furthermore , as the adoptive transfer of polyclonal adult CD8+ T cells led to expansion of M45-specific clones , it is likely that the low precursor frequency of MCMV peptide-specific CD8+ T cells accounts for the delayed clonal expansion of M45-specific CTLs in neonates . Finally , we aimed to further characterize T cell priming in non-lymphoid tissue by 2-photon microscopy of NIFs in the neonatal lung . In particular , we wondered if we could observe interactions between APCs and CD8+ T cells . Since CD11c is mainly expressed by dendritic cells and alveolar macrophages [60] , we infected neonatal CD11c-YFP transgenic mice with MCMV-3D . Four days later we adoptively transferred purified naïve CD8+ T cells from OTIxCFP mice ( Figure 8A ) . Numerous OTI T cells could be found in a dense network of CD11c+ APCs in the NIFs within 1 day of T cell transfer ( Figure 8B , Movie S4 ) . Interestingly , OTI T cells were in direct contact with APCs , but only occasionally with cells infected with the reporter virus MCMV-3D that carries the MHC class I immune evasion genes and is therefore expected to interfere with MHC class I peptide presentation ( Figure 8B+C ) . Most of the contacts observed between OTI T cells and APCs were stable and some lasted for more than 30 minutes ( Figure 8D ) . Of interest , APCs formed cell protrusions which connected OTI T cells with infected cells ( Figure 8B , Movie S4 ) . Most of the contacts between APCs and OTI T cells occurred when the APC itself was in contact with an infected cell ( Figure 8E ) . Histological analysis of NIFs revealed intensive synapse formation of OTI T cells with CD169+ macrophages which contained remnants of infected cells ( Figure 8F ) . These data support the idea that MCMV-specific cytotoxic CD8+ T cells can be primed by myeloid cells , potentially by CD169+ macrophages , of virus-induced NIFs in the lung and that these myeloid cells contribute to the local antiviral immune response .
In the present study we established an animal model for mucosal MCMV infection in neonatal and adult mice and describe the spatio-temporal distribution of virus infection at the single-cell level . Compared to classical MCMV infection models using systemic application ( i . p . or i . v . ) of the virus that leads to primary infection of multiple organs , the approach used in this study allows investigation of immune responses at mucosal surfaces . Our data suggest that infection of the respiratory tract serves as a previously underestimated entry organ for CMV in neonates and that other organs become infected after primary virus replication and hematogenous spread . This virus is known to persist in salivary glands and infectious virus can be found in saliva of infected humans . Thus , CMV may be transmitted via virus-containing saliva to the respiratory tract . This transmission route may be of clinical importance especially for postnatal infection of highly susceptible preterm low birth-weight infants . In neonatal and adult lungs type 2 alveolar epithelial cells were frequently found to be infected with MCMV , a cell type that also has been suggested as a target for HCMV [61] , [62] . Our observations are also in accordance with the model of cell-to-cell spread of infectious virions in the lung since we could visualize that different neighboring cell types became infected at later time-points after primary infection . The proximity of alveolar epithelial cells , fibroblasts and vascular endothelial cells within the lung [63] suggests the sequential infection of these cells as an imaginable route for virus particles to enter the blood stream from the pulmonary alveoli and spread systemically within the host . Indeed , infection of the gastrointestinal tract was evident in neonatal mice after hematogenous virus spread but not after virus administration via the oral route . After oral application , low pH in the stomach and bile in the duodenum supposedly prevent infection of the small and large intestine with the enveloped MCMV . Though , the neonatal stomach is not very acidic in the first days of life and therefore may allow the virus to enter the duodenum as an infectious particle . However , it was unexpected that neither the oral cavity nor the esophagus seem to be mucosal entry sites for MCMV . A study by Wu and colleagues suggested that neonatal mice can be infected after oral application of carrier-free MCMV as well as virus-containing milk cells [64] . However , these authors did not analyze infection of the gastrointestinal tract itself but instead found viral transcripts in the lung and other organs 4 days after oral delivery of virus . Therefore , it is currently not known which cells are first targeted by MCMV after oral application . Nevertheless , the data presented in our study suggest that after systemic spread from a primary site of infection such as the respiratory tract , MCMV can disseminate to any vascularized corner of the body and virus infection of the colon may actually occur from the “blood-side” and hike through the epithelium rather than start at the apical side of the epithelium . To prevent , or at least reduce , systemic dissemination of pathogens efficient antiviral defense mechanisms has to be induced very early after infection of mucosal tissues . Following MCMV lung infection neonatal mice failed to prevent the systemic spread of virus originating from infected cells of the lung . An impaired first line antiviral defense is most likely the cause for the high susceptibility of neonates to virus infection and explains prolonged viral replication in the lung and pronounced virus dissemination with subsequent infection of various organs . Still , virus spread to other organs was usually not observed during the first 5 days of infection in neonates and in adults , and the majority of infected cells was cleared after 3 weeks suggesting that local immune responses of the neonatal lung to some degree help to reduce systemic spread of the virus . Several of our observations support the hypothesis that the formation of NIFs contribute to control MCMV infection of the neonatal lung: i ) infected cells of the lung were efficiently removed at locations where NIFs were induced , ii ) remnants of infected cells were always found to be associated with NIFs , iii ) macrophages within the NIFs contained fragments of lysed , virus-infected cells indicating that NIF macrophages locally remove infectious virions , and iv ) NIFs provided an environment that allowed priming of antigen-specific cytotoxic T lymphocytes . Immunohistology identified NIFs to primarily consist of MCMV-infected cells and myeloid cells including macrophages ( F4/80+; CD169+ ) as well as DCs ( CD11b+CD11c+; CD103+ ) while only few lymphocytes and NK cells were present . These features clearly distinguish NIFs from induced BALT that develops after the clearance of infections and is characterized by large and separated T and B cell zones and that has been shown to act as a general priming site for T cells [49] . Although T cells are sparse , our data indicate that APCs can efficiently prime naïve CD8+ T cells directly in NIFs . T cell priming is a multistep process which has been extensively characterized in secondary lymphoid organs such as lymph nodes . There , following recognition of antigen presented by APCs T cells undergo an extensive proliferation and differentiation program that lasts for at least three days . During this period all lymphocytes are trapped within these organs by a process known as lymph node shut down , which also prevents the release of activated T cells . Therefore , it is unlikely that those OTI T cells that were observed to rapidly proliferate in NIFs 2 days after their adoptive transfer were initially primed in lung-draining lymph nodes . The idea that T cells are directly primed in NIFs is further supported by the finding that proliferating T cells were also present in NIFs of FTY720-treated mice , where T cell egress from lymph nodes is blocked . Furthermore , 2-photon microscopy studies revealed intensive interaction and synapse formation of APCs and OTI T cells within NIFs . In addition , approximately 75% of the APCs that contacted OTI T cells simultaneously interacted with infected cells indicating that APCs , which are not actively infected , actually cross-present viral antigens to naïve T cells that differentiate to mature CTLs . It is unknown which entry portal is used by naïve T cells to enter NIFs and what signaling molecules are involved . As data on essential molecules for homing of lymphocytes into the lung is sparse this needs to be addressed in future studies . Furthermore , it is currently unclear which subset of the CD11c+ APCs observed actually cross-presents antigen in the NIFs . We recently identified lung-derived CD103+ cells to cross-present antigen to CD8+ T cells in lung-draining lymph nodes [65] . Thus it seems possible that CD103+ DCs also cross present antigen directly in NIFs and indeed immunohistology identified few CD11c+ CD103+ DCs to be in direct contact with infected cells . Alternatively , some of the newly recruited monocytes and/or DC progenitors undergo a differentiation program within NIFs that allows the local generation of cross-presenting DCs . Interestingly , CD169+ macrophages were present at high frequencies within NIFs . CD169+ lymph node macrophages have recently gained considerable attention since they were identified to play important roles in controlling spread of lymph-derived virus , in presenting lymph-derived antigen to B cells and to cross-present lymph-derived apoptotic tumor cells to induce cytotoxic T cell responses [66] , [67] , [68] . In NIFs , CD169+ macrophages not only contained remnants but also contacted infected cells and simultaneously formed synapses with OTI T cells suggesting that these cells actually cross-present antigen and therefore contribute to the control of MCMV infection in the neonatal lung . Antibody depletion of CD8+ T cells clearly promoted virus replication in MCMV-3D-infected neonates emphasizing an important role of cytotoxic T cells in MCMV control . As we observed activated endogenous CD8+ T cells in MCMV-infected neonates at day 8 ( Figure 5A , CD44 expression and unpublished data ) it is likely that MCMV epitopes were recognized by CD8+ T cells at that time . Additionally , the cytotoxic CD8+ T cells activated in NIFs are seemingly fully functional since naïve OTI T cells adoptively transferred into neonates differentiated to cytotoxic effector T cells and efficiently reduced the viral load in all organs analyzed of mice infected with MCMV-3DΔvRAP . Activated OTI T cells had only a limited effect in mice infected with the MCMV-3D variant . These data indicate that the MCMV-encoded vRAP proteins , m06 and m152 , efficiently prevent killing of MCMV-infected cells by CD8+ T cells as shown by others before [56] , [69] , [70] , [71] , [72] . These observations suggest that , in principle , neonatal mice can prime CD8+ T cells and induce CTL-mediated antiviral immunity . Interestingly , the generation of CTLs in the present animal model is in line with a previous report showing expansion of CMV-specific CD8+ T cells in newborns upon HCMV infection [73] . This raises the question why particularly the CTL response should be responsible for the higher susceptibility of neonates to MCMV infection . First , the frequencies of lymphocytes are in general lower in neonatal than in adult mice [30] . Furthermore , our observation that M45-specific CD8+ T cells efficiently expand in neonates once adoptively transferred from adult donors suggest that a low precursor frequency - rather than a general defect in T cell priming in neonates - contributes to the late expansion of M45-specific CD8+ T cells . These findings are in line with reports showing that the neonatal TCR repertoire and hierarchy differ from that of adults [74] , [75] . Therefore , low numbers of MCMV-specific CTLs in combination with a reduced clonal repertoire and diminished variety of recognized viral proteins may account for the vulnerability of these young organisms . We also observed NK cells and CD4+ T cells in NIFs but it is currently unclear to what degree these cells contribute to the anti-MCMV response in these structures . NK cells have been proposed to lack multiple activating receptors during the very first days of life [76] . Additionally , neonatal myeloid cells have been reported to produce only low levels of IL-12 [77] and subsequent low IFN-γ responses by T cells and NK cells may diminish antiviral immunity in neonates . As CD4+ T cells also contribute to control of MCMV infection it is likely that low precursor frequencies of both , MCMV-specific CD4+ and CD8+ T cells , account for the vulnerability of neonates to MCMV infection . In summary , the high susceptibility of neonatal mice to viral infection may be the result of an impaired innate and a delayed adaptive antiviral immune response that allows prolonged local virus replication and extreme systemic viral spread with multi-organ disease and cachexia . In both , the adult and neonatal lung solitary infected cells were still present when all the infected cells had been removed from the NIFs . These findings suggest the existence of micro-anatomical niches which allow immune evasion of infected cells . Likely , innate immune responses are needed to allow migration of immune cells to places of viral infection which then leads to removal of MCMV-infected cells or inhibition of viral replication . Possibly , in some cells MCMV infection does not trigger these early responses and therefore the first steps of inflammation are not initiated . Alternatively , solitary cells might result from secondary infections with viruses released from other organs such as the salivary glands . The fate of these infected cells , apparently ignored by the immune system , clearly deserves further attention since it cannot be excluded that latent infection is finally established in such cells . The present study is to our knowledge the first to describe and profoundly characterize NIFs as well as solitary infected cells in the neonatal lung upon MCMV infection . Interestingly , CMV-associated interstitial pneumonia with formation of nodules is one among various reported lung manifestations of CMV infection in immuno-compromised adults [33] , [46] . As it is currently unclear what factors determine the type of lung manifestations in human CMV patients , the mouse model presented in this study might help to shed light on the pathogenesis of CMV lung disease as well as the definition of crucial antiviral immune responses to control CMV infection in the lung . In summary , this study provides profound insight into host-pathogen interaction upon viral challenge of the lung of neonatal mice . The localized accumulation of primarily myeloid immune cells at the site of infection represents an essential feature for the formation of NIFs in the neonatal lung . These structures allow the local induction of adaptive immune responses and moreover represent the anatomical correlate where the control of MCMV infection takes place .
Mice were all on a C57BL/6 background , bred at the central animal facility of Hannover Medical School under specific pathogen free conditions and/or purchased from Charles River Laboratories . ß-actin-eGFP mice [78] and ß-actin-eCFP [79] mice were crossed to ovalbumin-transgenic TCR ( OTI ) mice [52] and the F1 cross was labeled as OTIxGFP and OTIxCFP , respectively; CD11c-YFP [60]; Ncr1+/gfp [80] . All animal experiments were performed according to the recommendations and guidelines of the Federation of European Laboratory Animal Science Associations ( FELASA ) and Society of Laboratory Animals ( GV-SOLAS ) and approved by the institutional review board and the Niedersächsische Landesamt für Verbraucherschutz und Lebensmittelsicherheit ( AZ33 . 9-42502-04-10/0225 and AZ33 . 12-42502-04-12/0921 ) . MCMV mutants have been described previously [32] and were produced and titrated on mouse embryonic fibroblasts . MCMV-2D encodes Gaussia luciferase and mCherry , MCMV-3D carries additionally a sequence within the m164 ORF encoding the SIINFEKL peptide . The MCMV-3DΔvRAP mutant is identical to MCMV-3D except that it lacks the m06 and m152 ORFs . All reporter viruses lack the m157 ORF that encodes a ligand for the activating receptor Ly49H present on NK cells in C57BL/6 mice [81] . C57BL/6 wildtype or CD11c-YFP mice were mated and dams were kept with their litter . Neonatal mice were infected on their first day of life ( <24 h old ) ; “oral” inoculations were performed by repeated moistening of the mouth with fluid up to a volume of 10 µl ( for control applications we used 3×109 Fluoresbrite YG Microspheres , Polysciences Europe GmbH ) , for l . p . inoculations a volume of 10 µl was administered by probing of the laryngopharynx with a pipette and extension of the neck . Adult C57BL/6 wildtype mice ( 6–8 weeks old ) were anesthesized ( 100 mg/kg BW ketamine and 5 mg/kg BW xylazine ) and 20 µl of virus solution was applied to each nostril for “intranasal” infection . CD8+ T cells were isolated with MACS CD8+ T cell isolation kit ( Miltenyi Biotec ) from lymph nodes and spleen of OTIxGFP or OTIxCFP mice and had a purity of 85–95% . MCMV-2D and MCMV-3D-infected neonatal mice received equal numbers ( 5×106 cells ) of naïve CD8+ T cells via i . p . application ( Figure 6 ) . CD8 T cells were depleted by intraperitoneal application of RmCD8 . 2 mAb ( 25 µg/g body weight; Figure S4A ) . FTY720 was given subcutaneously ( 5 µg/g body weight ) on a daily basis . The first administration was given at the time of cell transfer ( Figure 7 ) . Right heart ventricle was perfused with PBS until blood cells were removed from the lung . Fragmented tissue was digested with Collagenase D ( Roche , 0 . 5 mg/ml ) and DNAse I ( Roche , 0 . 025 mg/ml ) for 45 min at 37°C , meshed through 40 µm Falcon® Cell Strainer and leukocytes isolated with Lympholyte®-M . Leica MZ16 epifluorescence microscope was used for whole organ images . For histology organs were fixed in 2% PFA and 30% sucrose for 30 min and embedded in OCT compound ( Tissue-Tek , Sakura ) . 7 µm-thick organ slices were stained after appropriate blocking with depicted antibodies . Images were taken with an AxioCam MRm camera ( Carl Zeiss ) attached to Axiovert 200M fluorescence microscope ( Carl Zeiss ) with PlanApochromat objectives 10×/0 , 45 , 20×/0 , 75 and 40×/0 , 95 ( magnification/numerical aperture ) and processed with AxioVision 4 . 8 software . Images of HE stained sections were taken with Olympus BX61 microscope and ColorView IIIu camera with UPlanSApo objectives ( 4×/0 , 16 and 40×/0 , 90 ) and processed with cell∧P 5 . 0 ( Olympus Europe ) . All images were processed with Microsoft Office Picture Manager . Cell strainers ( BD Falcon ) were used to prepare suspensions for FACS analysis or cell purification from lymph node or spleen cells . Cells were processed with LSRII Cytometer and data was analyzed with BD FACSDiva Software ( 6 . 1 . 3 ) or WinList 6 . 0 software . The following antibodies ( clones ) were used after adequate blocking of Fc receptors: B220-Cy5 ( RA3-3A1 ) , CD103-PE ( M290 ) , CD169-AlexaFluor647 ( MOMA-1 ) , CD11b-AlexaFluor488 ( M1/70 ) , CD11c-APC ( N418 ) , CD3-AlexaFluor488 ( 17A2 ) , CD3-PE ( 17A2 ) , CD31-biotinylated ( MEC13 ) , CD4-biotinylated ( GK1 . 5 ) , CD4-PerCP ( RM4-5 ) , CD44-eFluor450 ( IM7 ) , CD45-APC ( 30-F11 ) , CD69-PerCP/Cy5 . 5 ( H1 . 2F3 ) , CD8a-APC/Cy7 ( 53-6 . 7 ) , CD8b-Cy5 ( Rm CD8-2 ) , CD8b-AlexaFluor488 ( Rm-CD8-2 ) , F4/80-APC ( BM8 ) , NK1 . 1-PE ( PK136 ) , pro surfactant protein C ( AB3786 ) combined with anti-rabbit-Cy5 ( Jackson ImmunoResearch ) , TCR-Vα2-PE ( B20 . 1 ) . Streptavidin-Cy5 ( eBioscience ) , Streptavidin-APC/Cy7 ( BD-Pharmingen ) , Cell Proliferation Dye eFluor® 670 ( eBioscience ) , M45-tetramer-PE provided by Ramon Arens . Single organ preparations were performed after perfusion of supplying blood vessels with PBS . Organs were kept in PBS , homogenized with TissueLyser II ( Qiagen ) and supernatants were measured for luciferase expression after addition of “native Coelenterazine” ( Synchem ) with Lumat LB 9507 ( Berthold Technologies ) . For lung , salivary glands , gut and liver 1∶10 dilutions were performed for measurements . The following organs were analyzed: lung ( Figure 1 complete lung , Figure 5 lobes of right lung including trachea ) , gut ( from proximal esophagus to distal colon ) , salivary glands ( all sublingual and submaxillary ) , brain ( down to the bulb ) , spleen , liver ( complete liver of neonates , only left lobe from adults ) , kidney ( right only ) . Luciferase measurements of organs from non-infected animals were used as controls and data was normalized to means of control measurements to determine the detection limits . Neonatal lungs were explanted and 400 µm-thick lung slices were prepared with use of a Tissue Chopper ( McIllwain ) . Lung slices were fixed on a imaging chamber using tissue adhesive ( Surgibond ) and kept in oxygenated ( 95% O2/5% CO2 ) RPMI 37°C medium ( Invitrogen ) containing 5 g/L glucose . Imaging was performed with Olympus BX51 upright microscope equipped with a 20×/0 . 95 water immersion objective . A MaiTai Ti∶Sa pulsed IR laser ( Spectra-Physics ) was set to 920 nm for excitation of eGFP ( as well as Ncr1gfp/+ and DAPI for Figure 3 F ) or 860 nm for excitation of eCFP and YFP . A second laser excited mCherry with 1100 nm generated from an optical parametric oscillator ( OPO; APE , Berlin ) . Z-stacks of up to 30 images from 300×300×60–160 µm ( Movie S1 , S2 , S3 ) or 150×150×60–160 µm ( Movie S4 ) viewfields were acquired every 20–30 seconds to generate time-lapse series . Data was analyzed with Imaris 7 . x ( Bitplane Scientific Software ) and processed with MAGIX Video deluxe 2013 . 7 µm-thick lung sections of 2–4 animals per time-point and group were performed at comparable anatomical positions ( central lung , slices including right and left lobes and main bronchi ) . “Viable” infected cells were distinguished from cell “remnants” by the following criteria: morphology ( smooth edge , round shaped with or without elongations ) , nucleus ( clear non-fragmented DAPI signal present ) and cell size ( larger than 5 µm ) ( Figure 3F ) . Mean of two counted slices per animal was calculated for Figure 4D and E . Area of inflammation was determined by manual measurement of CD45+ stain signal using AxioVision 4 . 8 software ( Figure 4F ) . Neonatal mice were subcutaneously injected with 125 µM EdU and sacrificed within 4 h after injection . Histological staining was performed with Click-iT EdU Imaging Kit ( Invitrogen ) . Statistical analysis was performed with Prism 4 ( Graph-Pad Software , Inc . ) . Unpaired t-test for comparison of 2 groups or ANOVA one-way analysis for >2 groups . Statistical significance was depicted as follows: * , p<0 . 05; ** , p<0 . 01; and *** , p<0 . 001 .
|
Neonates are highly susceptible to a number of infections that usually cause disease only in immunocompromised individuals , most likely because of their incompletely developed immune system . Although this phenomenon has been frequently observed , immune responses of neonates remain largely undefined upon infections with viruses . There is lack of knowledge about the spatio-temporal dynamics of host-virus interaction , especially in comparative infection models of neonates and adults . In this study , with the use of virus reporter mutants , we provide elaborate insight into these aspects in the mouse model of CMV infection . We define hallmarks of virus tropism , early cellular immune responses and general infection dynamics , findings that are fundamental to understand neonatal antiviral immunity . Furthermore , we found that neonatal APCs induce T cell responses in nodular inflammatory foci of the lung , a process which was supposed to be restricted to lymphoid organs . However , the MCMV-specific T cell response was qualitatively different in neonates from that in adults , possibly explaining - in part - the higher susceptibility of newborns . These observations expand our understanding of where adaptive immunity can be initiated , highlights the importance of early local cellular immune responses and sheds more light on neonatal antiviral immunity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
Nodular Inflammatory Foci Are Sites of T Cell Priming and Control of Murine Cytomegalovirus Infection in the Neonatal Lung
|
Prion diseases are infectious neurodegenerative disorders of humans and animals caused by misfolded forms of the cellular prion protein PrPC . Prions cause disease by converting PrPC into aggregation-prone PrPSc . Chronic wasting disease ( CWD ) is the most contagious prion disease with substantial lateral transmission , affecting free-ranging and farmed cervids . Although the PrP primary structure is highly conserved among cervids , the disease phenotype can be modulated by species-specific polymorphisms in the prion protein gene . How the resulting amino-acid substitutions impact PrPC and PrPSc structure and propagation is poorly understood . We investigated the effects of the cervid 116A>G substitution , located in the most conserved PrP domain , on PrPC structure and conversion and on 116AG-prion conformation and infectivity . Molecular dynamics simulations revealed structural de-stabilization of 116G-PrP , which enhanced its in vitro conversion efficiency when used as recombinant PrP substrate in real-time quaking-induced conversion ( RT-QuIC ) . We demonstrate that 116AG-prions are conformationally less stable , show lower activity as a seed in RT-QuIC and exhibit reduced infectivity in vitro and in vivo . Infectivity of 116AG-prions was significantly enhanced upon secondary passage in mice , yet conformational features were retained . These findings indicate that structurally de-stabilized PrPC is readily convertible by cervid prions of different genetic background and results in a prion conformation adaptable to cervid wild-type PrP . Conformation is an important criterion when assessing transmission barrier , and conformational variants can target a different host range . Therefore , a thorough analysis of CWD isolates and re-assessment of species-barriers is important in order to fully exclude a zoonotic potential of CWD .
Prion diseases are fatal neurodegenerative disorders including Creutzfeldt-Jakob disease in humans , bovine spongiform encephalopathy in cattle , scrapie in sheep and goats and chronic wasting disease ( CWD ) in cervids [1 , 2] . According to the “protein only” hypothesis , prions are mainly , if not solely , composed of PrPSc [3] , a misfolded isoform of the host-encoded prion protein , PrPC . PrPSc results from the conversion of PrPC into a conformation enriched in β-sheets [4–6] which can act as a seed to bind and convert other PrPC molecules . These are incorporated into a growing polymer [7 , 8] which breaks into smaller oligomers , resulting in higher numbers of infectious nuclei . PrPSc is prone to aggregation and is partially resistant to proteases [9 , 10] . Despite the absence of a nucleic acid genome , various prion strains have been identified , displaying specific biological properties [11] . Prion strains can be differentiated by incubation time , clinical signs of the disease and biochemical properties such as conformational stability of PrPSc within one host species [12–14] . CWD affects elk ( Cervus canadensis ) , mule deer ( Odocoileus hemionus ) , white-tailed deer ( WTD; Odocoileus virginianus ) and moose ( Alces alces ) [15 , 16] . It is considered the most contagious prion disease with horizontal transmission favored by cervid interactions and environmental persistence of infectivity [17 , 18] . The disease is present in North America , South Korea and , recently , has been identified for the first time in Europe in free-ranging reindeer ( Rangifer tarandus tarandus ) in Norway [19] . The transmission of prions is governed by a species barrier between species , or a transmission barrier within the same species [20 , 21] . These barriers are regulated by the PrP primary structures of donor and recipient and emphasize the role of PrP structure in disease susceptibility . Species-specific allelic variants or polymorphisms were identified in cervid Prnp . WTD PrP can differ at residues 95 , 96 and 116 [22–24] . Mule deer PrP has a unique polymorphism at residue 225 . The contiguous residue 226 encodes the singular difference between deer and elk PrP and regulates strain selection [25] . In elk , PrP possesses a polymorphism at residue 132 [26 , 27] . Most polymorphisms have in common that the presence of a single non-wildtype allele is associated with reduced susceptibility to CWD . Transfer of prion strains between or within ruminant species expressing different PrPC primary structures can result in the acquisition of novel transmission properties [26 , 27] . This raises concerns on the zoonotic potential of prion agents from species consumed by human populations . The polymorphism at residue 116 ( A/G ) in WTD is of particular interest . The wild-type ( wt ) genotype encodes an alanine ( A ) which is highly conserved among species [28] . The variant encodes a glycine ( G ) ; however , the influence of this mutation on CWD susceptibility is unclear . Whereas one study reports on no differences in the incidence of CWD in heterozygous WTD [22] , others indicate that the 116G allele is found at a lower frequency in prion-infected than uninfected WTD [29] . Residue 116 ( 113 in human PrP ) is localized in the central hydrophobic core ( HC ) of PrPC , which is the most conserved domain of PrP and critically involved in prion conversion [28 , 30–32] . Mutations in the HC of human PrP are associated with heritable prion disease [33] . Therefore , we aimed to achieve a better understanding of potential structural changes in PrPC caused by this single amino acid substitution , how it affects conversion efficiency and subsequently , CWD prion conformation , infectivity and pathogenesis in vitro and in vivo . In wild cervids the most likely situation is that CWD prions from a certain host are transmitted to animals that express wt PrP . Our data demonstrate that prions from an animal with a PrP polymorphism can efficiently adapt to the wt PrP sequence . This may be due to a lower conformational stability of the WTD-116AG prions , which is likely the result of the predicted structural flexibility of PrP-116G , thus making it more prone to be converted . This is supported by an improved convertibility when recombinant PrP-116G is used as a substrate for conversion in real-time quaking induced conversion ( RT-QuIC ) assay . Our findings do not support earlier reports indicating a decreased susceptibility of WTD encoding the 116G allele to CWD infection , but indicate that investigations on adaptability and possible transmission of CWD prions with low conformational stability to non-cervid species are required .
We analyzed the Prnp genotypes of CWD-positive samples from Saskatchewan WTD ( S1 Fig ) and identified one isolate heterozygous at residue 116 , encoding A and G at this site . Single amino acid substitutions can substantially alter prion properties . Thus , we decided to assess the biochemical characteristics of wt and 116AG WTD CWD prions . We compared the proteinase K ( PK ) resistance of the two isolates by digestion of brain homogenates with different concentrations of PK ( 0 to 5 mg/ml ) . Western blot analysis indicated a dose-dependent decrease of the PrPres signals ( Fig 1A and 1B ) , significantly starting at 50 μg/ml of PK for the 116AG isolate and at a 10fold higher PK concentration for the wt isolate ( Fig 1B ) . Since differential PK resistance can arise from conformational variability , we compared the conformational stability of WTD isolates . The conformational stability assay ( CSA ) is a reliable tool to compare different prion isolates and distinguish prion strains [34] . It is used to determine the concentration of GdnHCl which is necessary to unfold 50% of the PrPSc ( referred to as [GdnHCl1/2] ) . The [GdnHCl1/2] of 116AG and wt prions were significantly different ( n = 5 , *P < 0 . 05 ) with 1 . 95 M for wt and 1 . 5 M for 116AG ( Fig 1C ) . A second WTD-wt isolate also showed a [GdnHCl1/2] of 2 M ( S2 Fig ) . This experiment revealed a significantly higher conformational stability of wt PrPSc compared to 116AG PrPSc . We took advantage of RT-QuIC assay to compare the seeding and amplification characteristics of wt and 116AG prion seeds , respectively , using different recombinant PrP ( rPrP ) substrates as well as by looking into informative parameters such as endpoint dilution , lag and log phase [35] . Serial dilutions ( 2x10-2 to 2x10-7 ) of wt and 116AG brain homogenates were used to seed quadruplicate RT-QuIC reactions with deer or mouse rPrPs as a substrate ( Fig 2A ) . Dilutions of which at least 50% of the replicates crossed a cut-off of approximately 50 , 000 RFU ( relative fluorescence units ) were considered positive . Endpoint dilution analysis of wt and 116AG prions ( 5 independent experiments ) showed that 116AG prions had a 100fold lower dilution endpoint than wt prions with both deer and mouse rPrP substrates ( Fig 2A ) . To confirm that the lower dilution endpoints were related to the seeding activity of those isolates and not due to a different initial amount of PrPSc in each isolate , we serially diluted the two isolates after PK digestion . The subsequent Western blot analysis ( S3 Fig ) indicated that wt and 116AG brain homogenates harboured comparable amounts of PrPres . Kinetics of conversion using 116AG prions as a seed resulted in extended lag and log phases compared to wt prions regardless of the rPrP substrate ( Fig 2B and 2C ) . The graphs in Fig 2B illustrate the time individual replicates needed to reach the RFU threshold , which signifies the lag phase . The y-axis ( Fig 2B ) represents the percentage of replicates ( individual RT-QuIC reactions ) that did not yet reach the threshold . At the beginning of the reaction , 100% of the replicates were negative and over time , the percentage of those replicates decreased incrementally . The curves indicated that the lag phase in 116AG seeded reactions is significantly extended starting with the 2x10-4 dilution , regardless of rPrP substrate ( Fig 2B ) . Finally , we calculated the time needed for each condition ( seed , substrate , dilutions ) to double the RFU during the amplification phase of the reaction to characterize the log or elongation phase which signifies the efficiency of amyloid fibril formation . Independent of the rPrP substrate used , 116AG prion seeds resulted in a significantly longer log phase ( 1 . 5 to 2 . 5 times ) compared to wt prion seeds ( Fig 2C ) . Besides , to investigate the infectivity of WTD isolates , we established primary cerebellar granular neuron cultures ( CGN ) derived from newborn tg1536+/+ mice [36] . We exposed CGN cultures to infected brain homogenates of either wt or 116AG WTD . Non-infected brain homogenates from WTD served as a negative control . We observed a progressive accumulation of PrPres in infected CGNtg1536+/+ cultures ( Fig 3 ) . In CGNtg1536+/+ cultures infected with wt CWD prions , PrPres accumulated as early as 7 days post infection ( dpi ) and progressively increased until the endpoint of the experiment ( 28 dpi; Fig 3 , left panel ) indicating a more efficient propagation compared to CGNtg1536+/+ cultures infected with 116AG prions ( Fig 3 , right panel ) . This result indicated that 116AG prions have a reduced infectivity in CGN cultures . Of note , this is the first evidence of primary neuronal cultures being infected with CWD agents . We performed transmission studies in transgenic tg1536+/+ mice overexpressing deer PrP approximately six to eightfold [36] . For each group , at least nine female mice were inoculated intracerebrally with WTD brain homogenates . Tg1536+/+ mice inoculated with wt prions had a significantly shorter incubation period than mice inoculated with 116AG prions ( P < 0 . 001 , 241 ± 20 vs 301 ± 26; S4 Fig , Table 1 . ) . Clinical presentation was similar in all mice; however , disease progression and duration of the clinical phase was clearly different , with one week in mice inoculated with WTD-wt and three months upon WTD-116AG infection . In addition , to assess differences between groups of mice inoculated with either of the two WTD isolates , coronal brain sections were examined histologically for spongiform changes and PrPSc aggregate distribution . Spongiosis and abnormal PrP aggregates were observed in cortex , hippocampus , corpus callosum , the habenular , thalamic and hypothalamic nuclei of two tg1536+/+ mice inoculated either with wt or 116AG prions . Despite a more intense staining upon inoculation of wt prions , PrPSc aggregates showed no consistent differences in terms of types of deposits between wt and 116AG inocula ( S5A and S5B Fig ) . Vacuolation in the cortex ( S5E and S5F Fig ) of a mouse inoculated with 116AG prions was less pronounced compared to the vacuolation of a mouse inoculated with wt prions ( S5C and S5D Fig ) . Similar to the characterization of WTD isolates , we analyzed the brain homogenates of the first passage in mice referred to as mWTD . Interestingly , assessment of PK resistance demonstrated that mouse brain extracts were slightly more resistant to PK compared to the WTD isolates regardless of the genotype of the inoculum ( wt or 116AG; Fig 4A ) . However , mWTD-wt ( blue bar ) brain homogenates harbored PrPSc which resisted digestion with higher PK concentrations of 5 mg/ml compared to mWTD-116AG ( red bar ) prions ( Fig 4B ) . CSA analysis of mWTD prions was in line with the findings for the WTD isolates . Although the [GdnHCl1/2] of PrPSc of mWTD prions was increased compared to the original isolates , mWTD-wt and -116AG prions still differed in conformational stability , comparable to what was observed in the WTD isolates ( Fig 4C ) . Again , mWTD-wt prions were significantly more stable than mWTD-116AG ( [GdnHCl1/2] 2 . 25 M for mWTD-wt vs . 1 . 85 M for mWTD-116AG; P < 0 . 05 ) . Next we performed RT-QuIC and endpoint dilution analyses , and it was evident that the seeding activities of mWTD-wt or mWTD-116AG prions ( 7 independent experiments ) in mouse brain homogenates ( Fig 5A ) were higher than those of the original isolates ( Fig 2A ) . In mWTD-wt seeded reactions , all dilutions up to 2x10-7 were positive regardless of the substrate . Brain homogenates containing mWTD-116AG prions had an at least 10fold lower dilution endpoint with both rPrP substrates ( Fig 5A ) despite comparable amounts of PrPres ( S6 Fig ) . The conversion kinetics in reactions seeded with mWTD-116AG prions were different from reactions seeded with mWTD-wt prions . Similar to the original WTD isolate seeds , the lag phase when seeding the reactions with mWTD-116AG prions was extended from the 2x10-4 dilution , independent of the substrate ( Fig 5B ) . However , except for one of the higher dilutions , the log phases of mWTD-116AG and mWTD-wt seeded reactions were similar with both substrates ( Fig 5C ) . The fact that this difference in the log phase was apparent with the high dilutions only reflects a very efficient recruitment and conversion of rPrP molecules to the elongating amyloid fibrils once sufficient seeds are formed during the lag phase . We infected CGN cultures derived from tg1536+/+ mice with mWTD-wt and mWTD-116AG or uninfected mice and assessed PrPres accumulation over time ( S7 Fig ) . Western blots showed a progressive accumulation of PrPres in cultures infected with mWTD-wt starting at 14 dpi until the endpoint of the experiment ( 28dpi; S7 Fig , left panel ) . Results were consistent between duplicates up to 14 dpi , and despite a discrepancy between the duplicates at 21 and 28 dpi in CGN infected with mWTD-wt , this preliminary result indicates a more efficient propagation compared to cultures infected with mWTD-116AG prions , where PrPres signals were weaker and increased slightly only after 21 dpi ( S7 Fig , right panel ) . Given the described differences in biochemical and amplification properties that were retained even on an identical host PrP genotype background , we wanted to verify whether this might reflect a novel CWD strain in the WTD-116AG isolate . Therefore , we performed a secondary passage of mWTD-wt or -116AG in tg1536+/+ mice ( Table 1 ) . Incubation times upon inoculation of tg1536+/+ mice with mWTD-116AG were significantly shorter than in the primary passage ( P < 0 . 0001 , 213 dpi ± 38 vs 301 dpi ± 26; Table 1 and S8 Fig ) . Surprisingly , 80% ( 4/5 ) of the mice inoculated with mWTD-116AG had a shorter incubation period compared to mice inoculated with mWTD-wt ( S8 Fig ) . Disease progression of most of the mice inoculated with mWTD-116AG prions was shorter than observed upon first passage , except for one mouse with the longer incubation period . Notably , incubation times between primary and secondary passage of WTD-wt prions were comparable , with rapid disease progression ( Table 1 and S8 Fig ) . To verify the conversion proficiency of the PrP-116G allele in seeded amyloid formation , we used the RT-QuIC technique to assess the ability of rPrP-116G to be converted by different CWD prion seeds . We used wt and 116AG prions ( Fig 6A ) , respectively , as seeds and either rPrP-wt ( 116A ) or rPrP-116G as substrates in RT-QuIC reactions . We found that rPrP-116G substrate was more efficiently converted ( Fig 6A ) than rPrP-wt . This is evident from an increased maximum RFU and a significantly shorter log phase for rPrP-116G substrate independent of the used CWD seed ( Fig 6B ) . Despite the more efficient amyloid formation and conversion of rPrP-116G , which might indicate a greater ability for adaptation to any CWD seeds , we noticed that storage at -80°C negatively impacted RT-QuIC results with rPrP-116G substrate . This provided a hint that rPrP-116G is less stable than rPrP-wt and led us to perform molecular dynamics simulations of both deer PrPs for verification . Computer simulations of protein dynamics provide valuable data to complement laboratory experiments [37] . The experimentally determined structure of the natively folded PrP was used to simulate conformational changes and physically realistic protein dynamics , including initial steps of misfolding [37] . The effect of pathogenic mutations on the molecular dynamics and conformation of PrP may be evaluated by applying mutations in silico . Thus , with tens of nanoseconds of MD simulation , the dominant contributions to intramolecular , atomic movements can be determined , making MD a useful and attractive method to analyze effects of single amino acid substitutions on PrP stability and identifying structurally flexible regions [38 , 39] . For both wt ( Fig 7A ) and the 116G polymorphism ( Fig 7B ) simulations were done with three different starting configurations of atom velocities ( labeled R1 , R2 and R3 ) to ensure reproducibility and consistency of the results . The final trajectories were analyzed specifically focussing on the structural stability and dynamic of both systems . The root mean square deviation ( RMSD ) of the backbone Cα atoms for the folded domains of wt and the 116G polymorphism were calculated and plotted in Fig 7C . The graph for the RMSD represents the average of the three independent MD runs over time , with the error bars corresponding to the standard deviation between the separate simulations . Comparing the RMSD values , it became apparent that the 116G polymorphism had higher RMSD values than wt and a much larger variation between individual simulations , indicating that the 116G conformation has more pronounced structural fluctuations than those of wt PrP . Another parameter determining the dynamic stability of a protein is the root mean square fluctuation ( RMSF ) value , which represents the flexibility of a specific residue around its average position . We calculated RMSF values for all six simulations and plotted averaged results for wt and 116G , with the error bars corresponding to the standard deviation between the separate simulations . Fig 7D shows that the 116G polymorphism affects the RMSF of most residues , but particularly those close to the 116G exchange , making the structure of 116G PrP less stable . In particular , we observed a substantially increased destabilization around the second β-strand and in loops LS2H2 , LH1S2 , LS1H1 . By calculating the radius of gyration ( Rg ) we determined the effective size of the proteins throughout their simulation ( 50 ns ) . This approach gave us a picture of the protein’s folding behaviour . To compare the packing behaviour of both wt and 116G , the radius of gyration of all Cα atoms was calculated and plotted . Fig 7E presents the Rg values for the final 10 ns of the simulation . It is readily apparent that the 116G polymorphism had a generally larger Rg value compared to the wt , implying that the 116G substitution affects the conformation of the whole protein . It is well-known that protein folding is an important contributor to protein stability , thus the more compactly folded wt protein may be more stable than the 116G polymorphism . The secondary structure ( SS ) plugin of the VMD analysis package [40] was used to assign the SS elements for all rounds of simulation . Subsequently , the dominant SS element for each residue for the last 20 ns of simulation was evaluated . However , only the percentage of α-helices , β-strands , and β-bridges was plotted ( Fig 7F , graphs represent the average of SS assignments from all three simulations for each system , with a more detailed analysis for all three simulations shown in S9 ) . New SS elements appeared in the N-terminal regions of both wt and 116G polymorphism ( residues 112 to 125 ) . In wt the N-terminal SS consisted predominantly of α-helices , while in 116G the SS elements were reduced by about ~12% and consisted mostly of β-strands and β-bridges . Furthermore , the first β-strand ( S1 ) was longer in wt PrP compared to the 116G polymorphism ( residues 131 to 137 and 131 to 135 , respectively ) . In contrast , the lengths of the first α-helix ( H1 ) and the second β-strand ( S2 ) were very similar in both systems . Another new SS element , which was seen in both wt and 116G involved residues 168 to 172 . In wt PrP residues 169 to 171 were mostly α-helical ( ~70% ) and to a lesser degree ( ~30% ) in a β-strand / β-bridge conformation . However , in 116G these residues were seen in equal proportions in both conformations ( ~50% for each ) , meaning they were more likely to be in a β-strand conformation than for wt PrP . The second helix ( H2 ) was very much preserved in both wt and 116G with a partial disruption of its final residues ( residues 193 to 198 ) . The same trend was seen in the third helix ( H3 ) , residues 223 to 228 had lost their helical structure for about 50% of the time in the simulation . Also , residues 200 to 202 had gained α-helical structure with an elongated H3 in both wt and 116G . An overall assessment of the SS throughout the simulations showed that PrP with the 116G polymorphism exhibited signs of structural disruption compared to wt PrP . Meaning the amino acids in wt PrP were more likely to be folded ( α-helix or β-sheet ) , rendering the protein more stable . In addition , the new SS elements ( residues 112 to 125 , and 168 to 172 ) of 116G preferred to adopt a β-structure ( β-strand / β-bridge ) , while wt PrP preferred α-helical structure . This finding matched previous MD studies indicating that mutated forms of the prion protein can adopt higher β-sheet content [41] . The averages for the total solvent accessible surface area ( SASA ) for wt and 116G PrP were computed over 50 ns of simulation and plotted in Fig 7G . 116G PrP had a slightly larger SASA value compared to wt PrP , indicating that the polymorphism made the structure of PrP more solvent exposed . In addition , a minor increase in exposure of hydrophobic residues of the PrP structure was seen in 116G compared to wt ( Fig 7H ) . An increase in hydrophobic exposure is an indicator of weaker interactions between solvent and protein , thereby rendering a protein more likely to undergo self-assembly [41] .
Polymorphisms in the prion protein gene Prnp are important modulators of susceptibility to and pathogenesis of prion diseases [42] . Here , we characterized the effect of the 116A>G polymorphism identified in a CWD isolate from a 116AG heterozygous WTD . We discovered that 116AG prions had a lower conformational stability , seeding activity in RT-QuIC and infectivity in CGN and mouse bioassays than wt ( 116AA ) prions . Biochemical differences and certain amplification characteristics were retained following passage in transgenic mice overexpressing wt deer PrP . This may indicate the isolation of diverse CWD strains arising from wt and 116AG CWD isolates . Notably , this is the first study revealing differences in conformational stability between PrPSc of CWD isolates , and moreover , even after passage in mouse models . Stability of PrPSc is associated with prion conformation and a widely accepted hallmark of strain variability [11 , 43 , 44] . Furthermore , we determined quantitative parameters for RT-QuIC to characterize seeding and amplification characteristics and established for the first time CGN cultures as a novel cell culture model for CWD infection . Previous studies have been contradictory with respect to the relationship between conformational stability of PrPSc and incubation periods [45–47] . Legname et al . showed an inverse correlation between stability of PrPSc aggregates and incubation time [47] . Efficient cell-free conversion of ovine PrPC into protease resistant forms was correlated to polymorphisms conferring susceptibility to scrapie [48] . We demonstrate that upon second passage of mWTD-116AG prions , lower conformational stability was correlated to a shorter incubation time . The fact that the differences in PK resistance and conformational stability between wt and 116AG prions were retained upon passaging indicates that the delay in incubation period observed in the primary transmission of 116AG probably was due to the single amino acid difference between inoculum and host PrP , creating a transmission barrier . This pattern was reflected by the duration of the log phase in RT-QuIC indicative of elongation of amyloid fibers , which was attenuated and almost equal between mWTD-wt and -116AG seeded reactions , in contrast to the significantly prolonged log phase of the WTD-116AG seeds in the WTD isolate . Heterozygosity of WTD could also lead to a co-existence of two conformers of PrP , wt and 116G , thus leading to competition and/or selection during prion fiber formation . Propagation in the CGN model still differed between the two mWTD prion inocula . This apparent discrepancy to the shortened incubation time in vivo and log phase in RT-QuIC validates that onset of clinical disease not necessarily correlates with in vitro propagation and is influenced by regional or host factors , or differential toxicity of prions . It also might reflect the differences observed in the lag phase of the RT-QuIC reactions in that sufficient seed formation is delayed and the mWTD-116AG infected CGN cultures not yet reached the exponential propagation phase . The 116 ( A>G ) polymorphism introduces a variation into the highly conserved hydrophobic core region ( HC ) of PrP [28] . One mutation in this domain ( 117A>V ) causes a heritable prion disease of humans where disease progression was not associated to amyloid formation [33] . The HC can facilitate PrPC-PrPSc interaction and the formation of β-sheet structure [49 , 50] , can act as a hinge region involved in PrPC-PrPSc conversion [30 , 31] and is of importance for the interaction with lipids which may facilitate the conversion of PrPC into PrPSc [32] . Intriguingly , using rPrP-116G substrate in RT-QuIC resulted in a more efficient conversion independent of the used seed compared to the rPrP-wt substrate . Substitution of A with G replaces a hydrophobic , aliphatic residue ( A ) with a more polar residue ( G ) . Thus , PrPC-wt may be capable of better van der Waals packing . In fact , MD simulations revealed that the structure of PrPC-116G is less stable than PrPC-wt and exhibits structural disruption and more solvent accessible surface area . This indicated a more flexible structure of PrPC-116G possibly more prone to conversion , resulting in prions with a lower conformational stability and PK resistance of original and mouse-passaged WTD-116AG prions . Interestingly , MD analysis indicated that the 116G allelic variant has a higher probability to form β-strands in the rigid loop region . The structural rigidity of this PrP domain has been discussed to be predictive of susceptibility to CWD infection , with higher flexibility of the loop conferring resistance [51] . Altogether , our data argue against a reduced susceptibility of WTD carrying the 116G allele , based on the increased effectiveness of amyloid formation of rPrP-116G , the predicted higher propensity of β-strand elements in the rigid loop region , and the new β-strand elements in N-terminal PrPC-116G ( residues 112 to 125 ) , predicted to be mainly α-helical in PrPC-wt . As a note of caution , with cell-free conversion assays dose , prion strain and route of infection cannot be considered , which all affect the ability of PrPC and PrPSc to interact and the progression of clinical disease . CWD strains are poorly documented . To date , there are few studies describing the existence of CWD strains [25 , 52] and the emergence of CWD strains with novel transmission properties [53] . First evidence of CWD strains was provided when transgenic cervid mice inoculated with a variety of CWD isolates revealed the identification of two CWD strains , termed CWD1 and CWD2 [25] . These two strains , with different incubation periods , exhibited indistinguishable biochemical properties of PrPSc . Interestingly , primary transmission of elk isolates displays either CWD1 or CWD2 profiles , but deer tend to harbor a CWD1/CWD2 mixture which was attributed to the singular difference between elk ( 226E ) and deer ( 226Q ) PrP and strain mutation [25] . Oral transmission of wt/wt deer prions into WTD expressing different PrPC primary structures ( wt/wt , S96/wt , H95/wt or H95/S96 ) [54] resulted in emergence of a strain ( H95+ ) with novel transmission properties in deer expressing H95-PrPC [53] . Transmission of these deer isolates into transgenic mice expressing deer wt ( tg33 ) or S96 PrPC ( tg60 ) resulted in differential strain propagation and revealed that H95/wt and H95/S96 deer accumulated a strain mixture ( Wisc-1 and H95+ ) while wt/wt and S96/wt carry the Wisc-1 strain mostly [53] . These results suggest evolution of cervid strains occuring by transmission between hosts expressing different PrPC primary structures . In contrast to this , Prnp-independent alterations of CWD strains upon transmission to different host species expressing the same PrPC were reported , highlighting the relevance of host-specific factors [55] . Our study favors the emergence of a new strain based on structural differences of PrPC-116G that could overcome the mutation effect of 226Q [25] and adapted efficiently to the wt PrP structure upon secondary passage in tg1536+/+ mice . Given the frequencies of non-wt alleles in cervids , transmission of CWD prions from animals expressing PrP allelic variants to animals harboring wt PrP is the most likely scenario among free ranging cervids . We suggest that CWD transmission between cervids expressing distinct PrPC molecules can introduce prion conformational variability that may modify transmission properties . Nevertheless , controlled oral inoculation as well as the generation of mouse models expressing PrPC-116G will be required to determine the degree to which the 116G allele affects strain properties and susceptibility to CWD infection . Our data add new insights into the association between structural stability of PrPC , conversion efficiency in vitro , conformational stability of CWD prions and infectivity and adaptation in vitro and in vivo . It improves our understanding of the impact of single amino acid substitutions on predicted structural properties and convertibility , suggesting that higher flexibility aligns with improved conversion but results in prions with a less stable conformation . However , this instability may enhance adaptation to PrPs of non-cervid species . Although until to date CWD prions have failed to transmit disease to transgenic mice expressing human PrP [56 , 57] as well as to cynomolgus macaques [58] suggesting that humans are resistant to CWD , this risk cannot be totally excluded . CWD prions arising from cervids expressing allelic variants that are converted to conformationally instable prions could be the missing link to determine the real zoonotic potential .
All isolates used in this study were prepared as 10% ( W/V ) brain homogenates in phosphate-buffered saline pH 7 . 4 ( PBS ) using a dounce homogenizer . Aliquots were stored at -80°C . The wt isolate was obtained by experimental oral infection [54] , and the 116AG from a free ranging animal at the terminal stage of disease . All work with animals was performed in compliance with the University of Calgary Animal Care Committee under protocol numbers AC14-0025 ( inoculation ) and AC14-0117 ( primary cultures ) and CCAC guidelines . The University of Calgary Animal Care Committee approved the study . The transgenic mouse line tg ( CerPrP132M ) 1536+/+ overexpressing wt deer PrP [34 , 36] was used to propagate the two WTD isolates . The relative level of PrP overexpression in the brain of these mice was about six to eightfold [36] . Six to eight weeks old female mice were anaesthetized and inoculated with 20 μl of a 1% brain homogenate of either of the two WTD isolates in the right parietal lobe using a 25 gauge disposable hypodermic needle . Mice were initially monitored weekly and daily when progressive clinical signs of prion disease were evident . At the experimental endpoint , animals were anaesthetized before being euthanized by CO2 overdose and their brains were collected and frozen at -80°C . For the second passage , animals were inoculated with 20 μl of a 1% brain homogenate of either of the two mWTD prions , corresponding to first passage mouse brain homogenate . Incubation time was expressed as the mean value of the days post inoculation ( dpi ) for all mice that tested positive for PrPSc after PK digestion ( PrPres ) . The statistical analysis of transmission experiments was performed using GraphPad Prism ( version 5 ) software and using the Mann-Whitney test . Brain tissues of one mouse each inoculated with either of the WTD isolates were fixed and paraffin embedded for histopathology . Coronal brain sections were performed and slices ( 4 to 6 μm thick ) were stained with hematoxylin and eosin to evaluate the sections for spongioform degeneration and immunostained for PrPSc deposition using mAb BAR224 . Briefly , brain slides were pretreated with high-pressure autoclaving ( 2 . 1 × 105 Pa ) for 30 min in citric acid ( 10 mM ) , pH 6 . 0 , at 121°C , followed by treatment with 98% formic acid for 30 min and 4 M guanidine thiocyanate for 2 h at room temperature . Tissue sections were scanned with a NanoZoomer 2 . 0RS scanner ( Hamamatsu Photonics ) and analyzed using NanoZoomer digital pathology software ( Hamamatsu Photonics ) . The anti-PrP monoclonal mouse antibodies used in this study were 4H11 ( 1/500; [59] ) or BAR224 ( 1/10 , 000; Bertin Pharma , Fr ) . For PrP analysis in brain extracts , brain homogenates ( 3 different animals ) prepared in PBS were either not digested or treated with different concentrations of PK ( 0 to 5 mg/ml; VWR , Ca ) as indicated for one hour at 37°C . The reaction was terminated by adding 1X pefabloc proteinase inhibitor ( VWR , Ca ) . Fifty μg of protein were separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis ( SDS-PAGE ) , and then electrophoretically transferred to PVDF membranes ( Millipore , Ca ) . PVDF membranes were probed using anti-PrP monoclonal antibodies followed by horseradish peroxidase-conjugated goat anti-mouse IgG antibody ( Sigma , Ca ) and developed using ECL-plus detection ( Amersham ) . Images were acquired on X-ray film ( Super Rx; Fujifilm ) or by using a digital imaging system ( Alpha Innotech , FluoriChemQ ) . FluoChemQ software ( Alpha Innotech ) was used to quantify and determine the relative values of PrPres signals . CSA was performed as previously described [34] with slight modifications . Briefly , 10% brain homogenates from WTD or mice upon primary passage of the WTD isolates ( 3 different animals ) were incubated with various concentrations ( 0 to 4 M ) of GdnHCl ( Sigma , Ca ) for 1 hour at 20°C under shaking conditions ( 450 rpm ) . Then samples were treated with 50 μg/ml of PK for an additional hour at 37°C and the reaction was stopped by adding 1X pefabloc proteinase inhibitor . The samples were then subjected to Western blot and PrPres signals were quantified as described above . The relative values of PrPres ( 5 independent experiments ) were plotted as a sigmoid curve against the GdnHCl concentration using GraphPad Prism ( version 5 ) . The GndHCL concentration required to denature 50% of PrPSc [GdnHCl1/2] was deduced from these curves . The statistical analysis to compare the different isolates was performed using GraphPad Prism ( version 5 ) software using unpaired student’s t-test . In this study we used the mature forms of deer , wt ( aa 24–234; construct kindly provided by B . Caughey , NIH Rocky Mountain Laboratories , Hamilton , MT ) , 116G ( aa 24–234 ) or mouse ( aa23-231 ) PrP cloned into pET expression vectors and expressed in E . coli Rosetta using the Express Autoinduction System ( Novagen ) . The 116G mutant was created by site-directed mutagenesis of the wild-type deer PrP in the pET plasmid . Inclusion bodies were prepared using the Bug Buster reagent ( Novagen ) and solubilized in lysis buffer ( guanidine-HCl 8 M , Na-phosphate 100 mM , Tris-HCl 10 mM , pH 8 . 0 ) for 50 min at room temperature and then centrifuged at 16 , 000 x g for 5 min at room temperature . Binding , refolding and elution using an AKTA Explorer system has been previously described [60] . Real time QuIC was performed as described [60–62] . Briefly , reactions were set up in assay buffer containing 20 mM sodium phosphate ( pH 6 . 9 ) , 300 mM NaCl , 1 mM EDTA , 10 μM Thioflavin T and 0 . 1 mg/ml rPrP substrate . Ninety-eight μl aliquots were added to the wells of a 96 well optical bottom plate ( Nalge Nunc International ) . Quadruplicate reactions were seeded with 2 μl of brain homogenate ( 10% ) from CWD-negative animals or CWD-WTD and mCWD-WTD isolates that were 10-fold serially diluted in RT-QuIC seed dilution buffer ( 20 mM sodium phosphate ( pH 6 . 9 ) , 130 mM NaCl , 0 . 1% ( w/v ) SDS , 1X N2 Supplement ( Invitrogen ) ) . The plate was sealed with Nunc Amplification Tape ( Nalge Nunc International ) and placed in a BMG Labtech FLUOstar Omega fluorescence plate reader that was pre-heated to 42°C for a total of 50 hours with cycles of 1 minute double orbital shaking ( 700 rpm ) incubation and 1 minute resting throughout the incubation . ThT fluorescence signals of each well were read and documented every 15 minutes then the values were plotted as the average of quadruplicate reactions by using GraphPad Prism software . CGN were mechanically extracted from the cerebella of 5 to 7-day-old tg1536+/+ mice and enzymatically dissociated , as previously described [63] . Briefly , cells were plated at a density of 1 . 9 x 103 cells/mm2 on plastic culture wells precoated with 10 μg/ml poly-D-lysine ( Sigma-Aldrich , Ca ) . Cells were cultured in Dulbecco’s modified Eagle’s medium-Glutamax I high glucose ( DMEM ) ( Life Technologies-Gibco , Ca ) supplemented with penicillin and streptomycin ( Life Technologies , Ca ) , 10% fetal bovine serum ( Life Technologies , Ca ) , 20mM KCl ( Sigma-Aldrich ) , and N2 and B27 supplements ( Life Technologies , Ca ) . Cells were incubated at 37°C in a humidified 5% CO2 atmosphere . Every week , the medium was supplemented with glucose ( 1 mg/ml ) ; in addition , the antimitotics uridine and fluorodeoxyuridine ( 10 μM ) ( Sigma-Aldrich ) were added to reduce astrocyte proliferation . CWD-negative brain homogenate or WTD ( 3 independent experiments ) and mWTD brain homogenates ( one experiment ) were added to CGN cultures as described [63 , 64] . Briefly , brain homogenates were sonicated and added at a final concentration of 0 . 01% to CGN cultures 48 h after plating . Four days later , the medium was removed from the cultures , and cells were washed twice in fresh culture medium . Fresh medium was then added , and no medium changes were performed for the remaining experiments . On different days post infection ( dpi ) , cells were washed twice with PBS and then incubated in lysis buffer ( 50 mM Tris-HCl [pH 7 . 4] , 0 . 5% Triton X-100 , 0 . 5% sodium deoxycholate ) ( Sigma-Aldrich , Ca ) for 10 min at 4°C . The protein concentration of each cell lysate was measured with the bicinchoninic acid ( BCA ) protein assay ( ThermoFisher Scientific-Pierce , Ca ) . Next , as previously described [63 , 64] , 50 μg of protein were digested with 5μg/ml of PK ( VWR , Ca ) for 30 min at 37°C , and the reaction was stopped by adding 1X pefabloc proteinase inhibitor to the mixture . Proteins were precipitated by the addition of methanol for 1 h at -20°C . The samples were then centrifuged at 16 , 000 xg ( 5417R rotor; Eppendorf ) for 20 min and submitted to Western blot . Using the SWISS-MODEL homology modelling server the initial model was generated [65] . The white-tailed deer PrP 112–233 sequence was uploaded and a template search was performed based on the SWISS-MODEL template library ( SMTL ) , a template search was performed . A total of 201 templates were produced by the homology modelling search . Out of these , the crystal structure of deer prion protein ( PDB: 4YXH [66] ) model 1 , chain A was chosen as the best template with a sequence identity of 99 . 01% , GMQE ( Global Model Quality Estimation ) score of 0 . 78 and a QMEAN4 score of -0 . 10 . Finally , the SWISS-MODEL homology modelling server was used to build the 3D model [65] . Since the template structure and therefore the subsequent model lacked the N-terminal sequence “MKHVGGAAAAGAVVGG” , including the site of the 116G polymorphism ( in bold ) , and two C-terminal residues “GA” , the Accelrys VS [67] software was used to add these amino acids to the model . Initial minimizations and equilibrations were run to relax the structure and allow the added peptide chain to fold naturally at 310 K for 10 ns . The complete construct 112–233 of wild type WTD prion generated by Accelrys VS [67] and equilibrated with MD was used as the control model . Finally , using the SWISS-MODEL homology modelling server , the model for the 116G polymorphism was generated based on the PDB file of the wt PrP structure ( PDB: 4YXH [66] ) . The wt and 116G polymorphism systems were subjected to minimizations , equilibrations and production MD simulations with the Gromacs v 4 . 6 . 3 package [68] and the OPLS ( Optimized Potential for Liquid Simulation ) forcefield [69] . To minimize each system’s energy , the steepest decent method was used . Then a single point charge extended ( SPC/E ) rectangular periodic water box was used to solvate the models . In order to keep the system neutral Cl− or Na+ ions were added . Next , energy minimizations with decreasing restraints on non-hydrogen protein atoms ( Kposre = 1×105 , 1×104 , 1000 , 100 , 10 and 0 kJ mol−1 nm−2 ) were made to relax first the solvent , then the protein . Heating to 310K ( Berendsen thermostat ) and isotropic temperature coupling ( NVT ) equilibration to adjust solvent density to 1g/cm3 followed the minimizations . Lastly , equilibration steps and production runs were executed with NPT ensembles at 310K temperature and 1 atm pressure . The short-range electrostatic and van der Waals interactions cut-off radii were equal to 14 Å each . Long-range electrostatic interactions were treated with the particle-mesh Ewald method ( PME ) , and all bond lengths were constrained with the LINCS algorithm with a fourth order of expansion . The production simulations were performed for 50 ns for each system . For both wt and 116G , two additional simulations using the same starting positions , but with different starting velocities of the atoms were run . To analyze the MD trajectories , radius of gyration , root-mean-square deviations ( RMSDs ) and root-mean square fluctuations ( RMSFs ) were calculated for the Cα atoms using the g_gyrate , g_rmsd , and g_rmsf programs of GROMACS [68] . The solvent accessible surface areas and the hydrophobic surface areas were also calculated using the scripts implemented in GROMACS [68] . Assignments of the secondary structure content and the snapshots of trajectories or graphical representation of the models were done in VMD [70] . Genomic DNA was extracted from brain homogenates using a commercial kit ( DNeasy Blood and Tissue Kit ( Qiagene ) according to the manufacturer’s instructions . Primers were designed using software “Primer3Plus” ( http://www . bioinformatics . nl/cgi-bin/primer3plus/primer3plus . cgi/ ) to amplify the sequence encoding mature cervid PrP ( residues 24–234 ) with high specificity by PCR . The forward and reverse primers recognize the N-terminal ( 5’CCTAGTTCTCTTTGTGGCCATGTG3’ ) and C-terminal signal sequences ( 5’TGAGGAAAGAGATGAGGAGGATCAC3’ ) of PrP , respectively . The primers were synthesized at the University of Calgary CoreDNA service . The PrP sequence of WTD isolates from CWD-infected animals revealed that one animal harboured the wild type PrP genotype ( A ) , wt , and the second one was heterozygous at amino acid 116 encoding both A ( wt allele ) and G , 116AG ( S1 Fig ) . The sequencing results were confirmed by cloning of the PCR products . Ten percent brain homogenate prepared in PBS from WTD or mWTD isolates were treated with 50μg/ml of PK for one hour at 37°C and the reaction was terminated by adding 1X pefabloc proteinase inhibitor . The samples were then serially diluted and run analysed by Western blot using mAb 4H11 ( S3 and S6 Figs ) .
|
Chronic wasting disease ( CWD ) is a prion disease which affects wild and captive cervids . Prion diseases are infectious neurodegenerative disorders , and the causative agent consists of abnormally folded prion protein termed PrPSc . Prions replicate without genetic information , and their three-dimensional structure is thought to encode heritable information necessary to propagate using the cellular prion protein PrPC as a substrate for conversion . In this study , we use in vitro and in vivo techniques to analyze the effect of a polymorphism at codon 116 ( A>G ) of the white-tailed deer prion protein on CWD prion conformation , propagation and pathogenesis . We observed differences in conformation , infectivity and seeding activity in vitro between CWD prions isolated from white-tailed deer encoding wild-type ( 116AA ) PrPC or 116AG-PrPC . In mouse bioassays conformational differences are retained , however , 116AG CWD prions resulted in significantly shortened incubation times upon passages . Molecular dynamics simulations suggest that the structure of 116G-PrPC is more flexible , which is supported by an improved convertibility in an in vitro conversion assay . Altogether these data indicate the importance of a variation in the most conserved PrP domain , and highlight the relationship between PrPC structural flexibility , prion conformation and conversion , and pathogenesis of prion disease in vivo .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"chronic",
"wasting",
"disease",
"animal",
"diseases",
"medicine",
"and",
"health",
"sciences",
"deer",
"molecular",
"dynamics",
"ruminants",
"engineering",
"and",
"technology",
"vertebrates",
"animals",
"mammals",
"animal",
"models",
"animal",
"prion",
"diseases",
"model",
"organisms",
"experimental",
"organism",
"systems",
"optical",
"equipment",
"zoology",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"zoonoses",
"chemistry",
"mouse",
"models",
"prisms",
"biochemistry",
"biochemical",
"simulations",
"computational",
"chemistry",
"equipment",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"biology",
"amniotes",
"organisms",
"prion",
"diseases"
] |
2017
|
Destabilizing polymorphism in cervid prion protein hydrophobic core determines prion conformation and conversion efficiency
|
Microsporidia ( Fungi ) have been repeatedly identified as the cause of opportunistic infections predominantly in immunodeficient individuals such as AIDS patients . However , the global epidemiology of human microsporidiosis is poorly understood and the ability of microsporidia to survive and multiply in immunocompetent hosts remains unsolved . To determine the presence of latent microsporidia infections in apparently healthy humans in the Czech Republic , the authors tested sera , urine and stool originating from fifteen persons within a three month period examined on a weekly basis . Sera , stool and urine samples originating from fifteen HIV-negative people at risk with occupational exposure to animals , aged 22–56 years , living in the Czech Republic were tested by indirect immunofluorescence assay ( IFA ) for the presence of specific anti-microsporidial antibodies , standard Calcofluor M2R staining for the detection of microsporidian spores in all urine sediments and stool smears and molecular methods for the microsporidial species determination . Specific anti-microsporidial antibodies were detected in fourteen individuals , asymptomatic Encephalitozoon spp . infection was found in thirteen and E . bieneusi infection was detected in seven of those examined . While E . hellem 1A and E . cuniculi II were the major causative agents identified , seven different genotypes of E . bieneusi were recorded . These findings clearly show that exposure to microsporidia is common and chronic microsporidiosis is not linked to any clinical manifestation in healthy population . Moreover , our results indicate much higher incidence of microsporidial infections among an apparently healthy population than previously reported . These results open the question about the potential risk of reactivation of latent microsporidiosis in cases of immunosupression causing life-threatening disease .
Microsporidia have emerged as causative agents of opportunistic infections in AIDS patients and other immunodeficient individuals . Several species of microsporidia can cause disease in humans . Intestinal microsporidiosis due to Encephalitozoon ( Septata ) intestinalis and Enterocytozoon bieneusi are most frequently reported among immunocompromised people including patients with acquired immune deficiency syndrome ( AIDS ) [1] , [2] and other immunocompromised patients such as transplant recipients [3]–[6] . Encephalitozoon cuniculi and E . hellem are less prevalent among immunodeficient patients [7] , [8] . Infections with microsporidia in immunocompetent individuals such as travelers have also been described [9] , [10] . Although the most common clinical symptoms related to encephalitozoonosis among immunodeficient patients are chronic diarrhea and malabsorption , they can also cause systemic diseases . While immunocompetent persons often have mild or self-limiting disease , AIDS patients can experience weight loss and increased mortality [11] . Since the studies examining the prevalence of microsporidiosis have been limited to patients who are infected with human immunodeficiency virus ( HIV ) or who have AIDS , the global epidemiology of human microsporidiosis is poorly understood . Variation of spore shedding intensity of microsporidia was shown in both human and animals [12]–[16] . However , to our knowledge there have been no reports on the spore shedding pattern of microsporidia in immunocompetent humans . Therefore we aimed to study the pattern of microsporidia spore shedding in a cohort of asymptomatic apparently healthy people .
The study was approved by the Hospital České Budějovice ethics committee ( protocol no . 202/07 ) . Written informed consent was obtained from every person prior to examination . Between September and December 2007 , a total of 180 individual stool and 180 urine samples were collected on the weekly basis for 3 months from fifteen HIV-negative people at risk of occupational exposure to various animals , such as farm ruminants , pigs , poultry and rodents . The male to female ratio was 8 ( 53% ) to 7 ( 47% ) with mean age of 35±11 years and range between 22–56 years . The samples were stored at 4°C in the dark without any conservation and examined immediately . Every specimen in the study was supplemented with data on the person's clinical symptoms ( e . g . , indigestion , abdominal pain ) . Prior to the study , serum samples were obtained from all individuals included and the presence of specific anti-microsporidial immunoglobulin G was tested by indirect immunofluorescence assay ( IFA ) . IFA was performed with purified whole spores of E . hellem , E . cuniculi or E . intestinalis grown in vitro in VERO E6 cells and semi-purified spores of E . bieneusi at the concentration 105 spores/well ( spores kindly provided by Dr . G . S . Visvesvara , CDC Atlanta , GA , USA ) . Sera were serially diluted ( 1∶8 , 1∶16 , 1∶32 , 1∶64 , 1∶128 and 1∶256 ) in PBS and compared with negative and positive control sera . Sera with positive fluorescence at titers greater than 128 were considered positive . Standard Calcofluor M2R staining [17] was used for the detection of microsporidian spores in all urine sediments and stool smears . Stained slides were examined by fluorescence microscopy using UV light with a wavelength of 490 nm and at a magnification of 1000× . Positive control slides were used for each examination . The DNA was isolated from the stool and urine samples using homogenization by bead disruption using FastPrep–24 Instrument ( MP Biomedicals , CA , USA ) and DNA was extracted using commercially available isolation kit QIAamp DNA Stool Mini Kit ( QIAGEN , Hilden , Germany ) according to the manufacturer's instructions . Acquired DNA was stored at −20°C . The nested PCR protocol by Katzwinkel-Wladarsch et al [18] amplifying the ITS region of Encephalitozoon spp . and Enterocytozoon bieneusi using microsporidia-specific primers was performed as described elsewhere [16] . As positive controls the following were used: DNA obtained from spores of E . intestinalis originally isolated from AIDS patients [19] and grown in vitro in VERO E6 cells in the Laboratory of Veterinary and Medical Protistology at the Institute of Parasitology ASCR , and DNA from spores of E . bieneusi of genotype D originally isolated from a pig [16] . PCR products were visualized on a 2% agarose gel containing 0 . 2 µg/ml ethidium bromide and directly sequenced on the ABI3730XL sequence analyzer ( Applied Biosystems , Foster City , CA ) . Sequences were aligned and completed using programs ChromasPro ( Technelysium , Pty , Ltd . ) BioEdit and Clustal X 2 . 0 . 6 and compared with sequences in GenBank .
Specific anti-microsporidial antibodies were detected in fourteen out of fifteen tested people; 87% of sera reacted with E . cuniculi , 47% with E . hellem , 13% with E . bieneusi and none with E . intestinalis ( Table 1 ) . None of the individuals demonstrated any clinical symptoms ( loose stool , indigestion , etc . ) . While no positive finding was revealed among samples of urine using microscopy , four stool samples originating from 3 persons were positive for microsporidia spores , which were subsequently molecularly characterized as E . bieneusi . During the twelve week long observation of spore excretion , microsporidia were molecularly detected in 34 urine samples ( 19% ) and 39 stool samples ( 22% ) originating from all fifteen tested people . Each of the person excreted microsporidial spores intermittently in irregular intervals ( Figure 1 ) . The concurrent infection with two species of Encephalitozoon , E . cuniculi and E . hellem , and Enterocytozoon bieneusi was detected in 7 individuals , co-infection with E . cuniculi and E . hellem in three cases and monoinfections with E . bieneusi , E . hellem or E . cuniculi in one or three individuals , respectively . No E . intestinalis infection was detected . Whereas both Encephalitozoon spp . infections were more often found in urine , E . bieneusi was detected equally in urine and stool samples ( Figure 1 ) . While E . hellem or E . cuniculi infection were caused mainly by predominant genotype 1A or II , seven different genotypes of E . bieneusi including novel genotypes CZ4–CZ6 were identified in both urine and stool samples . All other ITS sequences from the study samples were a 100% match to the reference genotypes from GenBank listed in Table 2 .
The actual extent of microsporidian infections is unknown . Microsporidia were , and still are , often overlooked and misdiagnosed because they are not specifically searched for in most diagnostic labs , they are rather small , and their staining with hematoxylin and eosin is not sufficient . Most of what is now known about human microsporidiosis can be attributed to the experience with patients infected with HIV [1] , [2] . However , with increased awareness and improved diagnostics , microsporidia have become more frequently reported also in immunocompetent individuals , producing asymptomatic infections [20]–[22] . Despite limited sample number our findings showed a well-supported correlation between spore presence in excretions and seropositivity , which discriminates the actual latent microsporidiosis from simple consumption and passage of spores through the intestinal tract . Intermittent spore shedding for a long period has been experimentally demonstrated for several hosts including rabbits with E . cuniculi [12] , wild-type mice with E . intestinalis [13] , pigs with E . bieneusi [15] , budgerigars with naturally acquired Encephalitozoon spp . infection [23] and HIV-positive patients with E . bieneusi [14] . The persistence of microsporidia despite resolution of the intestinal disorder suggests that microsporidia infection may cause clinical symptoms ( e . g . , diarrhea ) during the early stages of infection that could be overlooked and resolved even though the microsporidia persist . Our survey was performed on a limited sample size from a highly selected population , which could result in decreased statistical power . On the basis of present results it is obvious , that prevalence data of microsporidial infection reported by various authors reaching up to 38% the case of Encephalitozoon spp . and 51% for E . bieneusi , could be hampered by collection of only a single sample for diagnosis , especially in low level infections . While the twelve week sampling enabled us to detect E . cuniculi in 86% of tested people , E . hellem in 66% and E . bieneusi in 47% , the hypothetical individual single sampling performed at any day would identify E . cuniculi in only 0–27% of persons , E . hellem in 0–13% , and E . bieneusi in 0–13% . Based on data in the literature and our experience , it seems that the incidence of microsporidial infections is much higher than previously reported and microsporidia may represent neglected etiological agent of more common diseases . However , it is not known how extensive such silent infections are in asymptomatic carriers , including both humans and animals , which have been reported increasingly to harbour various species and genotypes of microsporidia [16] , [24] , [25] . Moreover , the fact that microsporidia DNA were detected in urine sediments suggests , that microsporidia are able to disseminate also in immunocompetent hosts despite previously reported protective T-cell mediated adaptive immunity together with several components of innate immunity [26] , [27] . Furthermore , the majority of prevalence studies currently rely on detection of spores in stool samples only . The results of this study clearly showed that infected seropositive person could excrete detectable amount of microsporidial DNA via urine , nevertheless examination of stool sample will be negative . Detection of specific antibodies seems to be more sensitive than one-shot detection of spores and can provide more accurate information about ongoing microsporidia infection . In conclusion , studies focusing on the epidemiology of microsporidiosis will more clearly define the environmental sources of microsporidia that pose a risk for transmission so that preventative strategies can be implemented . Since no data exist about latent infection in immunocompetent carriers , possible infection reactivation in these individuals and person to person transmission risk via organ donation , such epidemiological data must be compared with experiments that could solve this question definitively . Moreover , using detection methods with a high sensitivity , such as PCR , and consecutive sampling from every individual is recommended to provide more precise epidemiological data .
|
Microsporidia are a group of obligate intracellular parasitic fungi that have risen over the past two decades from obscure organisms to well recognized human pathogens . Out of 14 species reported to infect humans and causing more severe symptoms in immunocompromised individuals , microsporidia of the species Encephalitozoon and Enterocytozoon bieneusi are the most frequent causes of life-threatening chronic diarrhea and systemic disease in HIV patients and acute , self-limited diarrhea in immunocompetent persons . Although the diagnosis and clinical management of microsporidiosis cases have improved significantly recently , the epidemiology of human microsporidiosis is still unclear . To identify the occurrence of latent microsporidia infections in apparently healthy people , the authors tested sera , urine and stool originating from fifteen persons within a three month period . They found specific antibodies against microsporidia in sera originating from fourteen individuals , and using molecular tools , they detected microsporidial infection intermittently in all tested people . The presence of detectable amounts of microsporidial spores demonstrated that exposure to microsporidia is more common than previously believed and microsporidiosis is not linked to any clinical manifestations in healthy people . This finding should make the clinician more aware of the risk of this unapparent infection , its potential reactivation after immunosupression and consequences leading to life-threatening disease .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"biology",
"veterinary",
"science"
] |
2011
|
Latent Microsporidial Infection in Immunocompetent Individuals – A Longitudinal Study
|
While the static magnitude of thermal pain perception has been shown to follow a power-law function of the temperature , its dynamical features have been largely overlooked . Due to the slow temporal experience of pain , multiple studies now show that the time evolution of its magnitude can be captured with continuous online ratings . Here we use such ratings to model quantitatively the temporal dynamics of thermal pain perception . We show that a differential equation captures the details of the temporal evolution in pain ratings in individual subjects for different stimulus pattern complexities , and also demonstrates strong predictive power to infer pain ratings , including readouts based only on brain functional images .
Any scientific or philosophical examination of human perception invariably must take into consideration the long-lasting notion of the subjectivity of pain . Plato , Aristotle , Galen , and Darwin excluded pain from other sensory modalities and instead classified it with emotions . Avicenna ( or Ibn Sina ) , the 11th century Arab-Persian philosopher-physician , is credited to be the first to suggest pain as a specific skin sense; this idea was later reformulated by Descartes , who conceptualized pain signaling from the skin to the brain [1] , [2] . The notion of subjectivity and thus incommunicability of personal pain was seminal in Wittgenstein's abandonment of logic and shifting the emphasis of 20th century philosophical inquiry towards the study of language , in order to understand how such a private experience can be communicated at all [3] . More recently , D . Dennett has argued , based on modern neuro-scientific understanding that due to its subjective nature , and in contrast to visual perception , pain cannot be captured in computational models [4] . Indeed , the official definition of pain as accepted by the International Association for the Study of Pain states that pain is “an unpleasant sensory and emotional experience” , and expands to assert that , “pain is always subjective” [5] . In contrast , psychophysics from its inception in the 19th century has attempted to demonstrate that at least parts of human experience/perception can be captured quantitatively and described with simple models . Beginning with the work of E . H . Weber and culminating with S . S . Stevens's law of magnitude perception , statistical properties of pain have been quantified and modeled using simple equations [6]–[8] . Currently , statistics of pain are most commonly quantified with questionnaire-based tools , and these remain the main instruments with which efficacy of pain therapies are studied in clinical trials , for example [9] , [10] . Temporal profiles of pain perception , however , have been seldom studied [11]–[13] . Yet , with the advent of human brain imaging technology the need for tracking pain perception in time prompted a number of groups to study pain perception as a time-evolving phenomenon [14]–[17] . A result that has surprised the pain research community is the presence of strong temporal non-linearities in the relationship between the stimulus pattern and the corresponding ratings , including illusory perception of heat and warmth [16] which do not appear to fit any cogent framework and yet can be linked to brain activity [18] , [19] . With this as a starting point , we treat here time evolution of acute thermal pain perception as a dynamical system described by differential equations , the properties of which provide a general summary of the transformation of thermal heat parameters to pain perception space . Surprisingly , simple and interpretable first- and second-order differential equations with very few parameters accurately model time variability of pain perception in humans elicited by thermal stimulation patterns of varying complexity . The equations can be used to infer with high accuracy the response of individuals in modeling conditions that include access to the stimulus temperature and in ‘mind reading’ setups , i . e . when pain perception is solely inferred from functional images of the brain aided by the derived equations .
Given that perception of pain is a slow event and can be rated continuously , online continuous ratings of thermal pain can be readily generated [14]–[17] . When the stimulus intensity on the skin is monitored together with the resultant ratings of pain , one can view this as a system identification problem where the input and output are continuous time varying variables . We reason that behavioral and evolutionary constraints require thermal pain to display three basic features . First and foremost , it must signal the threat of tissue damage: this is obviously determined by the current value of the skin temperature . The signal magnitude must monotonically increase with the temperature , although not necessarily linearly ( as in fact , tissue damage is not linear with temperature ) . Following standard psychophysical practice , we consider the perceived magnitude of pain to be a positive quantity , i . e . we exclude the possibility of negative pain . Secondly , this magnitude must anticipate the possibility of damage , sounding the alarm of an imminent threat given the recent history of temperature values , independently of the current temperature . This information can be partially captured by the rate of change of the skin temperature . Finally , given its powerful hold on behavior , the intensity of pain perception must quickly decay once the threat of damage disappears , so as not to interfere with ongoing mental states [20] , [21] . Following these basic principles , we model pain perception as a dynamical system using a second-order differential equation: ( 1 ) Here is the instantaneous perception of pain at time , is the temperature , is the pain acceleration , and are , respectively , the pain's and temperature's rates of change . We explicitly constrain the dynamics to maintain the non-negativity of perception , , by imposing the boundary condition . The quantities are subject-specific constants . The first term in the right-hand-side represents the temperature-dependent “force” , whose functional form we model , for the sake of parsimony , with a step function ( Figure 1 inset ) : , that is , the acceleration of the perception of pain takes effect only after the threshold is exceeded . The second term is the decay of pain or “forgetting” , which helps perception return to its minimal value upon the removal of the injury threat presented by , and also dampens the oscillations that naturally arise in a second-order dynamical system . The constant has units of 1/time , and therefore can be considered the time scale of the forgetting process . The third and last term is less intuitive , but equally meaningful from a functional perspective . It can be thought of as a dynamic restoring force , similar to the elastic term in the equation that describes a mechanical oscillator . When the derivative of the temperature is small enough , the term is negative and has the effect of limiting the pain level upon the continuing presence of a supra-threshold stimulus , as well as eliminating any sub-threshold pain fluctuations . When the temperature changes quickly , however , the effect of this term is more interesting . In the event of a temperature increase , the term becomes a driving force that helps accelerate the perception of pain , to build up an alerting signal that anticipates the upcoming threat of the temperature reaching and surpassing the injury threshold . Similarly , when the temperature drops fast , the term becomes a restoring force , pushing pain perception faster than the decay term and the passive restoring force would allow . Notice that this creates an asymmetry in the rise and fall time-constants , even when the rate of temperature change is the same in absolute terms: if the temperature drops when the pain perception is high , the restoration is much faster than the rise , for a similar rate of change of the temperature . The constant determines the intensity of the restoring/driving force , while can be considered as a threshold above which fast changes in temperature become alarming . The different effects of the three terms are illustrated in Figure 1 , which depicts the evolution of pain perception averaged across subjects ( blue trace ) upon the presentation of an evolving temperature stimulus ( dashed trace ) ( figure 7 , from [16]; corresponding to our complex stimulus ) and the best-fit inferred model ( red trace ) . The temperature forcing term provides the basic effect of quickly increasing the magnitude of pain perception ( first arrow on the left ) . An equilibrium intensity is reached by the combined limiting effects of the restoring force and the decay term ( second arrow ) . The active form of restoring force ( i . e . when ) is most evident in the effect of the small kinks in temperature ( third and fourth arrows ) . In order to understand to what extent the complexity of the second-order dynamical system of Eq . 1 is warranted and the fit to the psychophysical pain ratings significant , we considered two null hypotheses and a model simplification to contrast our results . In the first place , we reasoned that the simplest approach for the nervous system to report thermal pain is by a direct correlation with the temperature , i . e . . This null hypothesis is , in fact , too simple: the linear proportionality implies that temperatures a few degrees below the skin injury threshold will be reported only with proportionally weaker intensity than those a few degrees above the threshold . Alternatively , we considered a model in which perception is linearly proportional to the temperature , but only once it has exceeded a subject-dependent threshold . For obvious reasons , we termed these two null hypotheses as the linear and threshold-linear models , respectively; in the latter case , the temperature threshold is estimated by optimizing the correlation between model and data . The linear null hypothesis has several disadvantages; most glaring among them is the fact that it reports sub-threshold temperatures , which do not necessarily pose a threat of injury , almost as intensely as those that do pose a threat . Similarly , the threshold-linear model is impervious to events that fall below threshold but may signal an imminent threat , such as a sudden increase in temperature . To further probe the significance of our model , therefore , we considered a simpler first-order system derived from Eq . 1 , assuming that the following conditions are satisfied: ( a ) the decay constant is sufficiently large , ( i . e . the time scale is short ) , and ( b ) the effect of the rate of change of the temperature is not significant , . Simple algebra leads then to the following first-order differential equation: ( 2 ) Where and are subject-specific constants . The functional form of this equation is similar to that of a leaky capacitor , with the forcing affecting now the rate of change of perception ( as opposed to the acceleration ) , and a restoring force that determines a unique time-constant for both rising and falling of perception . To test the relative merits of these models we performed psychophysical experiments , and contrasted model predictions . We designed two stimulation types: a simple stimulus in which the temperature ranges between a sub-threshold value and a supra-threshold value that is maintained constant during blocks [14] , and a complex stimulus in which the blocks of supra-threshold temperature are interspersed with shorter blocks of higher temperature values [16] ( see Methods for details ) . Figure 2A–F depicts an example of fitting a single subject's rating of a simple and a complex stimulus . Simple ( panel E ) and complex stimuli ( panel F ) are modeled using the first-order ( panels A and B , respectively ) and second-order models ( panels C and D ) . Observe that while for the simple stimulus the two models appear to fit similarly well , the complex stimulus highlights the ability of the second-order model to capture the subtleties of the rating . Similar results were seen in all subjects studied ( Figure S1 ) . The results of fitting the second-order model to the perceptual data for all participants are summarized in Figure 2G , showing the fit correlation for the second-order model contrasted with the null hypotheses . The increase of model performance over the null hypotheses is quite significant , reaching in some cases nearly 0 . 4 , while the mean model correlation is above 0 . 9 ( Wilcoxon matched-pairs signed-ranks test , Wp , ) . Similarly , the comparison with the first-order model ( Figure 2H ) shows that in all but two cases the second-order model is a better fit to the actual pain ratings ( Wp , ) . This increase in accuracy , however , may be explained by the model's larger number of parameters ( 5 ) compared with those for the simpler first-order model ( 3 ) , and the two null hypotheses ( 1 for linear-threshold , none for linear ) . To account for this , we computed the difference in the Akaike Information Criterion ( AIC ) between the model and the null hypotheses . AIC regularizes the goodness of fit with a penalty for the number of free parameters in the model; Figures 2I–J show the gain in AIC for the model over the null hypotheses , and the first-order model , respectively , suggesting that overfitting can be ruled out ( see Methods ) . To further assess our approach , we also compared the correlation between the derivatives of the rating and of the model ( Figure S2 ) , and again we observe that the second-order model outperforms the null hypotheses models ( Wp , ) but not the first-order model ( Wp , ) . We also considered the robustness and generalization capability of the modeling approach with respect to other sources of variability in the perceptual response . For that , we resorted to the concept of predictive modeling , a statistical learning approach that has gained increased acceptance in neuroscientific data analysis [22]: the parameters of a model are learned using training data , and then the goodness-of-fit evaluated on previously unutilized test data , as a means to estimate the model's generalization ability . We therefore computed the model parameters for each subject in the first run of the experiment , and estimated the response for the second , independent run using the same parameters . The results show that test and train correlations are still very similar ( Figures S6 , S7A ) . To understand the population effect of the stimulation paradigm and the modeling , we also fitted an average model of all the subjects , and then tested generalization efficacy of this model ( Figure S7B ) . While the simple stimulus condition is not significantly affected , the complex stimulus shows a large decrement in the generalization ability of the model , indicating that responses to higher temporal structure are dependent on individual sensitivity parameters . A more rigorous test of generalization , however , involves predicting one class of stimuli in one run ( i . e . complex in run 2 ) with parameters fitted to the other class and the other run ( i . e . simple in run 1 ) . Prediction of complex stimuli with parameters fitted to simple stimuli yields a group average of r = 0 . 68 , over r = 0 . 93 for the estimate . Prediction of simple for parameters fitted to complex yields r = 0 . 84 , very similar to the average of r = 0 . 89 for the estimate ( see Figure S7D–E ) . The higher efficacy of the latter setup is consistent with the idea that the more complex stimuli can reveal the full dynamical structure of the responses , and therefore be more robust to generalization . One of the practical applications of predictive modeling in neuroscience is its use in “mind reading” setups , i . e . the possibility of obtaining precise information about perceptual and cognitive states , such as words or images presented to subjects in the fMRI scanner , by applying a predictive model to fMRI data [22] . The ability to predict and reconstruct with high accuracy external stimuli under certain conditions has proved to have enormous implications for basic research and brain-machine applications [23]–[25]; however , predictive modeling of clinically relevant measures has shown to be more elusive . To further demonstrate the relevance of our findings , we analyzed the impact of including the analytic model in a predictive setup , as follows: ( a ) we trained a predictive linear model with regularizing constraints , the Elastic Net [26] , [27] , to infer pain ratings from full-brain fMRI traces , utilizing TR volumes ( i . e . the brain images acquired at each time point ) concurrent with the ratings as independent samples ( hereby labeled EN model ) ; ( b ) we trained a model as in ( a ) , but using up to 7 TR volumes previous to the time the ratings are reported , and using as predictors only voxels that have a time-lagged correlation with the target variable above a threshold ( 0 . 2 in this case ) ( EN w/lags model ) ; ( c ) we trained a model as in ( a ) and combined it linearly with the analytic second order model , Eq . 1 , trained on the same data using both temperature and pain ratings ( Combined model ) . Specifically , the model is trained to infer the pain ratings from fMRI traces , independently infer the temperature from fMRI traces , obtain a second estimate of the pain ratings through the application of the dynamical model to the inferred temperature , and then combine both predictions into one . Finally , ( d ) we trained an unconstrained , linear ordinary least-squares model ( OLS ) , with the same conditions as in ( a ) ( Figure 3A ) . With this setup , we then computed the predictive accuracy of the combined model to infer the pain ratings on unseen test data , using only the fMRI traces , and compared it with the predictions of the EN model , the EN w/lags model ( to compensate for the intrinsic use of the recent history in the analytic model ) , and the OLS model ( Figure 3B ) . The results are shown in Figure 3C–D , which displays for each subject the predictive accuracy of the EN , EN w/lags and OLS models in comparison to the Combined model . The Combined model shows a significant improvement in predictive accuracy over the other three models , including EN w/lags , which includes delayed information and helps it to predict better than EN . In all cases , the increase in accuracy is statistically significant ( Wp , ) . These results demonstrate that our dynamical model can be successfully combined with physiological measurements in order to obtain further insights into the mechanisms of pain perception , and eventually used as a scaffold for experimental manipulations . Moreover , given the high accuracy of the predictions , we conclude that “mind reading” of subjective pain perception is practically attainable . Besides the model's predictive efficacy , it is important to understand how consistent it is with respect to the known phenomenology . In particular , the distribution of threshold temperatures over the population ( Figures S1 , S4 , S5 ) closely matches classic values determined by rigorous psychophysical methods [28] . The other easily interpretable parameter of the model , the decay time-constant , also shows a reasonable distribution of values , as well as a good match between the second-order and the simplified first-order models ( Figures S1 , S4 , S5 ) . In order to assess the significance of each of the terms contributing to the description of the perceptual dynamics in Eq . 1 and Eq . 2 , we computed all pair-wise correlations between the corresponding fitted parameters in the second-order model . High correlation between two terms may indicate a redundancy in model , or perhaps an even worse inadequacy of the model to capture the essential features of the dynamics . Of all pairs ( Table S1 ) , only two reach statistical significance: between and ( r = 0 . 56 , p = 0 . 01 ) , and between and ( r = −0 . 53 , p = 0 . 017 ) . It is instructive to contrast these values with the result of performing a similar computation with the fitted parameters for the first-order model; in this case , the correlation between and is significant ( r = 0 . 72 , p = 0 . 0003 ) . A parsimonious interpretation of these results is that the simplification of the dynamics introduces correlations between terms that do not properly describe it . Given that the second-order model performs better , we conclude that the more complex model is also a better representation of the dynamics . Moreover , while the two correlations are significant , their actual value ( r0 . 5 ) implies that their contributions are not redundant . We tested more radical variants of the modeling approach , in order to test its goodness-of-fit in a “functional space” . In particular , Eq . 2 was expanded to incorporate two time-constants , slow and fast systems corresponding to the physiology of slow ( unmyelinated ) and fast conducting ( myelinated ) nociceptive afferents [29]; we determined that such models do not substantially improve prediction of pain ratings ( Figure S9 ) . In fact , the apparent presence of two time-constants in the perceptual dynamics is accounted for , in Eq . 1 , by the term , which models the decay of perception after the temperature drops below threshold as faster than the rising time-constant ( because is higher in the former than in the latter , see Figure 1 ) . A large psychophysical body of literature shows that static ratings of thermal pain , similarly to other sensory modalities , follow S . S . Stevens's power-law for perceived magnitudes [8] , suggesting that the dependence of dynamics of pain perception on temperature might be better modeled by a power function . As this law describes the stationary or steady-state response to pain , as opposed to its dynamical behavior , we cannot directly compare it against our model . However , we considered that it would be possible to extend the model to encompass power-law stationary responses . Given that this requires an additional parameter ( the exponent ) , it is more reasonable to consider an extension of Eq . 2 , in which the term driven by the difference between the current temperature and the threshold is modified by an exponent , leading to: ( 3 ) where and is an additional parameter . Performance of this new model was contrasted to Eq . 2 , yielding results that are comparable but slightly poorer , even though the model has one more parameter . To summarize , the mean correlation over simple and complex stimuli was 0 . 90 and 0 . 87 , compared to 0 . 92 and 0 . 88 for Eq . 2 . We also observe that as long as and are fitted for individual ratings , proportionality constant and the power parameter compensate for each other ( range for was 2 . 97 to −0 . 28 , mean = 1 . 0 and SEM = 0 . 3 ) , and and converge to the same optimal values as found for Eq . 2 ( performance measure between Eq . 2 and Eq . 3 using either r or SSE shows a correlation of 0 . 99 , p = 0 ) . Our model can capture , in a single framework , perceptual behaviors that are usually considered as disparate . Given that the perception of pain can be parceled into separate dimensions and as recent evidence suggests that the temporal dynamics of these modalities may have unique properties that depend on stimulus intensity [28] , we examined the properties of our models for the percept of burning . When subjects were instructed to report the magnitude of burning pain [28] , we observed similar rating profiles and model fitting to the perceived magnitude of pain , indicating that the modeling approach may be equally applicable to sub-modalities of pain . Similarly , our model encompasses the different behaviors associated with offset analgesia ( OA ) . While OA is usually defined by the de-sensitization to the same noxious temperature following exposure to a more noxious one [16] ( a feature essentially captured by our model , cfr . Figure 1 ) , other more subtle features have been reported in the literature under the OA characterization , of which we will consider the main two . The first one is the observation that temperature fall rates in the range of 0 . 1 to 0 . 5°C/sec are barely detected with continuous ratings of pain [16] . We tested whether our second order model will also show less sensitivity to stimulus offset rates , in comparison to the first order model , where perception fall rates should better reflect stimulus fall rates . Figure S10 shows that in fact these predictions are correct ( the model closely captures pain ratings as described in figures 3 and 4 in [16] ) . A second observation regarding OA is that pain perception magnitude for increasing intensities shows different patterns when the stimulus has an additional one degree perturbation ( offset stimulus ) in contrast to when the stimulus is kept at a constant level or returns to baseline [30] . Again our second order model captures these features better than the first order ( Figure S11 ) , and in fact our model replicates figures 2–5 in [30] .
Model simulation was implemented with standard integration algorithms in Matlab . To obtain the simplified Eq . 2 from Eq . 1 , we writeAssuming a large decay constant ( equiv . a short time scale to ‘forget’ ) , and that the effect of fast changes in the temperature profile is negligible , , we can drop the l . h . s . term to writeWhere and . Parameter estimates for first order and second order equations were calculated in Matlab using minimization of the least squares error between simulation and experimental data , and a random search technique over the parameter space . For each stimulus rating condition , three parameters were calculated for first order fitting and five parameters for second order fitting . Adequacy of fitting was measured by zero-lag Pearson correlation between model output and pain ratings . Overfitting of the model was investigated using the Akaike Information Criterion ( AIC ) , which penalizes the measure of goodness of fit with a term proportional to the number of free parameters [31] . When the residual squared error sum ( SS ) is known , the criterion can be written aswhere n is the number of samples , and k the number of parameters . is a constant that depends on the particular dataset used , but not on the model , and therefore can be ignored when making comparisons of between models for the same data . As even when is discounted , this measure still depends on the total number of samples , for presentation's sake we computed a normalized version , which we call here the Akaike gain for the model ( m ) with respect to the contrasting null hypotheses and first-order model ( c ) , asA positive value for indicates that the gain in accuracy of the model cannot be explained by the increase in number of parameters . For the first null hypothesis , i . e . perception proportional to temperature , the number of parameters is zero . The second null hypothesis , perception proportional to temperature over a threshold , has one free parameter that we estimate similarly to the analytic models . The Pearson correlation between the parameters for the second-order and first-order models was computed using all fitted parameters across subjects and stimuli ( Table S1 ) . The functional MRI data are the same used in an earlier study [14] . Here the thermal stimulus and related ratings of pain are used to compare results of full-brain machine learning with elastic net for predicting pain perception with and without incorporation of our quantitative model for pain perception , Eq . 1 . Herein , we learn a predictive model individually for each subject . We treat voxels as predictor variables , TRs as independent samples ( following [26] , [27] ) , and pain ratings as target variables , respectively . While the independence assumption among subsequent TRs does not hold in practice , and is used mainly for simplicity sake , it allows us to reach good predictive accuracy . We learn the model parameters using the first half of the experiment as training data , and then apply the model to the second half of the experiments , treated here as test data . Sparse predictive models were learned using a sparse regression method called the Elastic Net [32] , which enhances the basic LASSO regression [33] by combining <$>\raster ( 80% ) ="rg1"<$>1-norm ( sparsity-enforcing ) constraint with the <$>\raster ( 80% ) ="rg1"<$>2-norm ( “grouping” ) constraint . The rationale behind this extension is to overcome a known limitation of the LASSO: given groups of correlated variables ( e . g . , spatial clusters of voxels ) , LASSO may pick an arbitrary one from the group , as long as the resulting model predicts well; however , if the goal is neuro-scientific interpretation of the sparse model as a set of voxels relevant to the task , it is important to include ( or exclude ) voxels as groups ( clusters ) of highly-correlated variables , rather than single representatives of a group . This is achieved , to some extent , by controlling the grouping parameter mentioned above , that tends to enforce similar coefficients among highly correlated voxels ( e . g . , spatial neighbors ) . The Elastic Net and other models used in this paper are formally described below , and summarized in Table 1 .
The results show that acute thermal pain perception applied to healthy skin follows simple quantitative deterministic patterns . The dynamic model is derived from a behaviorally relevant interpretation of pain perception as a warning signal that quickly reports immediate threat of injury ( temperature above threshold ) , and approaching danger ( rapid temperature increases ) , and can also as easily discount the threat once it goes away or it is expected to do so ( temperature decreases ) . The model , using few parameters , can reproduce with high accuracy the dynamical transformation from stimulus to perception . Moreover , the model also has high predictive accuracy , and accounts for subjects' variability with simple and interpretable mechanisms . The model provides a summary of a relatively complex behavior , whose physiological correlates and mechanisms can be directly investigated through pharmacological manipulation and the design of targeted stimulus conditions . Temporal processing is ubiquitous in sensory systems , including the somatosensory pathway [36] , [37] . However , it is only in a few cases that spatio-temporal transformations can be functionally interpreted , beyond generic sharpening for enhanced localization [38] , or information compression [39] . We do not consider , however , that the perceptual dynamics captured by our model can be reduced to peripheral processing . In fact , as previously reported [14] , the BOLD response to a task similar to the one used in this report reveals a rich temporal structure across several cortical and sub-cortical areas compatible with the time scale of the perceptual ratings , such that the dynamics of pain perception may result from the emergent interaction of extensive networks . Moreover , given its ultimately non-linear nature , the model further predicts dynamical features of pain perception that may have unexpected behavioral relevance ( see Text S1 ) . The utilization of our analytic model within the “mind reading” setup highlights its predictive efficacy , and provides an additional validation step . A further reason for using the combined model , besides simply inferring pain from fMRI , is to go beyond the limitation of simple linear inference models such as Elastic Net , while keeping the non-linear model simple , tractable and interpretable . Given the nature of brain processes , we expect the true relationship between the high-dimensional fMRI signal and pain ratings to be a complex non-linear one . However , fitting an ad hoc non-linear model ( e . g . , a neural network ) to such high-dimensional data to predict pain rating directly could be computationally much more challenging than fitting a linear one . On the other hand , given an accurate analytical model linking temperature to pain , we may exploit it advantages in our combined nonlinear method , first obtaining an estimate of the temperature from fMRI data via simple and computationally efficient linear regression , and then using nonlinear model predicting pain from temperature . Though the combined predictive model involves inferring temperature as a hidden variable , it outperforms the direct EN model because it captures ( at least the temperature-to-pain part of ) the non-linear relationship between fMRI and pain perception . To some extent , we can consider the analytic model as a principled constraint in the temporal domain , similar to the spatial regularization imposed by EN . Our model can only provide a limited description of the full complexity of pain perception . In particular , the model accurately captures the perceptual dynamics in the time scale of seconds to minutes , most relevant for the functional interpretation of thermal pain as an “alarm signal” . Processes whose dynamics develop over longer time scales , such as habituation , sensitization , post-tissue injury , or following acute or chronic pain conditions [8] , [11] are beyond the model's descriptive capabilities . For instance , repeated testing of offset analgesia over multiple days in [16] results in sensitization changes , which however do not alter the quality of the responses . Nevertheless , our model can provide an analytic framework even in the context of these long-term adaptive processes , as it will be possible to study the effect of adaptation on the different parameters that control the short-term perceptual dynamics , for instance threshold and decay time-constant . Another class of perceptual behaviors that our model does not consider , unrelated to differences in time scale , are those derived from interactions between pain and cognitive and attention processes , which can significantly modulate the perception to objectively similar noxious stimuli [40]–[41] . Despite its limitations , the model provides a powerful tool with which peripheral and central mechanisms can be studied . As the model describes subjective reports of magnitude of pain , it may also generalize to magnitude perception across other sensory modalities . Moreover , as we have tentatively shown with the combined model of fMRI-based prediction , it should be possible to identify physiological processes associated with the proposed components of the perceptual dynamics , and so reduce the gap between phenomenology and theory .
|
We propose a model of thermal pain perception that accounts for its dynamical behavior , and can be used to predict subjective responses to thermal stimulation on individual subjects with high accuracy , close to 90% averaged over subjects ( over 65% for the null hypothesis ) . The model is based on behavioral considerations that include the need to signal current or approaching tissue damage , and the need to discount past danger . Moreover , we show that in a ‘mind reading’ setting , the combined use of sparse regression to infer pain perception from functional MRI recordings ( fMRI ) , and from the model applied to the stimulus temperature also inferred from fMRI , leads to equally significant predictive accuracy , close to 75% averaged over subjects . Our results demonstrate that a subjective percept such as pain displays a highly deterministic behavior .
|
[
"Abstract",
"Introduction",
"Results",
"Methods",
"Discussion"
] |
[
"cognitive",
"neuroscience",
"fmri",
"computational",
"neuroscience",
"biology",
"computational",
"biology",
"pain",
"neuroscience",
"neuroimaging"
] |
2012
|
Predictive Dynamics of Human Pain Perception
|
Autophagy plays a crucial role in health and disease , regulating central cellular processes such as adaptive stress responses , differentiation , tissue development , and homeostasis . However , the role of autophagy in human physiology is poorly understood , highlighting a need for a model human organ system to assess the efficacy and safety of strategies to therapeutically modulate autophagy . As a complete , cyclically remodelled ( mini- ) organ , the organ culture of human scalp hair follicles ( HFs ) , which , after massive growth ( anagen ) , spontaneously enter into an apoptosis-driven organ involution ( catagen ) process , may provide such a model . Here , we reveal that in anagen , hair matrix keratinocytes ( MKs ) of organ-cultured HFs exhibit an active autophagic flux , as documented by evaluation of endogenous lipidated Light Chain 3B ( LC3B ) and sequestosome 1 ( SQSTM1/p62 ) proteins and the ultrastructural visualization of autophagosomes at all stages of the autophagy process . This autophagic flux is altered during catagen , and genetic inhibition of autophagy promotes catagen development . Conversely , an anti–hair loss product markedly enhances intrafollicular autophagy , leading to anagen prolongation . Collectively , our data reveal a novel role of autophagy in human hair growth . Moreover , we show that organ-cultured scalp HFs are an excellent preclinical research model for exploring the role of autophagy in human tissue physiology and for evaluating the efficacy and tissue toxicity of candidate autophagy-modulatory agents in a living human ( mini- ) organ .
In recent years , autophagy has emerged as a pivotal actor in adaptive responses to stress and starvation [1–3] and in tissue homeostasis [4] , cellular differentiation [5] , and ageing [6 , 7] . Key concepts of autophagy have arisen from molecular genetic experiments in a number of model organisms , including mammals , in vivo and ex vivo [8] . However , the role of autophagy in human organ physiology is as yet incompletely understood due to the lack of human model systems and the difficulty of experimental manipulation . For this , it would be helpful to have an easily tractable , clinically relevant human organ model at our disposal . Moreover , such human models would be useful to assess the efficacy and safety of the ever-increasing number of strategies that are being proposed to therapeutically modulate autophagy to treat various human diseases and to slow tissue ageing [9–12] . On this background , we have turned to a complete , cyclically remodelled human ( mini- ) organ , i . e . , terminal scalp hair follicles ( HFs ) [13] . Human HFs can be easily microdissected from excess tissue removed during plastic or hair transplantation surgery and organ cultured in a well-defined , supplemented , serum-free medium [14] . Under these conditions , growing ( anagen ) HFs continue to produce a pigmented hair shaft and will continue their spontaneous organ remodelling activity for many days ex vivo . The organ culture of human HFs has not only permitted major advances in translational hair research but have also permitted novel insights into human tissue physiology and pathology , spanning diverse fields including metabolism , cellular differentiation , chronobiology , cell cycle control , immunology , ( neuro ) endocrinology , toxicology , and pharmacology [15] . Therefore , the value of HF organ culture as a model for biomedical research extends far beyond its importance for dermatology alone . After years of massive growth activity ( anagen ) , human scalp HFs spontaneously enter into a rapid , apoptosis-driven organ involution process ( catagen ) [16] , following the dictates of an as yet insufficiently understood , organ-intrinsic “hair cycle clock” [17–19] . We hypothesized that late-stage anagen scalp HFs , whose hair matrix epithelium proliferates at a higher rate than most malignant tumors , despite being exposed to a number of stressors , are likely to come under increasing pressure to maintain tissue homeostasis and may require a substantial autophagic flux [20] to maintain their growth . That the HF can recover from massive toxicological insults , such as during chemotherapy-induced alopecia [21] , and that the antimalarial agent chloroquine ( CQ ) , a major autophagy inhibitor used in the clinic , can elicit adverse hair effects , such as change in hair color and hair loss [22] , also encouraged the concept that the HF may engage in autophagy as a fundamental adaptive mechanism against stress . Here , we have tested the hypotheses that organ-cultured human scalp HFs need to maintain a substantial autophagic flux in order to sustain anagen and that these ( mini- ) organs are well suited to study both the role of autophagy in human organ physiology ex vivo and to test candidate agents that modulate autophagy in a therapeutically desired manner under clinically relevant conditions . In the following , we report evidence that confirms both working hypotheses .
The yeast homologous autophagy-related protein 8 ( ATG8 ) , Light Chain 3 ( LC3 ) , and sequestosome 1 ( SQSTM1 , also known as p62 ) are well-documented markers to monitor autophagy by fluorescence microscopy [23] . Lipid conjugated LC3 proteins ( LC3-II ) are specifically recruited on the membrane of autophagosomes from the initial stages of autophagy [24] . Differing from a diffuse cytoplasmic signal of the unconjugated LC3 form ( LC3-I ) , lipidated LC3-II–containing autophagosomes appear as fluorescent dots when assessed by indirect immunofluorescence ( IF ) [23] . However , visualizing the endogenous LC3 protein in compact tissues can be quite challenging [23] . Therefore , as a first step in characterizing the role of autophagy in cycling human HFs , we first established a suitable indirect IF microscopy protocol to detect LC3 in acetone-fixed cryosections of organ-cultured anagen HFs . For this , we used a specific anti-LC3B antibody that has a higher affinity for the lipidated LC3B form ( S1 Fig ) . Confocal microscopy with anti-LC3B antibody/Alexa555 ( red ) demonstrated the presence of cells with distinct perinuclear fluorescent signal ( Fig 1III and 1IV ) , consistent with the recognized cellular localization of autophagosomes [23] . Interestingly , LC3B-positive dots were most prominently seen in keratinocytes of the proximal hair matrix below Auber’s line , the most rapidly proliferating compartment of the HF epithelium [25] , and in the precortical hair matrix ( Fig 1I and 1II ) , i . e . , the epithelial compartment where undifferentiated HF keratinocytes become committed to undergo terminal differentiation into the cells of the inner root sheath , hair shaft cortex , or medulla and begin synthesizing large quantities of specific hair keratins [26] . To further support that punctate fluorescent signals do indeed correspond to LC3B-containing autophagosomes , we conducted confocal microscopy analysis in anagen organ-cultured HFs that were treated for 4 h with CQ ( 10 μM ) with a vehicle control . Because CQ blocks autophagy at its late stage by inhibiting lysosomal function , CQ induces the accumulation of autophagolysosomes enclosing lipidated LC3B [27] . Compared with vehicle-treated samples , confocal images of CQ-treated HFs showed a significant increase in the number of LC3B-positive fluorescent dots ( red , Alexa555 ) , primarily in matrix keratinocytes ( MKs ) ( Fig 2A and 2B ) . Thus , our IF method is suitable to monitor endogenous autophagy in human organ-cultured HFs by indirect fluorescence microscopy . Moreover , the fact that LC3B-fluorescent dots increased upon CQ treatment indicates that the intrafollicular autophagic flux was active in MKs and that autophagolysosomal function was preserved in cultured human HFs [23] . Next , these IF results were independently investigated by transmission electron microscopy ( TEM ) . These ultrastructural analyses supported the presence of an active autophagic flux in the human HF matrix by showing diverse cytoplasmic double-membrane structures belonging to autophagic vacuoles at different stages of autophagosome biogenesis [28] ( Fig 2C and 2D ) . Indeed , we observed a putative phagophore sequestering a portion of the cytoplasm to form an autophagosome ( Fig 2CI ) . We also recognized several initial autophagic vacuoles ( AVi’s ) with visible bilayers separated by a narrower electron-lucent cleft , typical of autophagy ( Fig 2CII , 2CIII and 2CIVa ) . Predominantly , these AVi contained morphologically intact cytosol with ribosomes and organelles ( see as an example the mitochondria in Fig 2CIII ) , which is a common feature of nonselective autophagy [23] . In addition , we observed late/degradative autophagic vacuoles ( AVd’s ) characterized by a partially or completely degraded internal membrane and electron dense cytoplasmic material and/or organelles at various stages of degradation ( Fig 2CIVb ) . We also recognized a putative autophagolysosome , i . e . , degradative autophagic vacuole , that has fused with a lysosome , characterized by lamellar internal membranes ( Fig 2CV ) . Quantification of TEM images showed no accumulation of AVd structures ( Fig 2D ) , supporting the notion that the autophagic flux in the matrix of human anagen HFs ex vivo is actively occurring . Next , we asked whether key autophagy readout parameters change when human anagen scalp HFs spontaneously enter into catagen ex vivo [15] by comparing anagen VI HFs with HFs that showed morphological criteria of early or middle catagen stages , using a battery of previously defined objective classification criteria [29] . Confocal images revealed a marked reduction in the number of LC3B-positive fluorescent dots during the transition from anagen to catagen ( Fig 3A and 3B ) , suggesting a more prominent autophagic flux during the proliferative stage of the HF cycle . To probe this hypothesis , we compared the levels of SQSTM1 in anagen and catagen HFs . SQSTM1 serves as a link between LC3 and ubiquitinated substrates [30] . SQSTM1 and SQSTM1-bound polyubiquitinated proteins become incorporated into the completed autophagosome and are degraded in autolysosomes , thus serving as an index of autophagic degradation [23] . In line with a reduction in autophagy-dependent protein degradation during catagen induction , confocal images showed that SQSTM1 fluorescence signal ( green , Alexa488 ) was significantly higher in the matrix of catagen HFs , compared with the matrix of HFs in anagen VI ( Fig 3C and 3D ) . Further supporting our IF data , quantitative assessment of autophagy-related structures in TEM sections from anagen and catagen HFs showed a significantly higher number of autophagic vacuoles in anagen versus catagen HFs ( Fig 3E ) . These observations raised the possibility that autophagy may serve as a process that maintains and prolongs anagen . If true , manipulating intrafollicular autophagy would be of profound clinical interest , because the vast majority of patients with hair loss or undesired hair growth seen in clinical practice shows a premature shortening of anagen ( leading to effluvium/alopecia ) or retarded entry into catagen ( resulting in hirsutism/hypertrichosis ) [13 , 19] . As mentioned above , treatments with the autophagy inhibitor CQ present recognized deleterious effects on hair viability [22] . Nevertheless , CQ may eventually affect the hair cycle through autophagy-independent cytotoxic effects related to the ability of this drug to inhibit lysosomal function or to intercalate in DNA [31 , 32] . Therefore , to evaluate whether autophagy has a pro-anagen function , we adopted a molecular genetic approach by knocking down the autophagy-related gene 5 ( ATG5 ) with iRNA , a gene that plays a fundamental role in the early stages of autophagosome formation [33] . Anagen HFs from three diverse human individuals were transfected with pool small interfering RNA ( siRNA ) sequences against ATG5 ( siATG5 ) or with nontargeting scrambled oligonucleotides , using our previously described transfection method for gene silencing in human HFs [34 , 35] . Confirming the silencing was successful , 48 h after transfection , the ATG5-fluorescent signal was drastically reduced in siATG5-treated HFs compared with control HFs ( Fig 4A ) . Additionally , the number of LC3B-positive dots was significantly reduced in siATG5 HFs , demonstrating that ATG5 silencing was functionally deleterious to intrafollicular autophagy ( Fig 4B and 4C ) . Further confirming a decrease in autophagic degradation when ATG5 was silenced , SQSTM1 fluorescence levels were elevated in siATG5 HFs , compared with control HFs ( Fig 4D and 4E ) . We then morphologically and immunohistologically assessed the hair cycle stage of each HF 96 h after transfection with siRNA sequences against ATG5 or Control , as previously described [29] . Confirming that ATG5 silencing persisted at this time point , IF analysis showed a significant 80% reduction in ATG5-fluorescence signal of siATG5-treated HFs , compared with control HFs ( Fig 5A and 5B ) . While the majority of control HFs progressed through the anagen-catagen transformation relatively slowly and were mostly in early catagen stage , siATG5-transfected HFs involuted much more rapidly and reached the middle and late stages of catagen development ( Fig 5C and 5D ) with less than 10% of ATG5-silenced HFs having retained their characteristic anagen VI morphology . To validate these morphological analyses , we assessed the hair cycle stage in siATG5 and control HFs by measuring the number of proliferative and apoptotic MKs in the hair matrix tips , below the line that represents the widest part of the hair bulb ( Auber’s line ) [15 , 29] . Compared with control HFs , we observed a significant reduction in the percentage of MKs that were positive for the proliferation marker , Ki-67 ( red , Alexa555 ) , in siATG5-treated HFs , with a significantly higher percentage of apoptotic ( Terminal deoxynucleotidyl transferase dUTP Nick End Labeling [TUNEL]-positive ) cells ( green , Alexa488 ) ( Fig 5E–5G ) . Therefore , experimental autophagy inhibition prematurely terminates anagen and promotes apoptosis-driven development . The above findings imply that , conversely , up-regulating intrafollicular autophagy should prolong anagen . Notably , the principal ingredients ( core mix ) of an anti–hair loss product on the market contains Galeopsis segetum extract , biotin , and N1-methyspermidine , the latter of which is a metabolically stable analog of the well-recognized autophagy-promoting agent , spermidine [36–39] . We thus decided to test whether this core mix was able to induce autophagy and prolong anagen in organ-cultured HFs . Confocal images showed a marked increase of LC3B-positive fluorescent dots in HFs treated with the core mix , compared with vehicle-treated HFs ( Fig 6A and 6B ) . Supporting this , the core mix treatment significantly lowered SQSTM1-fluorescent signal compared with the vehicle , demonstrating that the increase in LC3B-fluorescent autophagosomes depended on an increased autophagic flux , thus enhanced autophagy-mediated degradation ( Fig 6C and 6D ) . Quantification of autophagic structures in TEM sections from HFs treated with core mix or a vehicle further validated an induction of a bona fide autophagic flux upon treatment ( Fig 6E ) . This result also suggests that the N1-methylspermidine retains the ability to induce autophagy , as its desmethylated analog . To validate this concept , we adopted an in vitro cellular assay to demonstrate the pro-autophagic function of spermidine [36] . We thus evaluated the levels of lipidated LC3B and SQSTM1 in human bone osteosarcoma epithelial U2OS cells treated with equimolar doses of N1-methyspermidine and spermidine . Consistent with the results published by Pietracola and coworkers [36] , spermidine treatment increased the levels of the lipidated LC3B-II form and stimulated autophagy-mediated degradation of SQSTM1 ( Fig 6F and 6G ) . In addition , N1-methylspermidine–related effects on both LC3B-II and SQSTM1 levels paralleled the differences observed with the natural spermidine version ( Fig 6F and 6G ) . Further indicating that both compounds can also induce autophagy in keratinocytes , these results were extended and confirmed in the human NCTC 2544 keratinocyte cell line ( S2 Fig ) . Next , we decided to validate that the autophagy-enhancing ability of the core mix corresponded to a pro-anagen effect . We thus assessed the hair cycle stage of anagen HFs from three diverse human donors after a 5-d treatment with the core mix or vehicle . Despite a substantial variability in HF cycling among the diverse donors frequently observed in human HF organ culture experiments in which multiple patients were used [15] , morphological evaluation showed that the treatment with the core mix increased the percentage of anagen HFs from all donors ( Fig 7A ) . Moreover , the autophagy-inducing mix significantly enhanced the relative percentage of Ki-67 proliferative cells while reducing apoptotic MKs in the hair matrix tips , below the Auber’s line ( S3 Fig ) . Notably , the anagen-promoting effects of the core mix was observed even in catagen-primed HFs . Indeed , the treatment of anagen VI HFs from a donor that was already primed to enter catagen , as indicated by the fact that all vehicle-treated HFs had transitioned into catagen at the end of organ culture , preserved a clear anagen VI morphology in 16% of core mix–treated HFs ( Fig 7A , donor 3 ) . To further validate that the observed promotion of anagen was related to an induction of autophagy , we repeated the treatment in HFs silenced for the ATG5 gene . Supporting our previous analysis , siControl HFs treated with the core mix had an increased percentage of anagen HFs compared with vehicle-treated HFs ( Fig 7B , siControl ) . In marked contrast , core mix treatment failed to promote anagen in ATG5-silenced HFs from two independent donors ( Fig 7B , siATG5 #1 and siATG5 #2 ) . Collectively , our results support a scenario in which intrafollicular autophagy plays a fundamental anagen-maintaining role in HFs .
The current study unveils a crucial new role of autophagy in human hair growth control , namely for maintaining the HF growing stage , anagen . Moreover , we show that organ-cultured scalp HFs are an excellent preclinical research model for exploring autophagy functions in human tissue physiology and for evaluating the efficacy and tissue toxicity of candidate autophagy-promoting and -inhibitory agents in a living human ( mini- ) organ . Specifically , we present the first evidence that anagen hair MKs exhibit an active autophagic flux ex vivo , as documented by the presence of several LC3B-fluorescent perinuclear dots ( Fig 1 ) , which increase upon CQ treatment ( Fig 2A and 2B ) , and by demonstrating that autophagosomes representing different stages of autophagy are present in hair MKs ( Fig 2C and 2D ) . We further show that the number of autophagosomes decreases during the spontaneous involution of this ( mini- ) organ ( catagen ) ( Fig 3 ) , suggesting that intrafollicular autophagy may be modulated by several factors that are also involved in the regulation of HF cycling [40] . Because fibroblast growth factor ( FGF ) signaling is an important positive regulator of autophagy in chondrocytes [41] and is also involved in hair cycle control [40] , FGFs are among the most plausible regulators of intrafollicular autophagy . For example , fibroblast growth factor FGF7 ( also known as keratinocyte growth factor [KGF] ) is predominantly expressed in anagen and protects human HF from cell death induced by UV irradiation and chemotherapeutic or cytotoxic agents [42] , while FGF5 signaling controls catagen development [40] . Such candidate regulators of HF autophagy can now be probed in HF organ culture , using the methods and readouts reported here . Interestingly , the molecular controls that govern the anagen-catagen transformation in human HFs include profound changes in intrafollicular peripheral clock activity , whereby clock silencing prolongs anagen [18] . Recently , a connection between the circadian clock and autophagy has been reported in many systems [43–47] . For example , turnover of the clock protein brain and muscle ARNT-Like 1 ( BMAL1 ) involves both proteasomal and autophagic activities [48] . As BMAL1 knock-down in human HF significantly prolongs anagen [18] , it is conceivable that autophagy in the anagen hair matrix may impact on the peripheral clock in human HFs . Because inhibiting autophagy by ATG5 silencing induces premature catagen and enhances hair MK apoptosis ( Figs 4 and 5 ) , autophagic flux in the anagen hair matrix appears to be important for anagen maintenance . That this is not only an ex vivo phenomenon but also clinically relevant is suggested by the fact that CQ can cause telogen effluvium in patients taking this antimalarial medication [22] , which is caused by premature catagen induction [13 , 19] . Conversely , the principal ingredients ( core mix ) of a nutraceutical product used to treat hair loss potently promotes autophagy in organ-cultured human scalp HFs ( Fig 6A–6E ) and promotes anagen ( Fig 7A and S3 Fig ) , but fails to do so in ATG5-silenced HFs ( Fig 7B ) . Taken together , our data support a scenario in which intrafollicular autophagy is eminently targetable for the therapeutic modulation of human hair growth . In line with this concept , another agent that positively regulates autophagy , caffeine [49 , 50] , is sold as a hair growth–promoting cosmeceutical and has been shown to also prolong anagen and stimulate the proliferation of hair MK [51] . Autophagy inducers , which are the focus of intense ongoing research efforts [52–55] , are therefore promising agents for the treatment of hair growth disorders and drug-induced hair loss phenomena characterized by premature catagen induction [56] . However , currently available chemical inducers of autophagy have limited specificity for the autophagic process and produce several autophagy-independent effects [55] that may also affect the HF cycle . Our genetically-modifiable HF model provides a suitable human test system for evaluating the efficacy and specificity of additional autophagy inducers in promoting hair growth . Notably , HFs are continuously exposed to multiple , potentially noxious stimuli , ranging from contact with pathogens and the skin microbiome , UV light , and other DNA-damaging and/or oxidative stress-inducing external pressures , including drugs ( many of which cause hair loss as an adverse effect ) and internal stressors such as inflammation , metabolic disorders , and ageing [57–61] . To cope with this plethora of stressors , the human HF has established complex but highly efficient stress-response and stress-management systems [21 , 62–67] . The current data suggest that autophagy , a recognized stress-adaptive mechanism [2] , is prominently enrolled into these intrafollicular stress-response/-management systems to such an extent that down-regulating autophagy does not permit human HFs to sustain their growth under conditions of stress ( such as organ culture ) . This working concept can be followed up in vivo by studying human scalp skin xenotransplants onto immunocompromised mice , in which longer-term cycling of human HFs can be studied [16] . Interestingly , autophagy has also been found to be functionally important in regulating interfollicular epidermis ( IFE ) physiology and stress responses [68–71] . Thus , future functional comparative studies between intrafollicular and interfollicular autophagy may reveal an even more prominent function of autophagy in skin protection and function . Our study also introduces and validates scalp HF organ culture as an instructive , clinically relevant preclinical tool for translational , autophagy-related studies in the human system ex vivo . This can now be used to evaluate the toxicity of candidate drugs with regard to how they affect autophagy in human ( mini- ) organs . For example , many drugs , including antidepressants , anticonvulsants , antihistamines , and anticancer agents , negatively impact on autophagy by reducing lysosomal function ( lysosomotropy ) [72] . High throughput assays developed for measuring this [72] can now be complemented in a second step by human HF organ culture as a much more clinically relevant drug toxicity–screening assay . Here , the efficacy and tissue toxicity of candidate autophagy-promoting and -inhibitory agents can also be assessed directly ex vivo . Perhaps most importantly , the model introduced here provides the autophagy research community with an excellent tool for exploring the as yet insufficiently understood signals that regulate autophagy in human epithelial tissues as well as the functional roles of autophagy in a human ( mini- ) organ under both physiological and experimentally induced pathological conditions ( e . g . , chemotherapy [73] , interferon-gamma treatment [64] , ultraviolet radiation [66] , and oxidative stress [63] ) .
Discarded human scalp HFs or skin samples were obtained with informed , written consent following the “Declaration of Helsinki Principles . ” Full-length HFs used for the all the experiments were received and stored with ethical and institutional approval from the University of Manchester . A full list of patient numbers , sex , and age information is provided in S1 Table . Human osteosarcoma U2OS cells were acquired from the American Type Culture Collection ( ATCC ) . Human NCTC 2544 keratinocytes were kindly provided by Dr . Barbara Marzani ( Giuliani s . p . a , Milan , Italy ) . U2OS and NCTC 2544 were grown in Dulbecco’s Modified Eagle Medium High Glucose and Roswell Park Memorial Institute ( RPMI ) 1640 medium , respectively , containing 4 . 5 g/L D-Glucose , 4mM L-glutamine , 10% fetal bovine serum ( FBS ) , and 0 . 25 mM sodium pyruvate . U2OS cells were adopted as an established excellent model cell system to evaluate pro-autophagic activity of natural products [36] . IF microscopy staining for localization and quantification of autophagy proteins LC3B , SQSTM1/p62 , and ATG5 in situ was performed with anti-LC3B ( D11 ) XP Rabbit mAb ( Cell Signaling Technology ) , anti-SQSTM1/p62 ( abcam ) , and anti-APG5L/ATG5 ( EPR1755 ) ( abcam ) antibodies . Briefly , 5-μm-thick cryosections were fixed with cold acetone ( −20 °C ) for 10 min . After several washes in PBS , primary antibodies were incubated overnight at 4 °C . Alexa Fluor 555-conjugated donkey anti-Rabbit and Alexa Fluor 488-conjugated donkey anti-Mouse ( Thermo Fisher ) were adopted as secondary antibodies . Confocal images of HF derma papilla and of the area surrounding were taken by using a Confocal Microscope NIKON A1 , the 20 × 0 . 25 and 60 × 1 . 40 numerical aperture objective lens . Quantitative immunohistomorphometry in defined reference area using standardized light exposure was performed with Image J ( NIH ) software , as described [29 , 64] . Microscopic hair cycle staging was performed as previously described , using Ki-67/TUNEL immunostaining and Masson Fontana histochemistry [15 , 17 , 29] . For the quantification of positive Ki-67– and TUNEL-positive cells , intramesenchymal TUNEL+ cells were excluded from our quantitative immunohistomorphometry analysis and only epithelial TUNEL-positive cells were counted . In fact , TUNEL-positive cells in the dermal papilla and connective tissue sheath is a well-recognized artifact of hair HF organ culture [15 , 29] , because these regions of the human HF mesenchyme do not show apoptosis under physiological conditions [16] . Hair matrix cells were identified based on morphology and position relative to the dermal papilla [15 , 29] . Similarly , the quantifications of LC3B- , SQSTM1- , and ATG5-fluorescent signals were performed only on epithelial HF cells—specifically , the hair matrix cells and precortical matrix—while the connective tissue sheath and dermal papilla were excluded from analysis . Human scalp HFs were fixed in a mixture of 2% glutaraldehyde and 2% paraformaldehyde in 0 . 1 M cacodylate buffer for 2 h at room temperature and then processed for TEM as described in the literature [47] . Briefly , the samples have been washed in cacodylate buffer ( 0 . 1 M , pH 7 . 4 ) , postfixed for 2 h with 1% osmium tetroxide in the same buffer , extensively washed again , and then incubated overnight in a 0 . 5% uranyl acetate aqueous solution in the dark . After washing , the sections have been dehydrated in a graded alcohol series , and after a final dehydration in 100% propylene oxide , they have been infiltrated with low viscosity Spurr resin overnight and polymerized for 48 h at 65 °C . Sections of about 70 nm were cut with a diamond knife ( DIATOME ) on a Leica EM UC6 ultramicrotome . Bright field TEM images have been collected with a Schottky field-emission gun FEI Tecnai G2 F20 ( FEI , USA ) transmission electron microscope operating at an acceleration voltage of 200 kV and equipped with a 2k × 2k Gatan Ultrascan ( Gatan , USA ) charge coupled device ( CCD ) . For quantitative assessment of autophagy-related structures in TEM sections [74] , we acquired images of both the cytoplasm and the autophagic structures at magnifications that allowed clear identification of compartments and delineation of profile boundaries . More than 20 fields were randomly recorded for each HF to quantify the number of AVs/field/follicle . Analysis and statistics were further performed on the average number of AVs per field from at least three independent HFs per condition .
|
Human scalp hair follicles ( HFs ) experience a massive growth for years , until they spontaneously enter into a rapid , apoptosis-driven organ involution process , orchestrated by an organ-intrinsic “hair cycle clock , ” the molecular control of which remains unclear . Human HFs maintain in vivo–like characteristics , even after being removed from the body , and spontaneously run through a fundamental organ-remodelling process , traversing through a stage of growth ( anagen ) and destruction ( catagen ) as a ( mini- ) organ model . Here , we exploit this unique remodelling ( mini- ) organ to unveil a crucial new role of autophagy in the growth of human HFs . We show that hair matrix keratinocytes exhibit an active autophagic flux ex vivo during anagen , which is altered after the transition to catagen . We find that genetic inhibition of follicular autophagy induces premature catagen and enhances hair matrix keratinocyte apoptosis , suggesting that autophagic flux in the anagen hair matrix is important for the maintenance of this stage . Indeed , we find that the principal ingredients of a product used to treat hair loss induces autophagy in organ-cultured human scalp HFs and promotes anagen . We conclude that organ-cultured human HFs are a suitable ( mini- ) organ system to study both the role of autophagy in human physiology ex vivo and to test candidate agents that modulate autophagy under clinically relevant conditions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"cell",
"death",
"fluorescence",
"imaging",
"autophagic",
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"keratinocytes",
"hair",
"growth",
"integumentary",
"system",
"catagen",
"phase",
"scalp",
"cell",
"processes",
"epithelial",
"cells",
"physiological",
"processes",
"mathematics",
"statistics",
"(mathematics)",
"anagen",
"phase",
"research",
"and",
"analysis",
"methods",
"hair",
"imaging",
"techniques",
"animal",
"cells",
"biological",
"tissue",
"head",
"short",
"reports",
"cell",
"biology",
"anatomy",
"physiology",
"epithelium",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"statistical",
"data"
] |
2018
|
Autophagy is essential for maintaining the growth of a human (mini-)organ: Evidence from scalp hair follicle organ culture
|
The visual system must learn to infer the presence of objects and features in the world from the images it encounters , and as such it must , either implicitly or explicitly , model the way these elements interact to create the image . Do the response properties of cells in the mammalian visual system reflect this constraint ? To address this question , we constructed a probabilistic model in which the identity and attributes of simple visual elements were represented explicitly and learnt the parameters of this model from unparsed , natural video sequences . After learning , the behaviour and grouping of variables in the probabilistic model corresponded closely to functional and anatomical properties of simple and complex cells in the primary visual cortex ( V1 ) . In particular , feature identity variables were activated in a way that resembled the activity of complex cells , while feature attribute variables responded much like simple cells . Furthermore , the grouping of the attributes within the model closely parallelled the reported anatomical grouping of simple cells in cat V1 . Thus , this generative model makes explicit an interpretation of complex and simple cells as elements in the segmentation of a visual scene into basic independent features , along with a parametrisation of their moment-by-moment appearances . We speculate that such a segmentation may form the initial stage of a hierarchical system that progressively separates the identity and appearance of more articulated visual elements , culminating in view-invariant object recognition .
It is well established that the receptive fields ( RFs ) of neurons in the early visual cortex depend on the statistics of sensory input and can be modified by perturbations of those statistics during development [1]–[6] . This relationship has been studied theoretically in many ways . Phenomenological models have focused on the mechanisms of synaptic plasticity and axon-guidance , giving mathematical or computational accounts of how Hebbian-like learning rules may combine with sensory stimulation to drive the formation of cortical response properties [7]–[12] . Constrained optimality approaches look beyond the details of the synaptic learning rule , and ask whether the observed pattern of cortical responses has been selected to optimise a functional objective . Many early studies of this type were founded on the information-theoretic ideas of efficient coding and redundancy reduction [13] , [14] , and proposed that RFs had adapted to maximise the transmission of information from the periphery [15]–[18] . More recent work has generalised this approach to consider other possible objective functions with different representational or metabolic benefits . Two established alternatives are the sparseness and temporal stability objective functions . In the sparse-coding view neuronal properties are optimised so that neurons remain silent most of the time , responding vigorously to only a limited subset of all stimuli [19]–[21] . Thus every image is represented by relatively few active neurons . Such a representation makes it easy to detect “suspicious coincidences” [22] and reduces energy consumption [23] . It can also be related to the older objective of information efficiency [19] . Under the temporal stability objective , neuronal RFs are adapted so that their output firing rates vary slowly in time [24]–[26] . To achieve stability , neurons must learn to be insensitive to typical rapid transformations of their input , leading to invariant representations that simplify recognition tasks [27] . The generative modelling approach takes a complementary functional view . It is based on the Helmholtzian account of perception as inverse inference ( sometimes called analysis-by-synthesis ) . That is , that the goal of the perceptual system is to infer from sensation the environmental causes most likely to be responsible for producing the sensory experience [28] , [29] . In this view , sensory cortex implicitly embodies a model of how external causes interact to form the sensory input ( a causal generative model ) ; given a particular sensory experience , cortical processing inverts the model to infer the most likely causes of the sensory activity . Mathematically , this corresponds to an application of Bayes' rule . This general view that the brain carries out or approximates some form of probabilistic inference is supported by a number of psychophysical , anatomical , and physiological results ( see [30] , [31] for reviews ) . Many models that have been formulated in terms of the optimisation of an objective function could also be viewed as implementing inference within an appropriate generative model: the assumptions and structure of the model are implicit in the objective function . Thus , recoding based on the sparseness objective corresponds to inference within a generative model in which a number of independent , sparsely active causes combine linearly to form the image [20] . Similarly , the goal of redundancy reduction has led to models in which divisive normalisation reduces second-order dependence between linear recodings of an image [32]; in the generative view , this corresponds to joint modulation of the variances of otherwise independent sparse causes [33] , [34] . Finally , the temporal stability objective corresponds to a model with causes that are independent of one another , but stable or predictable in time [35] . A remarkable success of these functional models , whether formulated generatively or in terms of a representational objective function , is that , when used to learn an appropriate representation from a set of natural images , they yield elements that mirror a number of response properties of primary visual cortical neurons ( though some notable discrepancies do remain [16] ) . However , despite this success , the generative models involved match only the lowest-level statistics of natural images . Images generated from the learnt models have naturalistic textural properties , but none of the higher-level structure of the natural world . If this approach is to provide insight into higher processing within the visual cortex then appropriate structure must be introduced to the models . In the present study we focused on one basic structural aspect of the environment: The visual world is largely composed of discrete objects , which each contributes a set of discrete visual features to the image . Moreover , the objects , and therefore their associated features , usually remain in view for some time , although their precise appearances might change gradually due to changes in viewpoint , lighting or in the object's position . We thus formulated a model in which the identity of the visual elements present was signalled by a set of binary-valued variables , while their appearances each evolved separately under the control of continuous attribute variables . This independent control of appearance stands in contrast to a related idea of “content” and “style” [36] , [37] where the transformation of appearance is usually shared across the image or image patch . This comparison is taken up in greater detail in the Discussion . We fitted this model to natural video images , without using any additional information about which elements were present or what their transformations might be . We found that the model naturally learned biologically plausible features , with low dimensional manifolds of attributes . Many aspects of the learnt representation corresponded closely to both anatomical and functional observations regarding simple and complex cells in the primary visual cortex ( V1 ) . Thus , the model offers a functional interpretation for the presence of two main classes of cells in V1 . Complex cells represent the probability of presence of an oriented feature , while simple cells parametrise the precise appearance of the feature in the visual input . We speculate that a similar representation in the form of feature identities and attributes may continue up the visual hierarchy , ultimately contributing to view-independent object recognition .
Figure 1A illustrates the intuitions that underlie the general structure of the model . The image at each point in time—represented by a vector shown at the bottom of the figure—is composed from a set of visual elements illustrated by the objects in the top row . Only a small subset of all the possible elements contributes to any one image . The identity of these active elements is represented by a set of binary-valued variables , where means that the th element appears in the image at time . If active , the form of the element in the image may vary; for instance the object may appear at any position or orientation . Each element is thus associated with a set of possible contributions to the image , which form a manifold embedded within the space of all possible images . The configuration of element at time is then specified by a vector , with dimensionality equal to that of the manifold . We call the elements of this vector , , the attributes of the visual element . The shape of the manifold is described by a function , which maps this attribute vector to the partial image it describes . For concreteness , consider the rightmost panel of Figure 1A , which represents a model for a beverage can . The fact that the variable takes the value indicates that the object is present in the image at time . The arrow indicates the point ( encoded by ) on the manifold where the can has a particular position and viewpoint in the input visual space . If one of the attribute variables were to correspond to the orientation of the can , changing its value would trace a trajectory on the manifold , which would result in a rotation of the object in the image space . The set of partial images associated with all of the active elements then combine through a function , which could in principle implement occlusion , illuminant reflection , or other complex interactions , to yield the image: ( 1 ) where we have included an additive , independent noise term . In this abstract form the model is very powerful , and provides an intuitively satisfying generative structure for images . Unfortunately , for manifolds and combination functions modelling the appearance of entire complex objects and the interactions between them as illustrated in Figure 1A , the task of inferring the elements and their appearances from natural data is intractable . To explore the potential of the framework we adopted a simplified form of the model , taking the mappings to be linear ( equivalently , we defined the attribute manifolds to be hyperplanes ) and to sum its arguments . This allowed us to implement the selection of the active elements by multiplication: ( 2 ) where the basis vectors parametrise the linear manifold , and and are the number of identity variables and the ( maximum ) dimensionality of each attribute manifold respectively . In this simpler form , we expect the visual elements to correspond to more elementary visual features , rather than to entire objects ( Fig . 1B ) . The complete probabilistic generative model for image sequences includes probability distributions over the identity and attribute variables . We chose distributions in which objects or features appeared independently of one another , and where the probability of appearance at time t depended on whether the same feature appeared at time . The attributes of the feature evolved smoothly , again with a Markovian dependence on the preceding state . The formal definition of the probabilistic model is given in Methods . The parameters of the model specify the partial images generated by each feature ( represented by the basis vectors ) , the probability of each feature being active , and the degree of smoothness with which the appearance of the feature evolves . All of these parameters were learnt by fitting the model to natural image sequences . In previous work on sparse coding the number of basis vectors or components needed has been explored outside of the model fitting procedure ( for example [38]; but see [39] ) . Crucially , here we were able to learn the dimensionalities of the model—the numbers of visual elements and associated attribute variables—from the data directly , using Bayesian techniques described below and in Methods . Probabilistic models are often fit by adjusting the parameters to maximise the probability given to the observed data—called the likelihood of the model . In practice , image models have often been fit by maximising the data probability under settings of both the parameters and the unobserved variables ( in our case these would be the identity and attribute variables ) , a procedure which may be severely suboptimal [40] . Here , we adopted an iterative procedure called Variational Bayes Expectation Maximisation ( VBEM ) [41] , [42] to learn an approximation to the full probability distribution over the parameters and unobserved variables implied by the data—known as the VB posterior distribution . This posterior provides a more robust estimate of the parameters , with concomitant estimates of uncertainty , and can be used to determine the appropriate dimensionality of the model directly . More complex models can always be adjusted to give higher probability to any data set , and so the maximum likelihood approach would always favour a model with greater dimensionality . This effect can lead to overfitting , where an overly complex model is selected . However , because there are very many more possible parameter settings in a complex model , any one such parameter setting may actually be very improbable even though it might fit the data well . Thus , when considering the probabilities of parameter settings and models as in the Bayesian approach , a form of “Occam's Razor” comes into effect favouring descriptions complicated enough to capture the data well but no more so [43] . For models similar to the one developed here , one consequence of this “Occam's Razor” is that the posterior probability distributions on the values of any superfluous basis vectors concentrate tightly about 0 , effectively pruning the basis dimension away , and leaving a simpler model . In this context , the effect has been called Automatic Relevance Determination or ARD [42] , [44] . Bayesian estimation is well-defined only if a prior distribution—that is , an initial probability distribution determined before seeing the data—is specified . The prior on the basis vectors was of a form often used with ARD , with a so-called hyperparameter determining the concentration about a mean value of 0 . The prior distributions on the parameters that determine the temporal dependence of identity and attribute variables were broad enough not to influence the posterior distribution strongly . The exact definitions of the distributions over parameters , along with details of the fitting algorithm , are given in Methods . We used this model to investigate the visual elements that compose natural images , comparing features of the representation learnt by the model when fit to natural image sequences to the representation found in V1 . The input data were a subset of the CatCam recordings [45] , which consist of several-minute-long video sequences recorded by a camera mounted on the head of a cat freely exploring a novel natural environment . Temporal changes in the CatCam videos are caused partly by moving objects , but mostly by the animal's own movement through the environment . Cats make few saccades and use only small eye movements to stabilise the image during locomotion [45] , so that the amplitude and frequency of spatial transformations in the videos ( translation , rotation , and scaling ) is similar to that experienced by the animals . Computational constraints prevented us from modelling the entire video sequence . Instead , we fit the model to the time-series defined by the pixel intensities within fixed windows of size pixels over 50 frames . We initialised the model with 30 identity variables each associated with attribute manifolds of 6 dimensions and let the algorithm learn an appropriate model size by reducing the number of active attribute dimensions and identity variables by ARD . We performed a total of 500 VBEM iterations , at each iteration taking a new batch of 60 sequences of 50 frames , randomly selected from the entire dataset . Further computational details are given in Methods . Given an observed image sequence , the model could be used to infer a posterior probability distribution over the values of the identity and attribute variables at each point in time . We compared the means of these distributions to the firing rates of neurons in the visual cortex . The use of the mean was necessarily arbitrary , since there is no generally agreed theory linking probabilistic models to neural activity . The brain may well represent more than a single point from this distribution . For example , information about the uncertainty in that value would be necessary to weight alternative interpretations of the data . Once the model had been fit to the data , however , we found that the attribute variable distributions estimated from high-contrast stimuli were strongly concentrated around their means . Thus , many different choices of neural correlates would have given essentially identical results . It is also worth mentioning here that although the identity variables describe the presence or absence of a feature in the generative model and are thus binary-valued , the posterior probability of the feature being present ( which is the same as the posterior mean of the binary identity variable ) is continuous . Thus , neurons presumed to encode these posterior means would respond to stimuli with graded responses , which would take uncertainty about feature identity ( e . g . , under conditions of low contrast ) into account . Figure 2A shows the VB posterior mean basis vectors learnt from the CatCam data . Each row displays the basis vectors of the attribute manifold corresponding to a single identity variable . Since the manifold was a hyperplane , the set of possible feature appearances was given by all linear combinations of the basis vectors ( Fig . 3D ) . For every manifold , the mean basis vectors resembled Gabor wavelets with similar positions , orientations , and frequencies , but different phases ( Fig . 4A–C ) . Thus every point on the manifold associated with a single feature corresponded to a Gabor-like image element with similar shape , orientation , and frequency , but variable phase and contrast . When presented with a drifting sine grating of orientation and frequency similar to that of the basis vectors , the probability of the feature being present was found to approach 1 rapidly , and then to remain constant , while the means of attribute variable distributions oscillated to track the position of the sine grating on the manifold , as illustrated in Figure 3 . Attribute variables thus behaved much like simple cells in V1 , in that they responded optimally to a grating-like stimulus and oscillated as its phase changed , while identity variables responded like complex cells , being insensitive to the phase of their optimal stimulus . In electrophysiological studies , the classification of neurons into simple and complex cells is done using a relative modulation index [46] , [47] , which is defined as the ratio of the response modulations ( F1 ) to the mean firing rate ( F0 ) in response to a grating with optimal orientation and frequency , but varying phase . Cells that respond to phase changes with large oscillations have relative modulation larger than 1 and are classified as simple cells , while cells that are invariant to a phase change are classified as complex cells . We computed the relative modulation for the posterior mean values of the variables in our model . All identity units were classified as complex ( maximum F1/F0 ratio 0 . 28 ) and all attribute units that had not been pruned during learning were classified as simple ( minimum F1/F0 ratio 1 . 45 ) . The magnitude of relative modulations for attribute and identity units is comparable to that found in simple and complex cells in the primary visual cortex of macaque and cat , although the population distribution is narrower [47] ( Fig . S2 ) . By contrast to the standard energy model of complex cells [48] , here complex and simple cells did not form a hierarchy , but rather two parallel populations of cells with two different functional roles: the former coding for the presence of oriented features in its receptive fields , the latter parametrising local attributes of the features ( primarily their phase ) . To explore this connection further we compared the properties of simple cell RFs in V1 as reported in the physiological literature with the ‘RFs’ of the attribute variables . The RF of an attribute variable was defined by analogy to the conventional physiological definition . We fixed the posterior distribution over the parameters of the model to that estimated by VBEM from the natural data , and then examined the values of the attribute variables that were inferred given coloured Gaussian noise input . The RF was defined to be the best linear approximation to the mapping from this input to the inferred mean attribute value , a procedure equivalent to finding the “corrected spike-triggered average” or Wiener filter [49] ( see Methods ) . Although nonlinearities in the model and inference meant that these RFs differed slightly from the basis vectors associated with the attribute variables , we found them to be visually indistinguishable ( Fig . S1 ) . We then computed the orientation , spatial frequency and phase for the resulting RFs by fitting a Gabor function to each of the filters ( Methods; Fig . S1 ) . Figure 4 ( A–C ) shows the orientation , frequency , and phase for each pair of RFs associated with the same identity variable ( thus , a feature with a 4-dimensional attribute manifold contributed 6 points to each graph ) . In the visual cortex , neurons performing related computations appear to be co-located [50] , [51] . Since the responses of related neurons are highly dependent given a visual stimulus , this may reflect a computationally efficient solution that minimises wiring length [11] , [52] . We compared our data to the results reported in [53] for pairs of simple cells recorded from the same electrode in area 17 of the cat visual cortex ( Fig . 4D–F ) . In both the model and physiological reports , the two orientations in each pair of RFs agreed very closely; the frequencies slightly less so; while no relation was apparent in phase . The distribution of preferred frequencies and orientations in the RFs of attribute variables are shown in Figure 2 B , D . The distribution of frequencies is quite broad compared to that found in models based on sparse coding or independent component analysis ( ICA ) [16] , [54] , where RF frequencies tend to cluster around the highest representable value , and compares well with the width of the distribution in simple cells ( Fig . 2C ) [55] . The joint distribution of orientation and frequency ( Fig . 2E ) covers the parameter space relatively homogeneously . Note that the CatCam image sequences have less high-frequency power at horizontal orientations , and this bias is reflected in the results . Figure 5 shows the joint distribution of RF width and length in normalised units ( number of cycles ) in our model and for simple cell RFs as reported by Ringach [56] , [57] for area V1 in the macaque . The aspect ratios are similar in both cases ( again , contrasting with typical sparse coding results [58] ) , although the model results tend to have larger RFs , possibly again due to the particular content of the video . The model was initialised using a representation that contained 6-dimensional attribute manifolds for each feature . However , in the posterior distribution identified by VBEM , the probability of the basis vectors corresponding to many of these dimensions being non-zero vanished—that is , a model in which the image data were described using fewer dimensions was found to be more probable . In fact , the VB posterior representation was only slightly overcomplete , with 96 basis vectors representing an 81-dimensional input space , and with the dimensionality of most feature manifolds lying between 2 and 4 ( Fig . 6A ) . Given the proposed identification of identity variables with complex cells , this gives a prediction for the dimensionality of the image-subspace to which a V1 complex cell should be sensitive . The subspace-dimensionality of a complex cell may be estimated by finding the number of eigenvalues of the spike-triggered covariance ( STC ) matrix [59] that differ from the overall stimulus distribution . One study [60] has reported , for complex cells in the anaesthetised cat , a distribution of dimensionalities that peaked sharply at 2 , with only a few complex cells being influenced by 1 , 3 , or 4 dimensions . A more recent paper published by the same group has found a broader distribution in the awake macaque [61] . An analysis of the RFs of the identity variables made using an equivalent procedure revealed a comparable distribution for our results ( Fig . 6B ) . ( The number of significant eigenvectors returned by the STC analysis can be slightly different from the dimensionality of the attribute manifold because of the non-linear interactions with other variables in the model . ) The model distribution is skewed slightly towards a larger number of stimulus dimensions; although this may be because the sample in [61] included both simple and complex cells . A second study [62] performed a similar analysis using spatio-temporal stimuli and found 2 to 8 significant dimensions for complex cells . This broad range of dimensionalities agrees qualitatively with our results . Unfortunately , quantitative comparison with this study is unreliable as the physiological RFs were identified in effectively one dimension of space , and one of time , while the basis vectors of the attribute manifolds span two spatial dimensions , without a temporal aspect . A key aspect of our model is the temporal dependence of the identity and attribute variables . To ask what role this temporal structure had on the feature basis vectors found , we shuffled the order of frames in the CatCam database , and then refit the model using exactly the same procedure as before . When using unshuffled data , the learning process adapted the feature manifolds so that the inferred values of identity variables persisted in time , while the inferred attribute variables changed smoothly . In the shuffled data such a persistent and smooth representation cannot be found . Instead , learning adjusts the attribute manifolds so as to maximise the independence of the associated identity variables , grouping together attribute dimensions that tend to co-occur in single frames . This is similar in spirit to Independent Subspace Analysis [63] , or to a Gaussian Scale Mixture model [33] with shared binary-valued scale parameters [64] . Figure 7 shows the basis vectors and pairwise distributions of their properties found for the shuffled data . The VB posterior distribution concentrated on a more overcomplete representation ( 122 basis vectors representing 81 input dimensions ) than for the unshuffled data . Some manifolds were pruned away entirely , while the majority of those that remained preserved the maximum dimensionality of 6 . The basis vectors still resembled oriented features , although the fit of the linear RFs with Gabor wavelets was worse on average than that obtained for the unshuffled video , or seen in physiological data . The fractional error of fit ( sum of squares of the residuals divided by the sum of squares of the RFs ) was for simple cells [53] , for the model learnt from unshuffled data , and in this case ( Fig . 8 ) ( see Fig . S1 and S3 , for the reverse-correlation filters and Gabor fits ) . As shown in Figure 7 ( b–d ) , attribute variables associated with a single identity still agreed in orientation , but not in phase . However , in contrast to the model learnt from unshuffled sequences and to the physiological results , there was much poorer correspondence in spatial frequency ( compare Fig . 7C to Fig . 4B , E ) . According to their relative modulation index , identity variables would still be classified as complex cells ( maximum F1/F0 ratio 0 . 63 ) , and attribute variables as simple cells ( minimum F1/F0 ratio 1 . 34 ) . Despite finding a larger number of basis vectors , the model described a larger proportion of the shuffled data as noise , thereby fitting them more poorly . We evaluated the probability given to 50 new batches of 3000 frames each by the parameter distributions learnt from the shuffled and unshuffled data . As estimated by the VB approach , the probability assigned by the unshuffled model was more than times greater ( more precisely , the free-energy—a lower bound on the log probability that is maximised by the VBEM algorithm—was larger by NATS , i . e . between 1 . 7% and 4 . 5% greater; Methods ) . Overall , when deprived of temporal structure in the observations , the algorithm converged to a worse model of the video , and one which was less similar to the physiological data . It is interesting to note that despite these deficiencies in the representation learnt from shuffled sequences , the basis vectors of the attribute variables still resembled simple cell RFs . This observation stands in contrast to results from previous models of complex cells based on temporal stability , which had assumed a hierarchical organisation similar to the classical energy model [25] , [26] . In those models the only signal available to shape the simple cell RFs derived from the temporal stability imposed on the corresponding complex cells . If this signal were removed by shuffling the input frames , the simple cells would be unable to develop any sort of organised response . In our model , however , the independence effect discussed above was still able to provide a learning signal for the attribute manifold in the absence of temporal stability . Thus , we predict that even if stimulus temporal correlations were disrupted during learning , for example by rearing animals in a strobe-lit environment , simple-cell responses would still emerge; although the receptive fields ( defined by reverse correlation ) would fit Gabor wavelets less accurately , and anatomical subunits would be less well-grouped in spatial frequency . In fact , experimental evidence from Area 17 in strobe-reared cat seems to support our results . After strobe rearing at an 8 Hz frequency , the spatial RF structure of simple cells in area 17 remained intact except for their width , which was found to increase; and for direction selectivity , which was mostly lost [65] . Studies performed with lower strobe frequencies ( 0 . 67–2 Hz ) found other changes in the RF properties , including an increase in the number of cells classified as non-oriented , a slight decrease in orientation selectivity , and a reduction of the frequency of binocular cells [66] . In addition , given the increase in the dimensionality of the attribute manifold , we predict that an STC analysis of complex cells in strobe-reared animals would show a larger number of relevant dimensions .
We have investigated a new generative model for images which makes explicit the separation between the identity of a visual element and the attributes that determine its appearance . This structure within the model makes it possible to extract and bind together attributes that belong to the same visual element , and at the same time to construct an invariant representation of the element itself . We modelled identity with a set of binary-valued variables , each coding for the presence or absence of a different feature . Their appearances were described by manifolds , parametrised by a set of attribute variables . Both identity and attribute variables were assumed to exhibit temporal dependence within image sequences . We were also interested in determining the size of the model , i . e . , the number of attribute and identity variables required to optimally describe the input data . This was achieved by performing a Bayesian analysis of the model , which avoids over-fitting and involves defining an appropriate prior distribution over the generating basis vectors . As a result , after convergence of an iterative algorithm , only the basis elements needed to effectively match the data remained active and all redundant attribute directions were pruned away , avoiding overfitting the image data . The algorithm was applied to natural image sequences in order to learn a low-level representation of visual scenes . The filters associated with the individual attribute variables were shown to have characteristics similar to those of simple cells in V1 . The RFs of attributes associated with the same identity variable had similar positions , orientations , and frequencies , but different phases . As a consequence , the corresponding identity variable became invariant to phase change and behaved like a complex cell . In the standard energy model of complex cells and in several previous functional models , complex and simple cells form a hierarchy . Simple cells have the role of subunits and are regarded as an intermediate step on the way to the complex cell . Their phase-dependent information is then discarded as a first step towards the construction of an invariant representation . Here complex and simple cells do not form a hierarchy , but rather two parallel interacting populations of cells with two different functional roles: the first coding for the presence or absence of oriented features in its RFs , the latter describing local parameters of the features ( mainly their phase ) . A formal analysis of the model reveals that , indeed , the interaction between identity and attribute variables in our model is richer than in the energy model . In addition to a quadratic term similar to the one in the energy model inside an exponential , the interaction includes a divisive normalisation term , and dependence on the statistics of natural input and the prior probability of the feature encoded by the identity variable being present ( Text S1 ) . Intriguingly , some physiological data [67] and biophysical models [68] , [69] have also suggested a non-hierarchical relationship between simple and complex cells . However , these results have suggested a spectrum of “simple-” to “complex-like” behaviour within a single population . By contrast , our view preserves the notion of two distinct classes of cell with different response property and computational role , but which are organised in parallel rather than hierarchical populations . In Results , we showed that properties of RFs learnt within our model agreed with a broad range of existing physiological data . A further aspect of the model could be tested if it were experimentally possible to identify and record simultaneously from a complex cell and the simple cells that form the subspace related to it . First , a direct consequence of the non-hierarchical organisation of complex and simple cells is that increasing the probability of a feature being present in the visual input by stimulating the complex cell should result in the corresponding simple cells becoming active ( as they seek to describe the attributes of the feature whose presence has been asserted by activation of the complex cell ) . This is in contrast to the behaviour implied by the feed-forward energy model , where complex cells would not influence the activity of simple cells . A similar test might exploit the temporal persistence in the identity variable corresponding to the complex cell . Consider two sequences of visual stimuli which both end in a frame well-matched to the RF of one of the simple cells . If the preceding frames had matched the RFs of the other simple cells associated with the same complex cell , and therefore had activated the complex cell , the temporal persistence within the corresponding identity variable should maintain that activation and thereby facilitate the response in the simple cell . Conversely , if the preceding stimuli had fallen outside the feature manifold , the simple cell might be less strongly activated . The computational power of a class of models similar to the one in this paper has been investigated by Tenenbaum and Freeman [36] , and Grimes and Rao [37] . These models were based on the bilinear interaction between two sets of variables: content variables , which described the appearance of the input data ( e . g . , a prototypical handwritten digit , or the appearance of an image patch in a model of visual input ) , and style variables , which parametrised transformations of the content ( e . g . , the style of the digit or global translations of the patch ) . Tenenbaum and Freeman [36] showed that the rich nonlinear interactions between these two factors facilitated classification and extrapolation in a series of experiments using spoken vowels , letters in different fonts , and faces in different poses . Grimes and Rao [37] assumed a sparse prior distribution over content and style variables , and applied the model to translated natural images . The learnt basis vectors were shown to represent oriented features and to be largely invariant to local translation . Although learning was based on natural images , content and style play mathematically symmetric roles within these models , and thus could not be identified from the images alone . Instead , the content and style variables were partially fixed , so that all that needed to be learnt were the corresponding basis vectors and transformations . In this paper , the semantic difference between the identity and attribute variables , and the temporal persistence assumption , meant that the model could be learned in a completely unsupervised fashion from natural movies . In our model , the input images result from the combination of multiple visual elements , identified by the identity variables . The appearance and transformation of each of these elements is separately encoded by the associated attribute variables . Thus , the role of the attributes is a combination of the role of content and style variables in the previous models . In the model described here , the appearance manifolds associated with each feature are linear , and they combine additively to form the image . These choices are a matter of computational tractability , and have two main limitations . First , the additive combination function f is unable to model effects such as occlusion , shadowing , or reflective illumination . Linear models like sparse coding and ICA also assume the same kind of linear superposition , and it is unclear at this stage how much a more realistic f would influence the results at the level of small image patches [70] . Second , the linear feature manifolds do not allow global transformations of feature appearance , such as translation or rotation , to be captured by a single attribute dimension . Each attribute is , at best , able only to model a local , linearised version of the transform . However , global properties may still be approximated using several attribute dimensions , or by a hierarchical model in which a higher-order feature with a global translation attribute generates local features where needed at a lower level ( cf . [71] ) . Another simplification concerns the temporal aspect of V1 RFs . As in most computational models of V1 neurons , we did not attempt to match the temporal behaviour of early visual neurons , again because of computational constraints . Currently , the model defines a Markov temporal dependency for the variables in the model , which is intended to capture a simple timescale of persistence . This temporal model implicitly defines a spatio-temporal receptive field ( STRF ) for attribute and identity variables . However , the Markov assumption does not allow the model to express the more complex temporal behaviours observed in V1 neurons , such as direction selectivity . Instead , the resulting STRF is formed by the spatial RF , as shown in Fig . S1 B , decaying exponentially in time . In previous work , temporally extended RFs have been modelled by augmenting the input data with the pixel intensities of patches at neighbouring times , and then building a model of the augmented data set [26] , [72] . However , from a generative point of view this does not seem to be appropriate , as the model would independently generate pixel intensities in overlapping temporal windows , which would give multiple inconsistent proposals for the pixels values at any particular time . In our case , we would need to use a more complex model of temporal dependencies , for example by allowing temporal dependencies between attribute variables in the prior ( i . e . , by defining matrix in Eq . 9 to be full instead of diagonal , or by introducing a non-Markov structure ) . It may be possible to extend the model developed here so as to represent more complex visual elements . One approach is illustrated in Figure 9: In the schematic , high-level identity variables may represent entire objects . These generate lower-order elements , like parts of an object or image features . For example , the activation of an identity variable corresponding to a face would activate , with high probability at the lower level , variables coding for the presence of eyes , nose , and mouth . Similarly , high-level attributes , like the size and viewpoint of the face , would influence low-level attributes such as the position of its individual parts , and may also determine which parts are visible . The hierarchy may then be repeated down to individual image features . Such a hierarchical organisation would be closely related to the hierarchical nature of the environment . The connections between higher-order and lower-order identity variables , for example , would encode whole-part relationships , while the connections between higher- and lower-order attributes would encode structural constraints between the individual parts necessary to form the whole . Such a structure would allow the visual system to benefit from the advantages of a recognition-by-components architecture , including the ability to reuse known parts to form novel objects , and to express the wide range of possible configurations of articulate objects [71] , . The computer vision community has long been interested in the analysis of images for the categorisation and recognition of objects . A recent trend in the field has been to build hierarchical generative models of objects composed of sub-parts; this line of research has found that such a hierarchical representation can indeed increase the performance of the algorithm [74]–[77] . These computer vision models generally start by describing the image using a standard , fixed set of features , and pre-specify the transformations that these can undergo; the object model may also be pre-specified [74] or may be learnt from data [75]–[77] . Moreover , categorisation is typically supervised . Our approach is in many ways complementary , in that it starts from the bottom up , and requires no supervision ( see [70] , [78] , [79] for comparable bottom-up computer vision approaches ) . Our results show that it is possible to learn simple but meaningful features from natural images , and at the same time learn the transformations that they are subject to in natural vision . It remains to be shown , however , whether our method can be extended successfully to represent more complex objects . Algorithms related to the temporal stability principle have also been applied with some success to learning a high-level object representation [27] , [80]–[82] . In [80] , [82] , the representation is invariant to frequent transformations , such as translation , and the corresponding attribute information ( e . g . , position ) is discarded . In [27] , [81] it is shown that the representation learnt by the Slow Feature Analysis algorithm preserves the attribute information . However , the model does not make any semantic distinction between variables carrying these two kind of information , so that a readout system downstream of the sensory cortex would need an additional criterion in order to access them . We believe that the additional structure in our model will help in extracting a high-level representation of objects from natural scenes . Moreover , a readout system would have access to more structured information about the environment , and could access differentially the identity information – for example in recognition tasks , as identity is invariant to all possible appearances parametrised by the attributes – and the attributes – for example , to guide reaching behaviour . In the introduction we discussed how it is possible to interpret functional models based on constrained optimisation of an objective function from a generative perspective . From this point of view , concentrating on a single computational objective appears rather simplistic , given the complexity underlying any natural scene . We argued that by developing models in the generative framework , one is able to develop models of vision that are closer to the true visual generative process . A common critique of the generative approach is that it seeks to model every aspect of its input , while the visual system might be interested in extracting only a behaviourally relevant subset of the sensory information . This argument implicitly assumes that it would be easier and more useful for the visual system to extract only relevant information ( e . g . , object position ) while ignoring “nuisance” information ( e . g . , light reflections ) . On the other hand , the representation formed by the visual system has to be used for many different tasks , and as such it is almost impossible to decide a priori which information should be discarded . A complete generative account of the visual data is more flexible as it identifies and separates all the different causal influences that contribute to the scene , and makes them available for context-specific processing . By contrast , a system that selectively discards parts of the visual signal might find it difficult to adapt when that discarded information became relevant ( e . g . , in an hypothetical task where light reflection predicts reward ) . Moreover , it is in principle possible to define partial generative descriptions of the visual signal . The key is that generative models explain their input probabilistically up to a certain level of “noise” ( e . g . , the term in Eq . 1 ) . The noise term includes genuine noise in the input and more generally all aspects of the input that the model can not capture , or is not interested in capturing . Thus , by building a more complex model of noise , a generative model could selectively describe only the subset of aspects of the stimuli that it considers relevant: Suppose that in one task , all that was important was the identity of a visual feature , not its specific appearance . Then the attributes in our model would be regarded as “nuisance” variables . Ideal inference about the identities would proceed by integrating over the uncertainty in the “nuisance” variables – in essence , they would form part of a complex noise model . This integration may be explicit ( and possibly approximate ) as in our VB implementation . It may also be implicit in a model with a more flexible definition for the noise ( e . g . , by learning different noise parameters for different dimensions ) . This paper has presented a first step toward including constraints regarding the structure of the visual environment in computational models of vision . By taking into account the conceptual distinction between identity and attributes of visual elements , we were able to match more closely the physiological and anatomical organisation of V1 . Further steps in this direction will hopefully lead us toward the development of a more complete , probabilistic account of visual inference .
The generative model describes the probability of a sequence of image patches , each one described by a vector of pixel intensities , in terms of binary-valued identity variables and associated attribute vectors , each of dimensionality , . The generative process maps these hidden identity and attribute variables to observations according to Eq . 2 . Assuming Gaussian noise with variance along observed dimension , corresponding to a diagonal covariance matrix , the probability of observing an input sequence conditioned on a setting of the hidden variables is: ( 3 ) where denotes a Gaussian distribution over with mean μ and covariance . The prior distributions over the variables were defined according to the intuitions described in the introduction , namely that visual elements should appear independently of one another and for extended periods of time , and their appearances should vary smoothly . This was translated into a prior distribution over identity and attribute variables as follows . Identity variables were modelled as independent , binary Markov chains with initial-state probabilities and a transition matrix comprising probabilities : ( 4 ) ( 5 ) ( 6 ) Our intuition that objects are persistent in time is respected when the probability of remaining in the current state is larger than that of switching , i . e . when the transition probabilities and are larger than . While comparable results may have been obtained by setting these parameters to a suitable value , we chose to remain within the Bayesian approach and instead expressed our belief as a prior distribution over values of ( specified below ) . The attribute variables are continuous and their evolution was modelled by Linear State Space Models with initial variances , transition matrices and transition variances : ( 7 ) ( 8 ) ( 9 ) The matrices and were defined to be diagonal , so that attributes were uncorrelated; and were related by the equation , so that the variance of the attribute variables was 1 in the prior [35] . This imposed an absolute scale , eliminating rescaling degeneracy . Slowly-varying variables have a positive autocorrelation , and would thus have parameters between 0 and 1 , with larger values corresponding to slower variables . Again , we expressed the belief in smoothness softly , by imposing a suitable prior distribution over these parameters ( see below ) . The priors on the basis vectors were Gaussian , with precision hyperparameters : ( 10 ) These zero-centred Gaussian prior distributions discouraged large components within the basis vectors . The widths of the distributions are set by the which were learnt alongside the other parameters . This choice of prior [35] leads to a pruning of basis vectors during learning , through ARD [42] , [44] . Since the basis vectors of redundant attribute dimensions are free to match the prior , and as this is centred on the origin , they are driven to zero . The precision hyperparameter can then diverge to infinity , effectively eliminating the basis dimension from the model . As a result , only the dimensions of the attribute manifold that were required to describe the data without overfitting remained active after learning . For the remaining parameters we also chose conjugate priors . Conjugacy means that the posterior distribution has the same functional form as the prior , resulting in tractable integrals . Conjugate priors are intuitively equivalent to having previously observed a number of imaginary pseudo-observations under the model . By choosing the number of pseudo-observations we can regulate how informative the prior becomes . In summary , the prior over the image noise precision was taken to be a gamma distribution with parameters , the prior over the transition matrix T was Dirichlet with parameters , and the prior over was a nonstandard distribution ( due to the coupling between mean and variance of ) in the exponential family that required 4 hyperparameters to be specified ( , and ) . The complete directed graphical model showing the dependencies between variables is depicted in Figure 10 . In the Bayesian formulation the parameters of the model are formally equivalent to hidden variables , differing only in that their number does not increase with the number of data points . The goal of learning is then to infer the posterior joint distribution over variables and parameters given the data: ( 11 ) where indicates the ensemble of all parameters and all hyper-parameters ( in the following for simplicity we will omit the dependence on ) . Although this distribution is intractable ( as in most non-trivial models ) , it is possible to use a structured variational approximation to obtain a tractable alternative . The idea is to introduce a new factored distribution in which some dependencies between the variables are neglected , while keeping the rest of the distribution intact . Learning proceeds by functional maximisation of the free energy , i . e . , the lower bound on the marginal likelihood ( 12 ) The maximisation over can be understood as the minimisation of the Kullback-Leibler divergence between the factorised and the real posteriors [42] , [83] . The key factorisation underlying the VBEM algorithm Beal2003 is the one between hidden variables and parameters ( 13 ) Given this basic factorisation , the algorithm proceeds in a way similar to Expectation Maximisation ( EM ) by iteratively inferring the hidden variable distribution given the observations and averaging over the parameters ( E-Step ) ; and the parameter distribution given the observations and averaging over the hidden variables ( M-Step ) . We needed two further factorisations to achieve a tractable algorithm: one between the distribution over basis vectors and input noise , and one between different identity variables at different times ( i . e . , ) . Note that these approximations do not completely eliminate dependencies between the factorised variables , which still influence each other through their sufficient statistics ( for example their means ) . In particular , the method is much less constraining than the commonly used approach of Maximum A Posteriori ( MAP ) estimation , where the entire posterior distribution is collapsed to a single point by taking the values of variables and parameters at the mode . Although the derivation of the learning equations requires long algebraic computations , they are derived from the VBEM setting without any noteworthy deviation , and are described in Text S2 . The input data to our model were taken from the CatCam videos [45] . Since some sections of the video contain recording defects ( block artifacts or pixel saturation ) , we selected a subset that showed minimal distortion ( labelled b0811lux in the dataset ) . Observations comprised the time-series of pixel intensities in fixed windows of size pixels . The windows were placed to cover ( without overlap ) the central region of the video . In this way we obtained a total of about 300 , 000 frames . The input data were preprocessed by removing the mean of each frame to eliminate global changes in luminance and to compensate for the camera's global gain control mechanism . The data were then reduced in dimensionality from 400 to 81 dimensions with equalised variances , using principal components analysis ( PCA ) . Due to the self-similar structure of natural images [22] , this was spatially equivalent to applying the model to patches . The resulting vectors , however , were smoother and easier to analyse , since the square shape of the pixels became less important . Moreover , starting with larger patches allowed us to capture the temporal correlations that arose during faster movements of the cat ( e . g . , fast head movements ) , which would have been impossible with small patch sizes . The variance equalisation ( common in image modelling ) helped with convergence . It is unlikely to have affected the final result as it is a linear operation for which the learning algorithm could easily compensate . This has been confirmed in a run performed without dimensionality reduction ( Text S3 ) . We initialised the model with 30 identity variables ( ) and attribute manifolds of 6 dimensions ( ) and let the algorithm learn the model size by reducing the number of active attribute dimensions by ARD hyperparameter optimisation . The mean of the basis vectors were initialised at random on the unit sphere , and the priors over the parameters were chosen to be non-informative for the input noise ( 1 pseudo-observation , ) and more informative for the dynamic parameters ( 2000 pseudo-observations ) , favouring persistent identity variables and slowly-varying attributes ( , ) . ( Although we have no reason to think that attribute variables should have different timescales , the small differences in the value of kept the model from being degenerate , in the sense that every rotation of the identity subspace would otherwise be equally optimal . ) We performed 500 VBEM iterations , at each iteration using a new batch of 60 sequences of 50 consecutive frames taken at random from the entire dataset . After 300 iterations we started learning the precision parameters , updating their values every 20 iterations . Parameters were identical for the fit to shuffled data , the only difference being that the selected frames were not consecutive in time . At the end of the VBEM iterations we compared the free energy of the original model to that of the time-shuffled model on a novel set of 50 batches of 3000 frames each , taken from the CatCam data as described above . The free energies were computed for each batch separately . We also ran one additional fit ( not shown ) to check that the results obtained for shuffled data were not strongly influenced by our choice of priors on and , for which we took with 1 pseudo-observation , and = 0 . 5 with 1 pseudo-observation . The results obtained were very close to those shown for the shuffled data . In order to compare the properties of the learnt units to those of cortical neurons we proceeded in a way similar to that reported in the experimental literature . In electrophysiological recordings one does not have access to the complete input-output function of a neuron , , or to the equivalent of our basis functions , . Typically , one computes the best linear approximation to the input-output function by spike-triggered averaging [49] , [85] . We derived the linear RFs of the attribute variables by presenting coloured noise stimuli with the same spectrum as natural images and computing the correlation between stimulus and response . In practice , this was done by doing standard white-noise reverse correlation in the PCA space . Since the dimensionality of the image patches has been equalised for variance , white-noise stimuli in the PCA space have the same spectrum as natural images when projected back to the image space . Given coloured noise data , we inferred the posterior distribution of identity and attribute variable using the VBEM algorithm , where the distribution over parameters was kept fixed to the one inferred during the learning phase ( i . e . , we only performed the E-step of the algorithm ) . The signal was reverse-correlated with the mean of the distribution over each attribute variable , ( 14 ) For visualisation and analysis , the filters were projected back in image space using the pseudoinverse of the PCA matrix . Optimal parameters for the RFs derived in this way were computed by fitting a Gabor function to them . Gabor functions are defined as ( 15 ) where ( 16 ) ( 17 ) The parameters are the amplitude , coordinates of the centre , orientation , frequency , standard deviations of the axes of the Gaussian envelope , and phase of the grating . To avoid local minima we performed multiple fits starting at 10 different orientations between 0 and and 10 different phases between 0 and , and kept the parameters with minimal mean squared error for all 100 fits . Phase differences in the RFs of attribute variables ( Fig . 4C , 7D ) were estimated by fixing the global orientation and frequency of an entire attribute manifold to the one of the best fitted RF ( minimal mean squared error ) , and re-fitting only the phase parameter to the RFs of the other attribute variables . The normalised widths and lengths reported in Figure 5 were defined as the product of the frequency of the Gabor function and the standard deviations of the axes of the Gaussian envelope , i . e . , and [56] .
|
When we look at a visual scene , neurons in our eyes “fire” short , electrical pulses in a pattern that encodes information about the visual world . This pattern passes through a series of processing stages within the brain , eventually leading to cells whose firing encodes high-level aspects of the scene , such as the identity of a visible object regardless of its position , apparent size or angle . Remarkably , features of these firing patterns , at least at the earlier stages of the pathway , can be predicted by building “efficient” codes for natural images: that is , codes based on models of the statistical properties of the environment . In this study , we have taken a first step towards extending this theoretical success to describe later stages of processing , building a model that extracts a structured representation in much the same way as does the visual system . The model describes discrete , persistent visual elements , whose appearance varies over time—a simplified version of a world built of objects that move and rotate . We show that when fit to natural image sequences , features of the “code” implied by this model match many aspects of processing in the first cortical stage of the visual system , including: the individual firing patterns of types of cells known as “simple” and “complex”; the distribution of coding properties over these cells; and even how these properties depend on the cells' physical proximity . The model thus brings us closer to understanding the functional principles behind the organisation of the visual system .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"neuroscience/natural",
"and",
"synthetic",
"vision",
"neuroscience/theoretical",
"neuroscience",
"neuroscience/sensory",
"systems"
] |
2009
|
A Structured Model of Video Reproduces Primary Visual Cortical Organisation
|
Across diverse taxa , seminal fluid proteins ( Sfps ) transferred at mating affect the reproductive success of both sexes . Such reproductive proteins often evolve under positive selection between species; because of this rapid divergence , Sfps are hypothesized to play a role in speciation by contributing to reproductive isolation between populations . In Drosophila , individual Sfps have been characterized and are known to alter male sperm competitive ability and female post-mating behavior , but a proteomic-scale view of the transferred Sfps has been missing . Here we describe a novel proteomic method that uses whole-organism isotopic labeling to detect transferred Sfps in mated female D . melanogaster . We identified 63 proteins , which were previously unknown to function in reproduction , and confirmed the transfer of dozens of predicted Sfps . Relative quantification of protein abundance revealed that several of these novel Sfps are abundant in seminal fluid . Positive selection and tandem gene duplication are the prevailing forces of Sfp evolution , and comparative proteomics with additional species revealed lineage-specific changes in seminal fluid content . We also report a proteomic-based gene discovery method that uncovered 19 previously unannotated genes in D . melanogaster . Our results demonstrate an experimental method to identify transferred proteins in any system that is amenable to isotopic labeling , and they underscore the power of combining proteomic and evolutionary analyses to shed light on the complex process of Drosophila reproduction .
In addition to sperm , males of many internally fertilizing species transfer seminal fluid proteins ( Sfps ) to their mates during copulation . These proteins function in a variety of reproductive processes , including sperm capacitation , sperm storage and competition , and fertilization , and in some cases they affect female behavior and physiology [1] . Like other reproductive proteins , Sfps often evolve rapidly between species , underscoring their relevance to reproductive success [2] . Sfps are thought to interact with several classes of molecules , including other Sfps ( which may originate from seminal fluid of the same male or from a competitor ) , proteins native to the female reproductive tract , and pathogens that may be introduced during the course of mating . These interactions create opportunities for coevolution , leading to speculation that sperm competition , sexual conflict , sexual selection , and/or host–pathogen interactions could drive the rapid , adaptive evolution of many Sfps [3] . Because of their rapid evolution and their critical importance to reproductive fitness , Sfps may also be involved in the formation of new species [3–5] . As such , researchers have sought to identify and characterize Sfps in such diverse taxa as mosquitoes , crickets , honeybees , rodents , and primates [6–10] . Although Sfps are being studied in many species , they are best characterized in Drosophila melanogaster . Because the Drosophila mating system features multiple matings by females and sperm competition between males [11–14] , Sfps are thought to be especially important for reproductive success and for mediating conflict and competition . Previous studies have focused on three areas: ( a ) the effects of the full set of Sfps ( and especially of accessory gland proteins , or Acps ) on male and female fitness; ( b ) identification of putative Acps by expressed sequence tag ( EST ) sequencing , comparative genomics , and proteomics; and ( c ) functional analysis of specific Acps . Whole-organism work revealed that Acps mediate a “cost of mating” to females . The repeated receipt of Acps through multiple matings lowers female reproductive output by reducing female life span without a corresponding increase in egg production [15–17] . Furthermore , when a population of males harboring natural genetic variation was allowed to adapt to a static female genotype , male sperm competitive ability , mating success , and harm caused to females increased in 30–40 generations [18] . These dramatic evolutionary outcomes sparked much interest in identifying the specific proteins of seminal fluid . Screens for genes expressed specifically in the male accessory glands ( and encoding proteins predicted to be secreted ) identified ∼70 putative Acps [19 , 20] , several of which have been genetically or biochemically characterized ( reviewed in [21] ) . Work in related Drosophila species has revealed that many predicted Acps are subject to lineage-specific gene gain , gene loss , and/or copy number variation [22–24] . Additionally , several proteomic studies have examined both Acps and sperm proteins found in males [25 , 26] , and whole-genome , tissue-specific microarray analysis has increased the number of predicted Acps to 112 [21 , 27] . In spite of this considerable progress , less than one-third of the predicted D . melanogaster Sfps have been detected in mated females [21 , 28] . Furthermore , prior work to predict Sfps has often required that candidate genes show tissue-specific expression in the male reproductive tract . Identifying the set of transferred Sfps in an unbiased fashion is of critical importance , since it is these proteins that are the most likely to influence post-mating processes like sperm competition and sexual conflict . We have developed a mass spectrometry ( MS ) method that specifically detects male Sfps in mated female D . melanogaster . In addition to confirming the transfer of many predicted Sfps , we identified dozens of new seminal fluid components , including completely new classes of proteins . Evolutionary analyses show that positive selection and tandem gene duplication drive the evolution of seminal fluid between species , and comparative proteomics with additional species identified lineage-specific Sfps . We also used our MS data to estimate the relative abundance of each Sfp in seminal fluid and to discover previously unannotated genes encoding additional Sfps . Taken together , our experiments illustrate the power of combining proteomics with evolutionary biology to address fundamental questions about reproduction .
To distinguish between transferred Sfps and proteins native to the female reproductive tract , we metabolically labeled female flies using a diet enriched in 15N isotopes to create an isotopically “heavy” form of the female proteins [29] . Females were reared on yeast that was grown in media enriched in 15N . After one full generation of labeling , the 15N enrichment in detected fly peptides was ∼98 atom percent excess , and no peptides from whole female flies were identified with natural abundance nitrogen isotopes . These data confirm that isotopic labeling can be readily achieved in D . melanogaster and other drosophilids ( see below ) . Therefore , we reasoned that by mating unlabeled males to labeled females and then analyzing proteins found in the female reproductive tract by MS , transferred male Sfps could be identified by those peptides that showed natural abundance isotope distributions . We chose to label females instead of males , because MS resolution is best for unlabeled peptides , and we were interested in identifying male Sfps . We performed multiple biological replicates of mating experiments with different strains of males: Canton S ( a standard lab strain ) , sons of tudor females ( spermless males ) [30] , and , as a negative control , DTA-E males ( which are spermless and do not produce main cell accessory gland proteins ) [31] . In two DTA-E experiments , we detected 11 transferred proteins , including several known to be produced outside of the accessory glands ( Table S1 ) . Six total experiments with Canton S and tudor males identified a set of 138 high-confidence Sfps ( Table 1 and Table S2 ) , using a peptide-level q-value ≤ 0 . 01 within each experiment [32] . Just over half ( 75/138 ) of the transferred Sfps were previously predicted through tissue-specific expression profiling or other experimental or comparative genomic methods [21] , but only 19 were confirmed previously to be transferred at mating . Notably , we found only five previously documented sperm proteins [25] , confirming that our protein preparation protocol effectively selected for soluble , extracellular proteins . We did not detect 49 predicted Sfps [21] . These proteins may be transferred at low levels , immediately cleaved or degraded in the female , or have certain peptide sequences or post-translational modifications that complicate detection by shotgun proteomics . Alternatively , some may not be transferred at mating . We identified 63 novel Sfps , 45 of which were found in at least two biological replicates . Many of these proteins fell into the same functional categories as the previously predicted set , including proteases , protease inhibitors , mediators of an immune response , and proteins involved in lipid metabolism ( Table 1 ) . We discovered several new classes of proteins among the transferred Sfps . Most intriguing were six members of the odorant binding protein ( Obp ) family [33] . Obps are thought to shuttle small molecules through aqueous solutions by binding them in a small , hydrophobic pocket; they are traditionally associated with the olfactory nervous system [34] . We confirmed that these Obps are transferred in seminal fluid by performing MS on protein digests from dissected accessory glands and by confirming each gene's expression in the male reproductive tract with FlyAtlas [27] . Reproductive proteins of diverse species often evolve under positive Darwinian selection , which may indicate involvement in a coevolutionary process such as sexual selection , sexual conflict , or host–pathogen recognition [2] . We used coding sequence alignments from the 12 Drosophila genomes project [35 , 36] to calculate the rates of nonsynonymous substitution ( dN ) and synonymous substitution ( dS ) for all Sfps for which an ortholog was identified ( 116 of the total 138 ) . For each Sfp , we determined the whole-gene , pairwise dN/dS ( ω ) ratio between the D . melanogaster gene and an ortholog from a closely related species ( Figure 1 ) . By this conservative test , five Sfps showed evidence of adaptive evolution ( ω > 1 ) . However , prior studies have shown that when the whole-gene pairwise ω ratio exceeds 0 . 5 , or when the nonsynonymous substitution rate ( dN ) is elevated , there are often specific sites within the protein for which adaptive evolution can be detected with more sensitive methods [7 , 37] . Therefore , we used multiple species alignments to search for specific residues under selection for all genes with pairwise ω > 0 . 5 and/or pairwise dN > 0 . 05 . ( We did not test all Sfps , in order to minimize the number of statistical tests . ) We found evidence for adaptive evolution at specific sites for 16 of 36 proteins ( Figure 1 and Table S3 ) , including four proteins that were unidentified previously as Sfps . Nine of these tests for selection remained significant after applying a strict Bonferroni correction for multiple tests . These rapidly evolving proteins are attractive targets for future study . Previous studies found that some predicted Sfps are clustered throughout the genome [21 , 24] . We examined the chromosomal locations of the transferred Sfps and found similar patterns . We defined a cluster as genes with start codons located within 10 kb of each other . We identified 19 clusters of 2–5 transferred Sfps , which contain one-third ( 46/138 ) of the detected Sfps ( Figure 2A ) . For 17 clusters , all member genes are transcribed in the same direction , and 15 clusters contain genes that encode proteins with full-length homology to one another . Thus , most of the observed clustering can be attributed to tandem gene duplication . Four paralogous clusters contain at least one gene that was under selection in the sites analysis above . Previous work also found a dearth of Acps on the X chromosome . Consistent with this finding , the 13 transferred Sfps on the X chromosome were significantly fewer than would be expected by chance ( χ2 = 4 . 68 , 1 degree of freedom [df] , p = 0 . 03 ) , given the proportion of annotated genes on the X . One example of rapidly evolving tandem duplicates is the gene pair CG17472 and CG31680 . Across five species , CG17472 has evolved adaptively , with 21 . 3% of sites predicted to be under positive selection ( estimated ω = 3 . 36 , PAML M8 versus M8a comparison: χ2 = 15 . 38 , 1 df , p < 0 . 0001 ) . These duplicates flank a transposition hot spot and a third , pseudogenized copy of the locus ( Figure 2B ) . Additionally , CG17472 has duplicated along the lineage leading to D . simulans and D . sechellia ( Figure 2C ) . Examining the ω ratio on each branch of the phylogeny reveals a burst of positive selection on the CG31680 lineage immediately after duplication . Indeed , a branch model allowing for variable selective pressures along each branch ( shown in Figure 2C ) fit the data significantly better than a model with a uniform ω for all branches ( χ2 = 29 . 04 , 14 df , p = 0 . 01 ) . We used our MS data to estimate the relative abundance of each Sfp in seminal fluid . By counting the number of spectra associated with each Sfp in a given experiment and standardizing by the length of the protein and the total number of Sfp spectra detected in the experiment , we calculated a normalized spectral abundance factor ( NSAF ) [38 , 39] for each protein , which could then be averaged across all experiments ( Figure 3 and Table S2 ) . Notably , NSAF values were positively associated with the number of biological replicates in which a protein was found ( Figure 3 ) . Several of the most abundant proteins were previously characterized Sfps , such as Acp62F ( a protease inhibitor ) and Acp70A ( the sex peptide ) . However , several novel proteins were also in the top quartile for abundance , including Obp56f , Obp56g , and the tandem duplicate CG17472 . Although these NSAF measurements are only approximate , these data provide the first proteomic-scale view of the relative amount of each transferred Sfp , which may be useful for selecting candidates for further investigation . To examine the cross-species evolution of seminal fluid content , we used the predicted protein annotations of D . simulans and D . yakuba [35 , 36] to repeat our mating experiments with a wild-type strain of each species ( Figure S1 ) . Of the 63 Sfps detected in all three species , 19 were not reported previously as seminal fluid components . For Sfps that were detected in only one or two species , we investigated whether these proteins could be called as either lineage-specific gene gain or loss events . Most of the proteins had identifiable orthologs in the other species; our failure to detect these proteins may be due to changes in expression patterns , sequence substitutions that render MS identification more difficult , changes in the amounts of proteins transferred at mating , or the lower number of replicates ( two per species ) performed for D . simulans and D . yakuba . However , our data identify 13 lineage-specific Sfps across the three species ( Table S5 ) . For example , in D . melanogaster , CG6289 ( a predicted serine protease inhibitor ) has duplicated to form the lineage-specific gene CG6663 . Also , in D . yakuba , Acp76A ( another serine protease inhibitor ) has duplicated , and several other proteins appear to be either lineage-specific to D . yakuba or rendered nonfunctional in other species ( Table S5 ) . Some proteins detected for D . simulans and D . yakuba lacked annotated orthologs in D . melanogaster . For seven such proteins , we identified the syntenic region in D . melanogaster and performed reverse-transcriptase PCR ( RT-PCR ) to determine whether transcripts of the region were made . In five cases , we detected a transcript in D . melanogaster ( see Table S5 ) , and three of these putative loci were detected as proteins in D . melanogaster when searching for unannotated proteins in the D . melanogaster genome ( see below ) . Curiously , one of these genes , which we have annotated as Sfp53D , showed male-specific expression in D . yakuba and male-biased expression in D . simulans , but no sex expression bias in D . melanogaster ( data not shown ) . Sfp53D is therefore an example of the type of protein that would have been omitted from previous sets of predicted Sfps due to its lack of sex-specific expression . Based on these results , we reasoned that other Sfps may not be annotated as genes in D . melanogaster , which would make them impossible to detect by searching mass spectra against the annotated proteome . To detect additional unannotated Sfps , we first constructed a six–reading frame translation of the D . melanogaster euchromatic genome , which produced >5 . 8 million potential open reading frames ( ORFs ) . Then , to reduce computational search time , we applied the Hardklör algorithm [40] to predict which MS2 spectra from a tudor experiment came from male peptides containing only natural abundance isotopes . These spectra were searched against the six-frame database , and those that matched an ORF corresponding to an annotated protein were discarded . This procedure left 23 novel , putative ORFs that did not match any D . melanogaster gene annotation in FlyBase . For each putative ORF , we used rapid amplification of cDNA ends ( RACE ) and RT-PCR to confirm transcription of the region encoding the peptide and to define the full-length transcript . Through this method , we discovered 19 unannotated genes ( Table S6; GenBank ( http://www . ncbi . nlm . nih . gov/Genbank/ ) accession numbers EU755332–EU755350 ) , most of which showed no significant identity to the predicted proteins of the other sequenced Drosophila species . All 19 proteins have predicted signal sequences for secretion; many are encoded by only one or two exons , and all produce short polypeptides ( median length: 93 residues ) . Consistent with our clustering analysis , half of the genes were found in regions of the genome containing other annotated Sfps . Most of the novel proteins had no recognizable domains based on BLAST and structural homology searches , but we identified one C-type lectin and three enzyme inhibitors , including a putative protease inhibitor , Sfp24Ba . This protein was identified by three peptides , one of which is indicated in Figure 4A . Sfp24Ba is adjacent to another previously unannotated gene , Sfp24Bb ( an apparent tandem duplicate ) , and lies 25 kb upstream of the gene that encodes a transferred protease inhibitor , Acp24A4 ( Figure 4B ) . Comparative structural modeling ( Figure 4C ) suggests that this protein is a Kunitz-type protease inhibitor . The discovery of these 19 new Sfp genes in a model system that has been studied for over a century and for which comparative genomic analysis is now straightforward underscores the limitations of both computational gene prediction programs and the “whole-proteome” databases that are routinely used during shotgun MS analyses .
Our study provides a proteomic-scale view of the transferred Sfps in D . melanogaster . While we confirmed that 75 predicted Sfps are truly transferred at mating , we also identified a total of 82 genes ( 63 already annotated , 19 newly discovered ) previously unknown to encode seminal fluid products . By using data from the genome sequencing projects and by performing comparative experiments in D . simulans and D . yakuba , we identified many instances of positive selection , tandem gene duplication , and lineage-specific changes in seminal fluid content between species . Taken together , our experiments demonstrate how new proteomic methods can be combined with the vast amounts of genomic sequence data that are now available to gain considerable insight into the molecular players of a specific biological process . The two methodological advances presented here—the use of isotopic labeling to distinguish between the sexes , and searching MS data against an entire translated genome—should be applicable to many taxa . For example , worms , plants , rodents , and microorganisms are all amenable to isotopic labeling [29 , 41 , 42] . In any of these systems , differential labeling should readily allow the detection of proteins transferred from one organism ( during mating or another behavior , e . g . , courtship ) . Thus , our approach allows transferred proteins in a pre-specified biological process to be identified . Furthermore , our MS- and RACE-based method to identifying novel genes should be applicable to other organisms with sequenced genomes , particularly if their genome sizes are no more than 1–2 orders of magnitude greater than the D . melanogaster genome . Rodents , Arabidopsis , and humans all fall within this range; indeed , recent work in A . thaliana has found new genes using a similar approach [43] . Our results confirm that searching MS data against an entire translated genome , rather than only an annotated or predicted proteome , can identify a considerable number of new genes . Admittedly , this process might have been particularly useful for identifying Drosophila Sfps . As shown here and in previous work , these proteins are short , rapidly evolving , and relatively free of codon bias [19 , 22] , three features making them less likely to be detected by computational gene prediction programs . Nonetheless , our gene identification method was straightforward to perform , and because it was experimentally based , it offered automatic verification for the new genes and allowed us to immediately assign them to a specific biological process: male reproduction . One striking result from these experiments is that seminal fluid content in D . melanogaster appears to be considerably more complex than was previously predicted . The Obp genes identified reproducibly and at high abundance by MS are particularly attractive targets for further characterization . One hypothesis for the origin of these reproductive Obps is based on the fact that Obps are a large , 51-member family in D . melanogaster [33] . If some members of this family were functionally redundant , selection on the regulatory and coding sequences of some Obps might have been relaxed , allowing them to be co-opted from an olfactory function into a male reproductive function . Indeed , several of the identified Obps show accessory gland–specific patterns of expression , while others are expressed in both the accessory glands and the head [27] . The function of these reproductive Obps remains to be determined; they may present odorants or pheromones to odorant receptors in the female reproductive tract or play some other role , such as transferring small molecules to the female to elicit a behavioral response . If some of the Obps interact with a receptor in the female tract , the Or10a odorant receptor is one possible target , since its expression is up-regulated in female reproductive tracts in response to the receipt of Sfps [44] . While the selective pressures driving the evolution of Sfps ( and of reproductive proteins in general ) remain unclear , the important roles of tandem gene duplication and positive selection in the evolution of Sfps are consistent with the predictions made by models of sexual selection/conflict [45] . If males are engaged in a coevolutionary chase with females , driven by sexual selection or sexual conflict , duplication of an Sfp locus could allow males to better adapt to a particular allele or paralog of a female receptor [45] . Indeed , gene duplication followed by positive selection has been observed previously in a well-characterized reproductive protein , lysin , which allows abalone sperm to penetrate the egg vitelline envelope [46] . If Drosophila Sfps are coevolving with receptors in the female reproductive tract—or with other Sfps with which they interact—then gene duplication may be an important evolutionary strategy for males to increase their reproductive success . Tests of this hypothesis will require both functional data on the newly identified male proteins and the identification of their interacting female and/or male partners [47–49] . The rapid divergence characteristic of many Sfps has generated considerable interest in their potential role in speciation [3 , 4] . If proteins mediating processes such as sperm storage , fertilization , or post-mating behavior diverged quickly between allopatric populations , driven continually by coevolutionary forces such as sexual selection or sexual conflict , between-population matings may become less productive than within-population matings . Such a difference could exert pressure to further differentiate the mating systems or mating behaviors of each group , which could eventually lead to the formation of distinct species . Determining the transferred Sfps , and subsequently identifying their functions and evolutionary patterns , could therefore be important steps in identifying potential “speciation genes . ” In conclusion , this set of transferred proteins provides a rich resource for investigating long-standing evolutionary questions and for identifying the specific molecules and functional allelic variants that affect both sperm competition and male-female coevolution and conflict . The challenge ahead will be to apply the combination of genetic , biochemical , and evolutionary methods that have already yielded many insights into Drosophila reproduction to this novel collection of transferred proteins . Functional tests of individual Sfps are essential for understanding the causes of the dramatic post-mating changes in female behaviors . For example , several studies have used gene knockouts or RNA interference to identify the post-mating effects of specific Sfps [50–55] . Other experiments have associated naturally occurring variants in several Sfps with different measures of sperm competition [12 , 56] . We expect that both of these approaches will become more effective in the future , since they can now be targeted to those transferred Sfps identified here .
Fly stocks were maintained on standard media at 25 °C , except during isotopic labeling ( see below ) . D . melanogaster stocks included a wild-type lab strain , Canton S , and the strain used for genome sequencing , y; cn bw sp . To produce spermless males , homozygous tud1 bw sp females [30] were mated to either Canton S or y; cn bw sp males , and male progeny were retained for use in mating experiments . The DTA-E stock was used to produce males lacking both sperm and main-cell accessory gland proteins [31] . D . simulans strain W89 and D . yakuba strain Tai6 were used in additional mating experiments . The isotopic labeling procedure followed a previously described method [29] , with some modifications . Wild-type Saccarhomyces cerevisiae was grown to saturation in minimal media containing 2% glucose , yeast nitrogen base without amino acids and ammonium sulfate ( Difco ) , and 15N-labeled ammonium sulfate ( ≥ 99% 15N-enrichment; Spectra Stable Isotopes ) . Yeast cells were pelletted , resuspended in a small volume of sterile water , and lyophilized . This dried yeast was then mixed with water to form a “heavy” ( 15N ) yeast paste . Flies were isotopically labeled by allowing unlabeled females to lay eggs for 24–36 h onto an agar plate topped with a small amount of heavy yeast paste . Adults were then discarded , and eggs were allowed to develop to adulthood at 25 °C in a vial capped at the open end by the plate . Heavy yeast paste was added to the plate throughout development as the sole food and nitrogen source . Virgin females were collected over CO2 within 8 h of eclosion and stored in a separate vial with 15N yeast paste on an agar plate . Shotgun MS analysis of proteins from whole , first-generation 15N flies was used to confirm isotopic labeling . In parallel to 15N labeling , males of the strain to be tested were grown in standard vials . Males were collected while young ( 0–3 d old ) and aged in isolation in a standard vial . We performed 12 total mating experiments: three biological replicates each of Canton S and tudor males , and two biological replicates each of DTA-E , D . simulans , and D . yakuba males . For each experiment , males and females were aged to 2–5 d before mating . On the day before an experiment , approximately 40 labeled , virgin females were divided into three vials containing agar with a small amount of heavy yeast paste . Unlabeled males , in a ≥1 . 5-fold excess relative to females , were placed into three standard vials . The next day , males were transferred to the female vials without anesthesia . Mating was allowed to proceed for 2 h; vials were inspected several times during this period to confirm that copulations occurred . At the end of the mating period , flies were sexed over CO2: males were discarded , while females were kept on ice and immediately dissected in 50 mM ammonium bicarbonate . The lower female reproductive tracts were retained and stored in cold ammonium bicarbonate , while ovaries were excluded to prevent saturating the protein sample with the highly abundant egg yolk proteins . ( If ovaries had been included , a greater fraction of peptides identified by MS would have arisen from these female proteins , making it more difficult to detect peptides from lower-abundance male Sfps . ) It is unlikely that the removal of the ovaries diminished our ability to detect certain Sfps , as we identified all five Sfps ( Acp26Aa , Acp36DE , Acp62F , msopa , and Spn2 ) that had been shown previously to localize to the ovaries [28] . Because we sought to identify soluble , extracellular male Sfps , proteins were prepared in such a way so as to select specifically for soluble proteins . We also sought to reduce cell lysis and thus protein content from male sperm cells and female reproductive tract epithelial cells , since releasing intracellular proteins from these cells would dilute the concentration of transferred Sfps and render their identification more difficult . Female reproductive tracts were homogenized in the ammonium bicarbonate dissection buffer , which lacks any type of detergent and thus minimized cell lysis . The mixture was then centrifuged for 5 min at 18 , 000g . This process was repeated once , and the supernatant was retained . Protein concentration was estimated using a BCA assay ( Pierce ) . Proteins were prepared for tandem mass spectrometry and digested with trypsin as previously described [57] . Two samples each of Canton S and tudor , and one sample each of DTA-E , D . simulans , and D . yakuba , were analyzed by multi-dimensional protein identification technology ( MudPIT ) [58] . Protein digests ( 50 μg ) were bomb-loaded overnight onto a tri-phasic 100-μm internal diameter capillary column packed with 15-cm reversed phase material ( Jupiter C12 , 4 μm , 90 Å; Phenomonex ) at the tip of the column , then 4 cm of strong cation exchange material ( Whatman ) , then 3 cm more of C12 material . The columns were then placed on-line with either an LTQ ion-trap mass spectrometer ( ThermoElectron ) or an LTQ-FT Ultra mass spectrometer ( ThermoElectron ) and eluted over a 12-step gradient with increasing salt concentration as described previously [59] . We also analyzed additional samples using a single reversed-phase HPLC method . One sample each of Canton S , tudor , DTA-E , D . simulans , and D . yakuba was analyzed with 75-μm internal diameter capillary columns packed with 40 cm of Jupiter C12 reversed phase material . For each sample analyzed by reversed phase , four or five technical replicates of ∼6 μg of protein were analyzed by injecting the sample directly into an on-line column and running four-hour gradients to acquire mass spectra using data-dependent acquisition . Tandem mass spectra from each RAW mass spectrometry data file were extracted from the proprietary data format and stored in the MS2 file format [60] using in-house developed software . The charge-state of multiply charged MS/MS spectra were assigned a single +2 and +3 charge state using the charge-czar program [61] and searched against two databases using Sequest [62] . One database contained the annotated proteome of the appropriate species; the other database contained a set of “decoy” proteins , made by randomly shuffling the amino acids in each protein of the annotated database . Each database also included common contaminants ( or their shuffled counterparts ) . For D . melanogaster samples , the proteome was taken from the version 4 . 3 release of the D . melanogaster genome ( downloaded from NCBI; gene annotations and names were later updated to version 5 . 2 ) . For D . simulans and D . yakuba , the GLEANR protein predictions from the 12-genome Drosophila sequencing project were used [35 , 36] . Because the GLEANR sets were likely imperfect , these species' databases were supplemented with the best hit ( e-value cutoff = 0 . 01 ) obtained when the identified D . melanogaster proteins were searched using tblastn against the D . simulans or D . yakuba sequences in GenBank . After the database searches , the percolator program [32] was used to improve the discrimination between correct and incorrect peptide spectrum matches and to assign a q-value as a measure of the false discovery rate [63] . To determine the list of high-confidence Sfps in D . melanogaster , we used the following criteria . Proteins identified in at least two independent experiments were automatically included . For proteins identified in only one of the six Canton S and tudor experiments , we required additional evidence that the protein could plausibly be involved in reproduction . This criterion could be satisfied if a protein was included in the most recent and comprehensive set of predicted Sfps [21] and/or if the protein showed strong evidence of being expressed exclusively or predominantly in the male reproductive tract ( accessory glands or testes ) in the FlyAtlas dataset [27] . Because we performed fewer mating experiments ( two per species ) and had no genome-wide catalog of Sfps or expression data , it was necessary to use different criteria in defining the sets of transferred proteins in D . simulans and D . yakuba . For each species , all proteins found in both experiments were automatically included . Furthermore , we included proteins found in only one experiment if they met any of the following criteria: ( a ) at least two peptides were used to identify the protein in the experiments; ( b ) if a single peptide was used for identification , it was detected at least twice during the MS run; or ( c ) the protein was identified as a transferred or predicted Sfp in D . melanogaster . After determining the list of transferred proteins shown in Table S2 , functional information was acquired by examining FlyBase and the primary literature and was used to classify proteins listed in Table 1 . We classified a protein as a “sperm protein” if it was found at least twice in Canton S experiments but not in tudor experiments , and/or if it was previously documented as such [25] . We used BLAST and BLAT searches to determine whether any transferred proteins of each species could be called as lineage-specific ( Table S5 ) . Genomic locations of Sfps were determined by downloading from FlyBase ( release version 5 . 2 ) the chromosomal location of the first transcribed base of each gene , and recording the strand from which the gene was transcribed . Only euchromatic genes were considered and plotted , such that plots in Figure 2A do not indicate , for example , centromeric heterochromatin . Clusters were defined as genes that were within 10 kb of each other . For proteins encoded in a given cluster , we used pairwise BLASTP searches to determine whether the proteins showed evidence of paralogy . We used simulations to estimate a null distribution of the number of clusters that would be expected for a set of 138 genes distributed across the chromosomes in the same ratio as our Sfps . We extracted coding sequence annotations from http://www . flybase . org/ ( version 5 . 2 ) and noted the location of the start codon for each gene ( one isoform per locus ) . We then generated 1 , 000 sets of 138 genes by randomly selecting genes from each chromosome arm in the same ratio as the observed Sfps . The number of clusters in each set was counted; the median was 3 , and the range was 0–9 clusters . Therefore , we judged our observed 19 clusters to be significantly more than would be expected by chance . Several GLEANR-predicted proteins identified in D . simulans and D . yakuba lacked annotated orthologs in the version 4 . 3 release of the D . melanogaster genome ( one has since been annotated as CG12828 , and another is reported [64] in GenBank under accession number BK003861 , but is not yet recorded in FlyBase ) . We thus tested whether these genes ( GLEANR numbers: dsim_2617 , dsim_3447 , dsim_15012 , dsim_10234/dyak_792 , dsim_9514/dyak_14199 , dyak_12348 , and dyak_10591 ) were expressed in D . melanogaster and showed sex-specific expression . PCR primers were designed to amplify transcripts in both the species of identification and the syntenic region of D . melanogaster . Although several of these proteins were encoded by short , single-exon genes , primers were designed to span putative introns when possible . Total RNA was prepared from whole male and whole female flies of both species using the TRIzol reagent ( Invitrogen ) and subjected to rigorous DNase treatment using the Turbo DNase kit ( Ambion ) . First-strand cDNA from each sex was synthesized using the SuperScript III kit ( Invitrogen ) according to the manufacturer's instructions . This cDNA was then diluted and used in PCR reactions . As a positive control , we assayed for transcription of ribosomal protein L32 ( RpL32 ) using previously published primers [12] , modified as needed for D . simulans and D . yakuba . Negative PCR controls were performed by using template from cDNA reactions that lacked reverse transcriptase . For each D . melanogaster protein identified , we used coding-sequence alignments generated by the 12-species genome sequencing projects [35 , 36] to conduct molecular evolutionary analyses . We preferentially used the more recent Fuzzy Reciprocal BLAST-based alignments of D . melanogaster coding sequences with orthologs in any other species ( ftp://ftp . flybase . net/genomes/12_species_analysis/clark_eisen/alignments ) ; however , less than half of our Sfps were included in this set , so for the others we used the comparative assembly freeze 1 ( CAF1 ) GeneMapper alignments produced by S . Chatterji and L . Pachter . From these sources , we were able to analyze 116 of the 138 annotated Sfps from D . melanogaster . We first made pairwise estimates of dN/dS with model M0 of codeml in the PAML package [65] . When available , we used the D . simulans ortholog; otherwise , the D . sechellia ortholog was used . Alignments were obtained from one of the above sources and checked by eye using MEGA 4 . 0 [66] . For genes with pairwise dN/dS ≥ 0 . 5 or dN ≥ 0 . 05 , we expanded our PAML analysis to up to five species ( melanogaster , simulans , sechellia , yakuba , and erecta ) in order to search for specific sites likely to have evolved under positive selection . For each gene , we used only those species for which alignments were reliable , and coding sequence alignments were checked by eye and edited in MEGA 4 . 0 . We then tested for positive selection by comparing the likelihoods of codeml models M8 and M8a with a likelihood ratio test [67] . In model M8a , each codon is assigned to one of 11 classes , ten of which have an ω ( dN/dS ) value between 0 and 1 that is estimated from the data using maximum likelihood , and the 11th of which has ω = 1 , representative of neutral evolution . Model M8 differs in that the 11th class of codons can take any ω value; this value is estimated from the data and can be greater than 1 ( which indicates adaptive evolution ) . We corrected for multiple testing with a strict Bonferroni correction , though we note that among the 36 tests performed , only ∼2 would be expected to be false positives at a critical p-value of 0 . 05 . As shown in Figure 2B and 2C , we re-analyzed CG17472 and CG31680 by including both paralogs from D . simulans and D . sechellia . The M8 versus M8a test of CG17472 in the results section contains all orthologs ( including both duplicate copies in D . simulans and D . sechellia ) , but not CG31680 and its D . sechellia ortholog . For Figure 2C , we used the dnaml program in PHYLIP [68] to construct a phylogeny and to simulate 1 , 000 bootstrap replicates . We then used PAML to estimate ω for each branch of the phylogeny and performed a likelihood ratio test to compare the likelihoods of a model that allowed for ω to vary on each branch of the phylogeny versus a null model in which a uniform ω was estimated across all branches [69] . Relative protein abundance was estimated from D . melanogaster MS data by counting the number of spectra that positively identified each protein in a given MS run . This spectral count was normalized for the length of each protein and divided by the sum of all normalized counts for the entire MS run to produce an NSAF for each protein , as previously described [38 , 39] . This value was then averaged across all experiments in which a protein was detected , and identified proteins were ranked by their mean NSAFs . This rank should be interpreted as how common it was to identify ionizable and detectable spectra for a given protein , relative to the other unlabeled proteins . To identify unannotated Sfps , we first used nr6frame ( D . States , unpublished program ) to make a six–reading-frame translation of the entire Berkeley Drosophila Genome Project D . melanogaster genome , version 5 ( downloaded from ftp://hgdownload . cse . ucsc . edu/goldenPath/dm3/chromosomes ) . This program translates genomic DNA in all six reading frames; each reported ORF ends with a stop codon ( but does not necessarily start with a methionine ) . Across the four Drosophila chromosomes , over 7 . 6 million ORFs were generated . We filtered these ORFs to exclude those that contained only one type of amino acid ( mono-residue repeats ) , those that were too short to be confidently used in MS spectrum identification ( <11 residues ) , or those that could not produce a tryptic peptide due to a lack of a K or R residue . This filtering reduced the data set to >5 . 8 million ORFs . For searching this large database , it was computationally advantageous to filter the MS2 files in order to reduce the search time . We used data from three technical replicates of a tudor sample , collected with a 40-cm reversed phase column on an LTQ-FT Ultra instrument . We used Hardklör [40] to predict the isotope distributions that resulted from 15N-enriched peptides and removed their corresponding MS/MS spectra from the analysis . Because of the excess of labeled peptides within the sample , this filtering reduced the number of spectra that needed to be searched by ∼86% . The remaining spectra were then searched against the six-frame translation database using Sequest [62] , and identifications were filtered by DTASelect . Identified peptides matching annotated protein coding genes were discarded , leaving 23 ORFs that did not match a genome annotation . We designed primers in the genomic regions matching the identified peptides and performed 5′ and 3′ RACE to amplify transcripts from these regions ( SMART RACE Kit , Clontech-Takara ) . This method identified 19 unannotated genes , which were then confirmed with RT-PCR and sequencing of cDNA from whole males . SignalP was used to predict whether each novel protein is secreted [70] , and we used BLAST and PHYRE [71] searches to determine whether any protein had sequence or structural homology to other proteins .
Sequence data has been deposited in GenBank under accession numbers EU755332–EU755350 .
|
Across many species , males transfer both sperm and seminal proteins to their mates . These proteins increase male reproductive success by improving sperm competitive ability and modifying female behavior . In Drosophila , seminal proteins increase female rates of egg-laying and sperm storage and reduce a female's willingness to mate with subsequent suitors . Several male seminal proteins have been extensively characterized , and others have been predicted based on gene expression patterns , yet the full set of proteins that is transferred to females has not been defined . Here we introduce a new proteomic method that identifies transferred seminal proteins in recently mated females and quantifies their relative abundance . We confirm many of the predicted seminal proteins and discover a number of novel seminal fluid components . Some of these proteins show elevated rates of evolution , consistent with their involvement in sexual selection or sexual conflict , and many have arisen by tandem gene duplication . By using this method in three species of Drosophila , we identified lineage-specific components of seminal fluid . Additionally , we developed and validated a method to identify completely new genes in the D . melanogaster genome . These transferred proteins are now targets for follow-up genetic , biochemical , and evolutionary analysis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"evolutionary",
"biology",
"molecular",
"biology",
"genetics",
"and",
"genomics"
] |
2008
|
Proteomics Reveals Novel Drosophila Seminal Fluid Proteins Transferred at Mating
|
Bacterial growth in multicellular communities , or biofilms , offers many potential advantages over single-cell growth , including resistance to antimicrobial factors . Here we describe the interaction between the biofilm-promoting components curli fimbriae and cellulose of uropathogenic E . coli and the endogenous antimicrobial defense in the urinary tract . We also demonstrate the impact of this interplay on the pathogenesis of urinary tract infections . Our results suggest that curli and cellulose exhibit differential and complementary functions . Both of these biofilm components were expressed by a high proportion of clinical E . coli isolates . Curli promoted adherence to epithelial cells and resistance against the human antimicrobial peptide LL-37 , but also increased the induction of the proinflammatory cytokine IL-8 . Cellulose production , on the other hand , reduced immune induction and hence delayed bacterial elimination from the kidneys . Interestingly , LL-37 inhibited curli formation by preventing the polymerization of the major curli subunit , CsgA . Thus , even relatively low concentrations of LL-37 inhibited curli-mediated biofilm formation in vitro . Taken together , our data demonstrate that biofilm components are involved in the pathogenesis of urinary tract infections by E . coli and can be a target of local immune defense mechanisms .
It has been recognized that bacteria in their natural milieu seldom grow as non-differentiated , single cell organisms . Instead , they form multicellular communities , biofilms , showing coordinated behavior [1] . Classically , biofilm formation includes surface adherence , cell-cell interactions , and production of extracellular matrix [2] . The extracellular matrix contributes to the development of higher-ordered three-dimensional structures that offer advantages to the bacteria , such as increased resistance to antimicrobial substances , mechanical forces and to nutrient depletion [3]–[5] . During urinary tract infections ( UTI ) , the role of bacterial biofilms has previously been established in the presence of indwelling catheters [6] . However , uropathogenic E . coli also forms biofilm-like structures on and inside host cells in the absence of a foreign body [7]–[9] , and the ability to form biofilms has been related to persistence of bacteria in the urinary tract [10] . Curli belong to a class of fibers known as amyloids [11] and are involved in adhesion to surfaces , cell aggregation and , finally , biofilm development . Functionally and genetically , curli are linked to cellulose [12] , another extracellular matrix component of biofilms formed by bacteria from the family Enterobacteriaceae . Bacterial cellulose has mostly been investigated in soil bacteria of the family Rhizobiaceae , where this polysaccharide is required for the firm adherence and aggregation of bacteria at the root hair tip of plants [13] . Although the production of cellulose is common among many bacterial species , its biological function and role during infection is not entirely clear . When cellulose is expressed together with curli , the two substances produce a highly inert , hydrophobic extracellular matrix around the bacteria [14] . Biofilms built from curli and cellulose have widely been investigated on abiotic surfaces [15] , [16] and in commensal intestinal E . coli isolates [17] , [18] . Less information is available about the role of curli and cellulose during E . coli infection of the urinary tract [10] , [19] . Recently , we demonstrated that epithelial cells of the urinary tract up-regulate the production of the human antimicrobial peptide LL-37 upon infection with uropathogenic E . coli [20] . Thus , the cathelicidin LL-37 plays an important role in the protection against infections of the urinary tract . The proform of LL-37 , hCAP-18 , is mainly produced by epithelial cells and neutrophils [21] , [22] . After processing [23] , the active LL-37 peptide is released and exhibits its bactericidal activity by interaction with the bacterial cell membrane [24] . In the current project , we sought to study the presence of curli and cellulose in E . coli isolated from uncomplicated community-acquired UTI and their impact on early UTI pathogenesis . In addition , we here investigate the influence of LL-37 on curli-mediated biofilm formation in E . coli . We suggest that curli and cellulose protect the bacterium from immune defense mechanisms and in addition modulate the immune response of the host . We furthermore demonstrate an interaction of curli and LL-37 , especially that LL-37 inhibits the polymerization of CsgA , the major subunit of curli .
A total of 99 E . coli isolates were collected from urine of patients with UTI and 77 isolates were obtained from fecal samples of healthy individuals . Each isolate was assessed for biofilm formation using a standard microtiter assay ( see Materials and Methods ) . On average , uropathogenic bacteria adhered significantly better and formed more biofilm as compared to fecal isolates ( P<0 . 0001 , Figure 1A ) . Curli and cellulose production by all isolates was monitored on Congo red and Calcofluor containing plates . To better mimic the host environment , we chose to analyze bacteria grown at 37°C . Based on the uptake of Congo red and fluorescence after the exposure of Calcofluor plates to UV light , we could identify that approximately half of the uropathogenic and commensal E . coli isolates expressed curli ( 54% and 45% , respectively ) , 30% of the uropathogens and 16% of the commensals expressed curli and cellulose together . This morphotype was significantly associated to uropathogenic E . coli ( P = 0 . 032 ) . The expression of cellulose alone was rarely detected in either collection ( 5% and 10% , respectively ) . Nearly all isolates were positive for expression of type 1 fimbriae , irrespective of their origin ( 99% of the uropathogenic and 92% of the fecal isolates ) . To confirm the expression of curli in vivo , we collected fresh urine samples from patients with community-acquired E . coli UTI . Bacteria were analyzed directly from the urine by dot blot analysis and electron microscopy . Ten of seventeen investigated clinical E . coli isolates ( 59% ) reacted with antibodies towards CsgA . This was in line with their Congo red/Calcofluor phenotype and the overall prevalence of curli in uropathogenic E . coli investigated here ( 54% ) . Bundles of curli expressed during UTI were visualized by electron microscopy and their identity was confirmed by gold-labeled antibodies ( Figure 1B ) . The relevance of curli and cellulose expression on bacterial adhesion and immune induction in target cells was investigated by the interaction of bacteria with human cells in vitro ( Figure 2 ) . Bladder ( UROtsa , T24 ) and renal ( A498 ) epithelial cells were infected with the uropathogenic E . coli strain No . 12 , producing curli and cellulose; and its isogenic mutants lacking curli and/or cellulose . The wild-type strain and its mutants also expressed type 1 fimbriae to similar extent . The total number of bacteria after 30 min of cell infection was determined . Curli expression resulted in an increased number of cell-associated bacteria in the presence or absence of cellulose ( P<0 . 0001 , Figure 2A ) . Likewise , levels of IL-8 were significantly higher in supernatants of cells infected with curliated E . coli compared to those induced by the respective non-curliated strain ( P = 0 . 001 and P<0 . 0001 for cellulose-expressing and lacking strains , respectively , Figure 2B ) . Cellulose , on the other hand , reduced the ability of bacteria to adhere ( P<0 . 0001 and P = 0 . 001 in the presence and absence of curli , respectively , Figure 2A ) . In curliated bacteria , cellulose expression significantly reduced the induction of IL-8 ( P = 0 . 001 , Figure 2B ) . The role of curli and cellulose on adherence and IL-8 induction was confirmed by complementation of the curli and cellulose-deficient mutants which restored the wild-type phenotype ( Figure 2C+D ) . To confirm the role of curli and cellulose during the initial infection steps , mice were infected with the isogenic E . coli strains . After 1 h of infection , the expression of curli increased the number of bacteria significantly only in the absence of cellulose ( P = 0 . 026 , Figure 2E ) , whereas the comparison between the wild-type strain and the curli-lacking mutant was not significant . However , comparing the pair of curliated strains with the pair of non-curliated mutants , the effect of curli on adherence in vitro was supported ( P = 0 . 007 ) . Similar to the cell culture experiments , the expression of curli alone increased the induction of MIP-2 ( P = 0 . 001 , Figure 2F ) . Interestingly , the inhibitory effect of cellulose was even more pronounced in vivo , and was also observed in the absence of curli ( P<0 . 0001 and P = 0 . 001 in the presence and absence of curli , respectively , Figure 2F ) . Moreover , none of the strains expressing cellulose induced MIP-2 levels significantly higher than levels in control mice inoculated with sterile PBS ( 32–52 pg/ml ) . In the initial stages of UTI , curli promoted colonization ( Figure 2A+E ) . We further investigated the later course of UTI . Mice were infected with isogenic strains expressing curli and/or cellulose , and kidneys were analyzed 48 h post infection ( p . i . ) . MIP-2 is the major neutrophil chemoattractant in the urinary tract [25] . Corresponding to immune induction ( Figure 2F ) , the curliated mutant was more efficiently eliminated after 48 h p . i . than the wild-type strain with cellulose ( P = 0 . 011 , Figure 3A ) . To further investigate the role of cellulose in this process , we induced neutropenia in mice prior to infection . Neutrophil-depleted and control mice were infected with curliated bacteria with or without cellulose . Clearance by neutrophils was more efficient for bacteria lacking cellulose ( Figure 3B ) . In neutrophil-depleted mice , the number of cellulose-deficient bacteria after 48 h was as high as those of the wild-type strain . To understand the mechanism underlying the more efficient infection by curliated bacteria , we specifically investigated the antimicrobial activity of bladder and renal epithelial cells on adhered bacteria . For this purpose , bacteria were coincubated with cells for 30 min and adherent bacteria were then subjected to a staining procedure allowing the discrimination between live and dead bacteria . Dependent on the expression of curli and cellulose , 19% to 67% of cell-associated bacteria were killed . Curli but also cellulose protected bacteria from antimicrobial activities of the cells ( Figure 4A , P<0 . 001 for curli and P = 0 . 024 and P = 0 . 003 for cellulose , respectively ) , most efficiently when both structures were expressed together . Bladder and renal epithelial cells are known to produce cathelicidins in response to E . coli infection , in particular LL-37 in humans and mCRAMP in mice , respectively [20] . To relate the observed antimicrobial activity of uroepithelial cells to this peptide , bacteria were exposed to conditioned medium from cells stimulated with phenylbutyrate to enhance LL-37 production [26] . After 30 min of incubation , the number of curli-producing bacteria was almost unchanged ( 99% and 102% of the inoculated concentration ) , whereas bacteria lacking curli were reduced to 93% and 85% in the presence or absence of cellulose , respectively ( Figure 4B ) . The most pronounced difference was observed due to curli in the absence of cellulose ( 102% versus 85% , P = 0 . 006 ) . Hence , we chose these two mutant strains for neutralizing experiments . Prior to inoculation , the activity of LL-37 in the culture medium was inhibited by neutralizing antibodies . While the number of viable curliated bacteria did not differ after 30 min ( Figure 4C , left ) , the number of bacteria lacking curli was significantly higher in the presence of LL-37-specific antibodies compared to the samples treated with an irrelevant isotype control antibody ( Figure 4C , right , P = 0 . 047 ) . We further investigated the influence of curli expression on bacterial sensitivity to LL-37 and mCRAMP more specifically by a broth dilution method . When bacteria were initially grown in biofilm , the concentration of LL-37 at which bacterial growth was inhibited to 50% ( IC50 ) was 12 µM for the curliated strains . However , the IC50 for the non-curliated strains was only 6–7 µM ( Figure 4D ) . At 10 µM LL-37 , the relative growth of the curliated strains was significantly higher than growth of the non-curliated strains ( P = 0 . 001 and P<0 . 001 in the presence and absence of cellulose , respectively , Figure 4E ) . Similar results were obtained for the mouse cathelicidin mCRAMP ( Figure 4F+G ) . The IC50 value was higher for the curliated strains than for the non-curliated strains ( 7 versus 4 µM , Figure 4F ) . At 5 µM mCRAMP , the relative growth differed significantly between the curliated and the non-curliated strains ( P = 0 . 034 and P = 0 . 037 in the presence and absence of cellulose , respectively , Figure 4G ) . The same bacteria were then pre-grown planktonically , where curli expression is suppressed . When grown under such conditions , no significant difference in the resistance against both LL-37 and mCRAMP was observed between the strains ( data not shown ) . These data indicate that curli is a biofilm component that counteracts the bactericidal effect of cathelicidins and may contribute to the increased resistance of E . coli growing in biofilm . In contrast to curli , cellulose did not influence the IC50 of cathelicidins ( Figure 4D–G ) . In order to elucidate one possible mechanism that could influence the increased resistance of curliated bacteria against LL-37 , the binding of LL-37 to wild-type curli and recombinant CsgA was assessed . A precipitation assay showed a pronounced decrease of LL-37 in supernatants from samples containing wild-type curli or polymerized CsgA ( Figure 5A ) . Further , LL-37 binding to both monomeric and polymeric CsgA was demonstrated by surface plasmon resonance ( Figure 5B ) . By comparing the response during loading of the peptide in a time frame of 0–180 s , the sensogram of LL-37 demonstrated a higher association with CsgA than the control peptides , i . e . scrambled LL-37 ( sLL-37 ) and the vasoactive intestinal peptide ( VIP ) [27] . Furthermore , the binding curves reveal that the control peptides had faster dissociation rates than LL-37 , indicating a weaker binding to CsgA . This was especially pronounced for the binding to polymeric CsgA . Determination of binding constants was precluded , since LL-37 and in particular CsgA forms oligomers and polymers , respectively , and thus generate several different assemblies . At concentrations below the IC50 for bacterial growth , LL-37 inhibited curli-mediated biofilm formation with a reduction of more than 80% at 2 . 5 µM for both the wild-type ( data not shown ) and the cellulose-negative E . coli strain ( Figure 6 ) . To investigate the specificity for the inhibitory capacity , sLL-37 and VIP were analyzed in the same assay . Our results showed that the same concentration of these peptides reduced biofilm formation by only 10% , which is a significantly lower reduction than the effect of LL-37 ( P = 0 . 001 , Figure 6B ) . This indicates a sequence-specific inhibition of curli-mediated biofilm by LL-37 . To explain a possible cause for the inhibition of biofilm formation by LL-37 , the effect of LL-37 on curli formation was investigated . For this purpose , we utilized monomeric recombinant CsgA , the major subunit of curli , which spontaneously polymerizes [28] . CsgA polymerization was monitored with thioflavin T ( ThT ) , a fluorescent dye that binds to polymerized , but not to monomeric CsgA . Our results demonstrated that CsgA polymerization started immediately after incubation at 37°C and reached a stationary phase after approximately 300 min ( red line in Figure 7A ) . After prolonged incubation , the fluorescence declined , most likely due to degradation of ThT and/or precipitation of fibers [28] . The polymerization was inhibited by LL-37 in a dose-dependent manner , and at a molar ratio of 1∶1 ( CsgA∶LL-37 ) fiber formation was completely inhibited ( Figure 7A , left panel ) . The control peptides sLL-37 and VIP had a less pronounced effect on CsgA polymerization than LL-37 ( Figure 7A , right panel ) . Similar results could be achieved using confocal microscopy . After 20 h incubation of CsgA alone , we could clearly detect fiber structures stained with ThT ( Figure 7B ) or Congo red ( data not shown ) . In line with the results described above , these fibers were not detected when LL-37 was present in a molar ratio of 1∶1 , giving a fluorescence of only 0 . 05 arbitrary units compared to CsgA alone ( 1 arbitrary unit , Figure 7C ) . The control peptides sLL-37 and VIP reduced the fluorescence intensity to approximately 0 . 2 arbitrary units , suggesting a lower inhibitory capacity than LL-37 . Thus , inhibition of CsgA fiber formation was evidently stronger for LL-37 ( P = 0 . 007 and P = 0 . 011 versus sLL-37 and VIP , respectively , Figure 7C ) . To confirm the inhibition of CsgA polymerization we sought to analyse the stability of the CsgA monomer in the absence and presence of LL-37 . Freshly purified , monomeric CsgA ( 10 µM ) was incubated for 20 h at 37°C without or with different concentrations of LL-37 and was subsequently separated by SDS-PAGE . After staining with Coomassie Blue , bands corresponding to LL-37 and/or CsgA in monomeric , dimeric or tetrameric form were visualized . When CsgA was incubated alone , monomers were not visible although sometimes dimers and/or tetramers could be observed . This finding indicates spontaneous formation of CsgA oligomers and/or larger polymers that can not migrate into the gel due to their size . However , in the presence of LL-37 , a band migrating at 15 kDa , the predicted size of monomeric CsgA , could be observed ( Figure 8A ) . This was already seen at a molar ratio of 16∶1 ( CsgA∶LL-37 ) . To exclude degradation of CsgA as a possible explanation for the lack of the gel band , polymerized CsgA was treated with 90% formic acid , dissolving polymeric CsgA into monomers . After this treatment , a band corresponding to monomeric CsgA was detectable in the gel ( Figure 8A ) . The impact of LL-37 on the structure of curli was investigated with circular dichroism ( CD ) spectroscopy . In line with previous findings , polymeric CsgA exhibited a beta-sheet conformation with a minimum around 218–220 nm ( Figure 8B ) [28] . Furthermore , the low signal amplitude indicates a decreased solubility due to polymerization ( Figure 8B ) . In contrast , CsgA together with LL-37 displayed a random coil structure as has been described for monomeric CsgA [28] . This result indicates that LL-37 is able to stabilize CsgA in an unstructured form .
In the present study , we show that the majority of uropathogenic E . coli from uncomplicated community-acquired UTI adheres stronger and produces more biofilm compared to commensal bacteria . Two major extracellular components in E . coli biofilm are curli and cellulose . We here sought to explore their role in the course of UTI and their interaction with the human antimicrobial peptide LL-37 . During early stages of UTI , curli promote colonization and immune induction . Cellulose in contrast reduces MIP-2 induction , followed by impaired bacterial eradication by neutrophils . The antimicrobial peptide LL-37 produced by uroepithelial cells and neutrophils in the urinary tract interacts with curli-mediated biofilms . Curli bind LL-37 , and thus protects the bacterial cell against the bactericidal activity of LL-37 . On the other hand , by binding to CsgA monomers and likely also shorter oligomers , LL-37 inhibits CsgA polymerization and curli formation . We here for the first time provide evidence that curli are present on E . coli in fresh urine of infected patients that are not catheterized ( Figure 1B ) . The expression of curli or cellulose was equally common among E . coli isolates from UTI and commensal fecal isolates . However , the combined expression of curli and cellulose was the most common phenotype among uropathogenic isolates whereas most of the fecal isolates expressed only curli . In the pathogenesis of infection , curli fimbriae have previously been implicated in the attachment and invasion of host cells , interaction with host proteins and activation of the immune system [11] , [29] , [30] . Cytokine induction by Salmonella has been associated with binding of CsgA to toll-like receptor 2 [31] , which is expressed on bladder and renal epithelial cells [32] . In our clinical samples , urinary IL-8 levels did not correspond to curli expression , and did not differ between isolates expressing different biofilm morphotypes ( data not shown ) . However , virtually all tested clinical UTI isolates expressed type 1 fimbriae . Since type 1 fimbriae and other bacterial factors such as lipopolysaccharides are potent inducers of IL-8 , the impact of curli on IL-8 induction was possibly masked [32] , [33] . In addition , it can not be ruled out that the lag time between onset of symptoms and the first visit to the hospital , when urine samples were obtained , also influenced the results . However , we did observe a clear correlation between curli expression and IL-8 induction in bladder and renal epithelial cells ( Figure 2B ) as well as MIP-2 in mice infected with the isogenic strains ( Figure 2F ) . Interestingly , curli-dependent IL-8 induction was also observed in A498 kidney cells , which have been found to lack toll-like receptor 2 [32] , [34] . It has been reported for this cell line , that IL-8 induction is probably increased by type 1 fimbriae-mediated attachment [32] , [35] . We speculate that adherence enhanced by curli could similarly increase the immune induction in A498 cells in our experiments . Another explanation would be that there is an alternative route not yet identified mediating the immune response . Recruitment of neutrophils is mediated by IL-8 and MIP-2 in humans and mice , respectively [33] , [36] . The crucial function of MIP-2 and neutrophils in the defense of the urinary tract [25] is illustrated here by less efficient elimination of the curliated , highly immunogenic cellulose mutant strain in neutrophil-depleted mice ( Figure 3B ) . In contrast , clearance of the wild-type strain expressing cellulose is not significantly influenced by the lack of neutrophils ( Figure 3B ) , consistent with low levels of MIP-2 detected in wild-type infected mice ( Figure 2F ) . It is well known that in wound healing bacterial cellulose itself does not induce inflammation [37] . The role of bacterial cellulose in the pathogenesis of infections , however , has previously not been established . Our results suggest a protective role against the immune system . Cellulose might mask bacterial surface structures , hence avoiding immune recognition and cytokine induction , or alternatively , actively decrease the immune response . In the cell culture model , we see a significant reduction of IL-8 after infection with the wild-type strain compared to the mutant expressing only curli , whereas there is no reduction due to cellulose in the absence of curli . It is reasonable to believe that cellulose might be able to cover the relatively short curli fibers but not longer structures such as type 1 fimbriae , which are expressed by all four strains utilized in this study . In the mouse model , we see a reduced MIP-2 induction in the presence of cellulose also in the absence of curli ( Figure 2F ) . Moreover , the inhibitory effect of cellulose on MIP-2 induction appears to be stronger . Thus , the production of cellulose might be an efficient protection for bacteria not only against environmental conditions but also against immune defense mechanisms in vivo . We have recently shown that LL-37 plays a crucial role in urinary tract innate immune defense [20] . Here we demonstrate increased resistance of curliated bacteria towards the antimicrobial properties of uroepithelial cells ( Figure 4A–C ) , which is at least partly based on increased resistance against LL-37 ( Figure 4C–E ) . We confirmed the relevance of this observation for the mouse infection model by investigating the susceptibility of wild-type and mutant E . coli against mCRAMP , the murine LL-37 ortholog . Our results revealed that the curliated strains are more resistant also against the mouse cathelicidin ( Figure 4F+G ) , indicating a similar interaction with curli as demonstrated for LL-37 . Curli fibers are mainly composed of polymerized CsgA . By precipitation of LL-37 in the presence of wild-type curli or recombinant polymeric CsgA , we demonstrate binding between the peptide and the protein ( Figure 5A ) . Antimicrobial peptides including LL-37 kill their target cells by a peptide-bacterial membrane interaction that leads to lysis of the bacterium [24] . Our finding suggests that LL-37 might be trapped in the net of curli covering the bacterial surface . This prevents LL-37 from reaching the bacterial membrane and lysing the cell . In contrast , bacteria without curli lack this protection and are more easily killed during adherence and invasion into the uroepithelium . A similar protective role against cationic antimicrobial peptides has been observed for the biofilm components alginate in Pseudomonas aeruginosa [38] and the polysaccharide intercellular adhesion ( PIA ) in Staphylococcus epidermidis [39] . We also observed partial protection mediated by cellulose , the polysaccharide component in E . coli biofilm ( Figure 4A+B ) . However , this effect was not as pronounced as the protection by curli and could not be related to LL-37 , since cellulose production did not influence bacterial susceptibility to the cathelicidins LL-37 or mCRAMP in vitro ( Figure 4D–G ) . Interestingly , we demonstrate that LL-37 inhibits the formation of curli-promoted biofilm formation in vitro ( Figure 6 ) . We also show that LL-37 prevents CsgA polymerization ( Figures 7+8 ) , and speculate that LL-37 inhibits biofilm establishment by direct interference with CsgA assembly . The binding of LL-37 to both monomeric and polymeric CsgA might block reactive surfaces that are crucial for the CsgA-CsgA interaction [40] . There are five segments in the CsgA amino acid sequence that are conserved and share similarity to each other . They are characterized by conserved Ser , Gln and Asn residues [11] , [40] , but these repeats also contain acidic residues that may contribute to an electrostatic interaction with the cationic peptide LL-37 . Based on the general structures of amyloids [41] , each of these repeats is predicted to form a strand-loop-strand motif in a strong hydrogen bonding network [11] , which might be prohibited by the binding of LL-37 . Considering the function of curli during infection , the prevention of curli generation would provide an effective host defense mechanism . Our in vivo results demonstrate , despite the initial advantage of curliated bacteria , that they are eradicated more efficiently at later stages of infection . The expression of many virulence factors is highly regulated by environmental conditions , and this has also been shown for curli . Curli are maximally expressed in stationary phase and participate in the initial stage of biofilm , i . e . irreversible attachment , whereas expression might be down-regulated during biofilm maturation correlating to later stages of infection [11] , [42] . Reduced curli expression by bacteria colonizing the kidney makes them more vulnerable towards LL-37 . In addition to increased bacterial sensitivity , incoming neutrophils release high amounts of LL-37 that contributes to the antibacterial defense . In growing bacteria , the generation of new curli fibers might be inhibited by LL-37 , reducing both protection and ability to colonize the host tissue . Remarkably , biofilm-inhibitory concentrations of LL-37 were much lower than bactericidal concentrations and within a range which can be present in vivo [20] , [43] . In contrast , subinhibitory concentrations of exogenous antimicrobial drugs , e . g . aminoglycosides , seem to stimulate bacteria to produce biofilm [44] . Moreover , bacteria grown in biofilm are less susceptible to most exogenous antimicrobial agents [45] . Biofilm inhibition by antimicrobial polypeptides has previously been described for Pseudomonas aeruginosa . Both LL-37 [46] and lactoferrin [47] increased bacterial surface motility mediated by type IV pili . A direct interaction with biofilm components was not investigated in these studies . The effect was rather related to an influence of LL-37 on the bacterial gene expression profile [46] or an influence of lactoferrin on free iron [47] . These and our findings stress an important anti-biofilm role of antimicrobial polypeptides in host defense . It is likely that the anti-biofilm activity is a general strategy for these host defense molecules to keep potential pathogenic bacteria more vulnerable to killing in various tissues , including the urinary tract . In conclusion , we demonstrate that uropathogenic E . coli by expressing curli are able to modulate the immune response and display increased virulence . Cellulose , on the other hand , may reduce adherence and immunogenicity by masking bacterial surface structures , thereby evading the immune system . We also show that defense mechanisms in the urinary tract interfere with these biofilm components; curli protect the bacteria from being killed by LL-37 , in contrast LL-37 is inhibiting the formation of curli fibers . This inhibition might be an important host defense mechanism in the protection against UTI .
The studies have been approved by the ethics committee of the Karolinska University Hospital , and written informed consent has been obtained from the patients and parents of the children , respectively , in accordance with the ethics permission . The clinical study included 98 patients with UTI; 36 children [20] and 62 adult women [48] . One woman suffered from two episodes of UTI with different E . coli isolates . The diagnostic criterion of acute UTI was the presence of ≥105 CFU of E . coli per ml of freshly voided urine . Except for bacteriuria , the diagnostic criteria of acute pyelonephritis were: body temperature ≥38°C and laboratory signs of systemic inflammation , either C-reactive protein ≥20 mg/liter or erythrocyte sedimentation rate ≥20 mm/h , respectively . In addition , fecal commensal E . coli isolates were collected from 77 adults in connection with routine outpatient health examination . None of them had a history of symptomatic UTI or recent gastrointestinal disease , and their urine did not yield E . coli on cultivation [48] . Two human bladder epithelial cell lines and one human renal epithelial cell line were utilized . Virus-immortalized bladder epithelial cells UROtsa were kindly provided by Prof . Scott Garrett , Department of Pathology , University of North Dakota and cultured as described previously [20] , [49] . Bladder epithelial cells T24 ( HTB-4; American Type Culture Collection ( ATCC ) , Rockville , MD , USA ) were cultured in McCoy's 5A medium containing glutamine ( Invitrogen Life Technologies , Carlsbad , CA , USA ) supplemented with 10% FBS ( Invitrogen ) . Human renal epithelial cells A498 ( HTB-44; ATCC ) were cultured as described before [20] . For further investigation , E . coli isolate No . 12 from a child with pyelonephritis was chosen . This was a typical isolate expressing both curli and cellulose as well as type 1 fimbriae and yielding an approximately median level of biofilm as measured on microtiter plates . One-step knockout of bcsA and csgBA was carried out according to the protocol of Datsenko and Wanner with modifications [17] , [50] . The following mutants were constructed using oligos listed in Table 1; WE1 bcsA::Cm , deficient in cellulose production; WE11 csgBA::Cm , deficient in curli production; and WE16 csgBA bscA::Cm , deficient in both cellulose and curli production . Expression of curli and cellulose in strain No . 12 was confirmed by these knockouts . Production of type 1 fimbriae under the culture conditions used for experiments was confirmed by yeast agglutination ( see below ) . For relevant control experiments , complementation equivalents for the mutants WE1 and WE11 were constructed . The complementation of strain WE11 was achieved as follows: First , the chloramphenicol cassette of the curli-deficient strain WE11 csgBA::Cm was removed by Flp-catalyzed excision as described elsewhere [51]; resulting in the Cm-sensitive , curli synthesis-deficient strain WE11_1 . The removal was confirmed by PCR using oligos Csg fw and Csg rev ( Table 1 ) . The DNA region comprising the csgBA operon was amplified from strain No . 12 using the above mentioned oligos and cloned into vector pWSK29 [52] . Finally , the obtained plasmid pWSK29-csgBA was transferred into strain WE11_1 . The ability of strain WE11_1 containing pWSK29-csgBA to produce curli was demonstrated by morphotype assessment on Congo red plates . The complementation of the cellulose production-deficient strain WE1 proved to be more complex , since plasmid-based complementation approaches failed . Therefore , strain B23 was constructed which originates from wild-type strain No . 12 and carries an inducible promoter upstream of the bcs operon . In short , the previously described kmRExTET cassette [53] which contains the anhydrotetracycline ( aTc ) inducible tetA promoter was amplified using oligos kmRExTET-bcs fw and kmRExTET-bcs rev ( Table 1 ) and inserted upstream of the bcs operon using the protocol of Datsenko and Wanner with modifications [50] . Insertion of the kmRExTET cassette was confirmed by PCR using oligos bcsE control fw and yhjQ control rev ( Table 1 ) . In the resulting strain B23 , cellulose production became an aTc-dependent occurrence due to the insertion of the kmRExTET cassette as previously communicated [54] . In the absence of aTc , the morphotype of strain B23 is consistent to the morphotype of strain WE1 bcsA::Cm . In presence of aTc , cellulose production in strain B23 is restored to comparable levels than in the wild-type strain No . 12 as judge on Congo red plates . Thus strain B23 grown under inducing conditions can be used as a complemented equivalent . All assays were performed using both bladder ( UROtsa , T24 ) and renal epithelial ( A498 ) cells grown on 24-well plates ( Costar ) . Experiments were performed in quadruplicates and repeated at least three times independently . Wild-type and mutant E . coli strains were cultured for 24 h at 37°C on LB agar plates without salt to promote the formation of biofilm . Medium was supplemented with ampicillin ( 100 µg/ml ) or aTc ( 50 ng/ml ) if appropriate . Colonies were scraped off and suspended in PBS . Bacterial cell clusters were then removed by centrifugation at 150×g for 10 min . The number of bacteria was determined spectrophotometrically at 600 nm and confirmed by viable count on blood agar plates after serial dilutions in PBS . To access the inhibitory activity of epithelial cells on E . coli growth and viability two experimental settings were employed . Before ELISA analysis , urine was centrifuged at 350×g for 10 min to remove cells and larger particles . ELISA kits for human IL-8 or mouse MIP-2 were obtained from R&D systems ( Abingdon , UK ) . IL-8 or MIP-2 levels were determined according to the manufacturer's instructions . The lower detection limit for IL-8 and MIP-2 was 31 . 3 pg/ml and 15 . 6 pg/ml , respectively . The urinary levels of creatinine were analyzed colorimetrically , and the levels of IL-8 were expressed as IL-8/creatinine ratios . Urine was collected from patients with E . coli UTI and without having a catheter . A drop of urine was incubated on carbon/Formvar-coated 400-mesh copper grids ( GilderGrids , Lincolnshire , UK ) for one minute . Immunostaining was performed as described previously with minor modification [56] . Briefly , grids were blocked with 1% BSA/PBS for 5 min , then incubated with anti-CsgA [28] ( 1∶200 in 0 . 1% BSA/PBS ) for 60 min at 37°C , followed by incubation with anti-rabbit IgG-10-nm gold antibodies ( 1∶15 in 0 . 1% BSA/PBS; Sigma-Aldrich ) for 30 min at 37°C . Grids were stained with 2% tungstophosphoric acid ( Merck , Darmstadt , Germany ) at pH 6 . Analysis were performed using a FEI Tecnai Spirit electron microscope ( Eindhoven , The Netherlands ) at 80 kV accelerating voltage . Urine was centrifuged at 300×g for 10 min to remove cells and larger particles . Bacteria from the supernatant were collected by centrifugation at 3500×g for 10 min . A 2-µL aliquot of the pellet suspended in a minimal volume of PBS was transferred to a nitrocellulose membrane ( Invitrogen ) and air dried for 15 minutes . Immunostaining was performed as described previously [56] . Briefly , membranes were blocked with 1% milk/1% BSA/PBS for 2 h at room temperature , incubated with anti-CsgA [28] ( 1∶5000 in 1% milk/1% BSA/PBS ) for 1 h , followed by incubation with anti-rabbit horse-radish-peroxidase conjugated antibodies ( 1∶3000 in 1% milk/1% BSA/PBS; Bio-Rad Laboratories , Hercules , CA , USA ) for 1 h . Mouse experiments were approved by the Northern Stockholm Animal Ethics Committee and experiments were carried out according to FELASA guidelines and in compliance with the Committee's requirements . The susceptibility of wild-type and mutant E . coli strains to synthetic LL-37 and mCRAMP ( Innovagen AB , Lund , Sweden ) was determined using a broth microdilution method . Briefly , bacteria were grown overnight at 37°C on LB plates with or without salt , inhibiting or promoting the formation of biofilm , respectively . Then , bacteria were suspended in PBS and diluted in LB broth without salt in a concentration of 105 CFU/ml . The bacterial concentration was verified by viable count after serial dilutions in PBS . Bacteria ( 90 µl suspension ) were grown in 96-well plates in the presence of 10 µl aqueous solution of synthetic LL-37 or mCRAMP in final concentrations ranging from 0 . 6 µM to 20 µM in 2-fold dilutions . After 20 h , bacterial viability was measured colorimetrically by reduction of Alamar blue ( BioSource International , Camarillo , CA , USA ) for 1 h at 37°C [57] . The IC50 was determined as the peptide concentration that gave 50% reduction of the absorbance at 570 nm relative to bacteria grown without peptide . Plates were filled with 90 µL bacterial culture of planktonic cells and 10 µL of aqueous solution of LL-37 in the same concentrations as in the susceptibility assay . As control peptides , sLL-37 ( Innovagen AB ) and VIP [27] with similar structural properties as LL-37 were utilized . The plates were incubated without shaking at 37°C for 20 h . After incubation , the amount of biofilm formed was determined as described above . CsgA-His6 and CsgG , a lipoprotein that is required for CsgA secretion [58] , were overexpressed in LSR12 ( C600::csg ) . The purification procedure of recombinant CsgA-His6 was performed as previously described with some modifications [59] . The filtrate was incubated with nickel-nitrilotriacetic acid ( Ni-NTA ) agarose ( Invitrogen ) for 1 h at 4°C with shaking , centrifuged at 200×g for 5 min and transferred to a polyprep chromatography column ( Bio-Rad , Hercules , CA , USA ) . To confirm protein identity of CsgA-His6 , the isolated protein was prepared and subjected to SDS-PAGE as described below , and thereafter the protein was transferred from the gel to a polyvinylidene fluoride ( PVDF ) membrane ( Invitrogen ) at 160 mA for 60 min . The band of 15 kDa , corresponding to the molecular weight of CsgA-His6 , was excised from the membrane stained with Coomassie Blue and analysed with N-terminal sequence analysis as has been described [60] . Since it has previously been shown that CsgA-His6 has the same polymerizing properties as wild-type CsgA [59] , we refer to CsgA-His6 as CsgA . The cellulose-deficient isogenic mutant of E . coli No . 12 was used to purify wild-type curli , as has been described [17] . Colonies were harvested and suspended in 0 . 05 M Tris-buffer by using an omnimixer . Bacterial cell debris was discarded by centrifugation at 8000×g for 15 min and curli protein was precipitated by adding 0 . 1 M MgCl2 and 0 . 15 M NaCl . The aggregated curli were centrifuged at 16000×g for 15 min , and the pellet was dissolved in 10 mM TRIS , 1 mM EDTA , pH 7 . 5 with 2% 3- ( ( 3-cholamidopropyl ) dimethylammonium ) -1-propanesulfonate ( CHAPS ) . After incubation for 45 min at 95°C , the solution was centrifuged at 20000×g for 10 min . The pellet containing curli was washed three times with water . Finally , curli were suspended in PBS and used for binding studies with LL-37 . Samples were prepared in LDS ( lithium dodecyl sulphate ) sample buffer ( 4∶1 ) and incubated at 70°C for 10 min . Electrophoresis was performed using 4–12% Bis-tris NU-PAGE gels ( Invitrogen ) at 200 V for 35 min . Gels were stained with Coomassie Blue for 1 h and destained over night in 90% water , 8% methanol and 2% acetic acid ( vol/vol/vol ) . Sample preparation and SDS-PAGE was carried out as described above and Western blot was performed as previously described [20] . The antibodies used were monoclonal mouse anti-LL-37 ( 0 . 6 µg/ml in 5% fat-free milk/PBS ) [55] and horse radish peroxidase-conjugated anti-mouse IgG ( diluted 1∶5000 ) . Samples containing 40 µM CsgA in 50 mM potassium phosphate buffer ( KPi ) and 0 . 02% NaN3 , pH 7 . 2 , with and without 10 µM LL-37 were incubated for 60 h at 37°C . The samples were then assayed with a Jasco J-810 spectropolarimeter from 190 to 250 nm in a quartz cell with a 1-mm path length at 20°C . The spectrum of buffer alone and LL-37 in buffer was subtracted from the spectra for CsgA alone and for CsgA together with LL-37 , respectively . Data were compared with student's t-test , Mann-Whitney U test or Fisher's exact test as appropriate . P values of less than 0 . 05 were considered as statistically significant .
|
Most infections of the urinary tract are caused by uropathogenic E . coli . On abiotic surfaces , these bacteria are able to form biofilms , which protect them from various adverse environmental conditions . In this study , we sought to investigate whether two E . coli biofilm components , curli fimbriae and cellulose , provide a similar protection against innate immune defense mechanisms of the urinary tract . We put special emphasis on the interaction with the human antimicrobial peptide LL-37 , which plays a crucial role in the protection against uropathogenic E . coli . We demonstrate that curli expression specifically reduces bacterial sensitivity to LL-37 by binding the peptide before reaching the bacterial cell membrane and exhibiting its bactericidal activity . A more general protection is mediated by cellulose , possibly by hiding immunogenic surface structures of the bacterium . In addition to providing protection , curli are also targeted by the immune system . The formation of new curli fibers is inhibited in the presence of LL-37 . Moreover , curliated bacteria show higher immunogenicity than their non-curliated counterparts . Cellulose expression , on the other hand , appears to impair initial host colonization . In conclusion , our findings demonstrate an example of the tight interplay between bacterial virulence factors and the host immune defense .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"immunology/cellular",
"microbiology",
"and",
"pathogenesis",
"infectious",
"diseases/urological",
"infections",
"urology/urological",
"infections",
"immunology/immune",
"response",
"immunology/innate",
"immunity",
"infectious",
"diseases/bacterial",
"infections",
"immunology/immunity",
"to",
"infections"
] |
2010
|
Uropathogenic Escherichia coli Modulates Immune Responses and Its Curli Fimbriae Interact with the Antimicrobial Peptide LL-37
|
Human T-lymphotropic virus type 1 ( HTLV-1 ) causes leukaemia or chronic inflammatory disease in ∼5% of infected hosts . The level of proviral expression of HTLV-1 differs significantly among infected people , even at the same proviral load ( proportion of infected mononuclear cells in the circulation ) . A high level of expression of the HTLV-1 provirus is associated with a high proviral load and a high risk of the inflammatory disease of the central nervous system known as HTLV-1-associated myelopathy/tropical spastic paraparesis ( HAM/TSP ) . But the factors that control the rate of HTLV-1 proviral expression remain unknown . Here we show that proviral integration sites of HTLV-1 in vivo are not randomly distributed within the human genome but are associated with transcriptionally active regions . Comparison of proviral integration sites between individuals with high and low levels of proviral expression , and between provirus-expressing and provirus non-expressing cells from within an individual , demonstrated that frequent integration into transcription units was associated with an increased rate of proviral expression . An increased frequency of integration sites in transcription units in individuals with high proviral expression was also associated with the inflammatory disease HAM/TSP . By comparing the distribution of integration sites in human lymphocytes infected in short-term cell culture with those from persistent infection in vivo , we infer the action of two selective forces that shape the distribution of integration sites in vivo: positive selection for cells containing proviral integration sites in transcriptionally active regions of the genome , and negative selection against cells with proviral integration sites within transcription units .
The completion of the human genome sequence has made genome-wide studies of retroviral integration possible . Such studies have led to a good understanding of the targeting preferences of retroviruses for integration . However , the genomic distribution of integration sites has not been widely studied in the context of persistent infection , where infected cells are subjected to additional selection forces such as the immune response . Here we tested the hypothesis that in vivo selection influences the genomic distribution of proviral integration sites in persistent HTLV-1 infection . Genome-wide studies have revealed that proviral integration is not random , but that each retrovirus has distinct target site preferences [1]–[4] . For instance , HIV shows a bias towards integration into genes , whereas MLV integration is biased towards CpG islands and the transcriptional start sites of genes [1]–[4] . Each retrovirus also targets a characteristic inverted repeat consensus nucleotide sequence at the site of integration [5]–[11] . The different targeting preferences of the individual retroviruses are determined by several known factors , including the properties of the viral integrase [12] , DNA binding proteins [13] , [14] , cellular targeting proteins [15]–[19] and the chromatin structure at the point of integration [20] , [21] . A recent study of HTLV-1 integration sites isolated from HeLa cells infected in vitro showed that HTLV-1 has integration site preferences most similar to those of avian sarcoma-leukosis virus ( ASLV ) [22] . Both ASLV and HTLV-1 were found to target genes , transcriptional start sites and CpG islands . Although in each case the bias was statistically significant , the magnitude of the effect was lower than observed in integration with other retroviruses ( HIV-1 , SIV , MLV and FV ) [22] . Since the Integrases of ASLV and HTLV-1 are more closely related in sequence to each other than to the other retroviruses , this observation supported the idea that Integrase is a major determinant of retroviral integration site targeting . HTLV-1 is associated with the neoplastic disease Adult T cell Leukemia ( ATL ) and the inflammatory condition HTLV-1-associated myelopathy/ tropical spastic paraparesis ( HAM/TSP ) . The mechanisms of pathogenesis of HAM/TSP have not been elucidated . However , it is well established that the proviral load of HTLV-1–the proportion of peripheral blood mononuclear cells ( PBMCs ) that contain a provirus–varies widely among HTLV-1-infected individuals and is strongly correlated with the risk of HAM/TSP [23] . In addition , there is evidence from the study of the cell-mediated immune response that the rate of HTLV-1 proviral expression at a given proviral load correlates with the outcome of infection [24] . An individual patient's proviral load remains constant over time [25] . In contrast , proviral loads show large variation between patients [23] . The causes of this proviral load variation are not yet clear . It was recently demonstrated that variation between individuals in the efficiency of the cytotoxic T lymphocyte ( CTL ) response determined approximately 30% of the observed between-individual variation in proviral load [26] . An additional 13% of proviral load was determined by the rate of provirus expression , independently of CTL efficiency [24] . We hypothesised that the CTL-independent variation in proviral load was due to molecular factors that affect proviral load and provirus expression . We postulated that one such factor was the integration site of the virus . The proviral load of HTLV-1 appears to be determined by a dynamic balance between viral replication and the host immune response [27] . HTLV-1 has two potential routes of infection: infectious transmission ( across the virological synapse [28] ) or mitotic replication of an infected cell ( instigated by the viral transactivating protein Tax [29] , [30] ) . The sequence of HTLV-1 is stable within an individual , indicating that the proviral load in vivo is maintained chiefly by proliferation of infected cells [31] . This interpretation is supported by the observation of large clones of infected cells with a common integration site in vivo [29] , [30] . We have hypothesised that infectious transmission of HTLV-1 is important early in infection whilst mitotic replication may be responsible for maintaining proviral load later in infection once persistent infection has been established and reached equilibrium with the immune response [27] . There are few existing data on the distribution of HTLV-1 integration sites in non-malignant cases of infection in vivo . Previous studies suggested that genes are favoured targets for the integration of HTLV-1 in ATL [32]–[34] whereas integration sites isolated from asymptomatic carriers ( ACs ) and from patients with HAM/TSP showed no preference for transcriptionally active regions [32] , [35] . The power of these studies was , however , limited by the small number of integration sites and by investigation of few genomic parameters . In this study we compared the genomic characteristics of HTLV-1 integration sites resulting from infection in a cell culture system in vitro with those of integration sites from freshly isolated PBMCs , in which the integration sites have been subject to years of in vivo selection . The results suggest positive selection for cells that possess a provirus integrated in a transcriptionally active genomic region and negative selection against cells with a provirus integrated in a transcriptional unit ( i . e . a gene ) during persistent infection in vivo . Moreover , frequent integration in transcriptionally active genomic regions was associated with a high level of HTLV-1 proviral expression and with the inflammatory disease HAM/TSP .
Studies of other retroviruses have shown that integration occurs in a statistically-defined consensus sequence , identified by analysis of a large number of integration sites [5]–[11] , [22] . It has been suggested that this consensus nucleotide sequence preference is an inherent property of the retrovirus [5] . Therefore , it should not be influenced by in vivo selection . In agreement with this , both in vivo and in vitro integration sites showed an identical consensus sequence at the point of integration ( Figure S1 ) . The datasets of HTLV-1 integration sites were analysed for proximity to genomic features of transcriptional activity including transcription units , transcriptional start sites and CpG islands [3] . Here ‘transcriptional unit’ denotes a full-length unprocessed RNA transcript . The observed frequency of HTLV-1 integration near each feature was compared to the expected frequency derived from analysis of the MRC sites using statistical analyses ( χ-squared and logistic regression ) . A recent study of the integration sites of HTLV-1 in vitro reported significant targeting of transcription units and a weak but significant targeting of regions containing transcriptional start sites and CpG islands [22] . Our in vitro co-culture system revealed identical integration preferences . We therefore combined both sets of HTLV-1 in vitro integration sites and compared the combined dataset to the MRC sites for proximity to genomic features of transcriptional activity . Analysis of the proportions of HTLV-1 integration sites and MRC sites within a specified distance ( window size between 1 kb and 25 kb ) of a CpG island showed a significantly higher frequency of HTLV-1 integration sites than MRC sites at all distances above 2 kb ( 3 . 5% of HTLV-1 integration sites lying ±2 kb of a CpG island compared with 2 . 0% of MRC sites; observed/expected ( O/E ) ratio of 1 . 75; p = 0 . 0059 , χ-squared , Figure 1A ) . In vitro there was no significant difference in the frequency of HTLV-1 proviral integration sites within CpG islands themselves , and ±1 kb of a CpG island , compared to MRCs , but statistical power was limited by the small numbers of sites in this very small genomic region ( 0 . 13% vs 0 . 24% within a CpG island , Table 2; 1 . 13% and 0 . 91% ±1 kb , Figure 1 ) . Compared to the MRC frequencies , there were also significantly more HTLV-1 proviral integration sites in the vicinity of gene transcriptional start sites . This was evident at all distances investigated ( 2–25 kb from a transcriptional start site ) ( i . e . 3 . 2% of HTLV-1 integration sites lay ±2 kb of a RefSeq gene transcriptional start site compared with 1 . 7% of MRCs; O/E = 1 . 88; p = 0 . 026 , χ-squared , Figure 1A ) . There was also a statistically significant excess of proviral integration sites present in transcription units compared to MRC frequencies using a number of different gene annotation databases ( e . g . 44 . 5% in RefSeq genes compared with 34 . 0% of MRCs; O/E = 1 . 31; p = 3 . 3×10−9 , χ-squared , Table 2 ) . Further , by investigating the gene density in genomic regions containing an HTLV-1 provirus , we observed that the proviral integration sites were associated with regions of high gene density: there was a significantly increased frequency of HTLV-1 provirus integration compared to MRC sites in regions of higher gene density at all region sizes investigated ( from 25 kb to 8 Mb from the integration site; p values ranged from 3 . 1×10−19 to 1 . 1×10−8; logistic regression analysis , Figure 1B ) . The integration sites identified in persistent infection in vivo were also associated with transcriptionally active regions as shown by their proximity to CpG islands , transcriptional start sites and their positioning in gene-dense regions . There were three times as many HTLV-1 proviral integrations ( 1 . 92% ) as MRC sites ( 0 . 61% ) lying within a CpG island ( Table 2 , p = 0 . 007 , χ-squared analysis ) . In addition , there was a statistically significant increased frequency of HTLV-1 integration compared to MRC sites near CpG islands at all distances investigated ( 1 kb to 25 kb from a CpG island ) ( Figure 1A ) . For instance , compared to the 3 . 0% of MRC sites , 11 . 5% of the HTLV-1 proviral integration sites from persistent infection lay ±2 kb of a CpG island ( O/E = 3 . 83; p = 4 . 1×10−14 , χ-squared ) . There was also a significant excess frequency of HTLV-1 proviral integration sites near gene transcriptional start sites: 5 . 4% of HTLV-1 sites lay within ±2 kb of a RefSeq gene transcriptional start site compared with 2 . 1% of MRCs; O/E = 2 . 57; p = 0 . 0002 , χ-squared ( Figure 1A ) . The integration sites identified in persistent infection were also found in regions of higher gene density than MRC sites ( Figure 1B ) . However , in contrast to the in vitro integration sites , the frequency of integration in transcription units in vivo was not significantly different from expectation ( the expected value was calculated , as before , from the distribution of MRCs ) ( Table 2 ) . Thus , although the integration sites isolated from persistent infection were shown to be associated with transcriptionally active genomic regions , this did not include an association with transcriptional units . The methods used in this study and by Derse et al used different restriction enzymes ( NlaIII and MseI respectively ) . The proportion of MRCs that lie within a given distance from a CpG island differed according to the restriction enzyme used: 3 . 0% of the MRC in vivo sites generated with NlaIII lay ±2 kb of a CpG island , compared with 2 . 0% of the MRC in vitro sites ( of which 266 were generated with NlaIII and 527 generated with MseI ) . Also , 2 . 1% of NlaIII in vivo sites lay within ±2 kb of a transcriptional start site compared to 1 . 7% of the in vitro MseI/NlaIII sites . That is , the respective restriction enzyme sites are differently distributed in the genome with respect to the features associated with transcriptional activity , and it was therefore essential to use different MRC sets in each case . We therefore compared each dataset to its respective MRC set before comparing the experimental datasets to each other . To test for evidence of in vivo selection on the distribution of integration sites , we compared the genomic distribution of 313 in vivo integration sites with the combined set of 793 in vitro integration sites using a logistic regression model . First , the logistic regression model was used to test for non-random frequency of integration across all chromosomes in vitro and in vivo . The chromosomal distribution of HTLV-1 integration sites in vivo differed significantly from the MRCs ( Figure 2; p = 0 . 00095 ) . Using χ-squared analysis and correcting for multiple comparisons , there was a significant excess frequency of integration in Chromosome 13 in vivo ( p = 0 . 014 ) . The logistic regression model also suggested non-random chromosome distribution of HTLV-1 integration in vitro ( Figure 2 , p = 0 . 016 , χ-squared analysis ) but this bias was not statistically significant after correcting for multiple comparisons ( p>0 . 1 ) . However , logistic regression confirmed a significant difference ( p = 0 . 006 ) between the chromosomal distribution of sites in vivo and in vitro . Compared with the in vitro integration sites , there were significantly more integrations in vivo into transcriptionally active genomic regions , i . e . near a CpG island or a transcriptional start site . The ratio of the numbers of observed HTLV-1 integration sites to expected ( MRC ) sites lying within a specified distance ( ranging from 1 to 25 kb ) from either a RefSeq gene transcriptional start site or a CpG island were compared in vitro and in vivo . Although both datasets showed a greater than expected frequency of proviral integration in transcriptionally active regions , there were also significantly more integrations near both CpG islands and transcriptional start sites in vivo than in vitro ( Figure 1A; p<0 . 005 , logistic regression analysis ) . There were also significantly more integrations lying within a CpG island ( p = 9×10−4 , logistic regression analysis ) ( Table 2 ) in vivo than in vitro . In addition , we analysed the gene density in a range of window sites ( ±25 kb to ±1 Mb ) around each integration site . Logistic regression analysis showed significantly more integrations into regions of higher gene density in vivo than in vitro at all window sizes tested ( p = 0 . 032 at ±25 kb; p = 2 . 5×10−6 at ±100 kb; Figure 1B ) . However , although the observed association with integration into transcriptionally active regions was stronger in vivo than in vitro , there were significantly fewer integrations into transcription units in vivo ( Acembly , RefSeq and Ensemble gene definitions tested; p = 2 . 5×10−4 , 0 . 044 , 0 . 0065 respectively ) ( Table 2 ) . Whereas HTLV-1 proviral integration frequency in transcription units exceeded expectation in vitro , in vivo there was no deviation from expected frequencies . We wished to compare the distribution of integration sites between individuals with a high proviral load and those with a low load; between those with high levels of expression of the viral protein Tax and those with low levels; and between patients with HAM/TSP and ACs . In each comparison , the individuals in Table S1 were split into 2 equal groups based on the above parameters ( i . e . above and below the median value of the respective parameters ) and the integration sites from the two respective groups of subjects were compared using a logistic regression model as set out in Protocol S1 . There were no significant differences in the genomic characteristics of the integration sites between individuals with a high proviral load and those with a low proviral load ( data not shown ) . However , differences in the distribution of integration sites were apparent according to the level of expression of Tax and the disease status . The HAM/TSP group and high Tax group each showed a consistent tendency to a higher frequency of integrations both in the vicinity of genes and in regions of higher gene density . Of the integration sites isolated from high Tax-expressing individuals , 38% were located in RefSeq genes , compared with only 33% of sites from low Tax-expressing individuals . The high Tax-expressing group also had a significantly higher gene density in the region surrounding the integration site than the low Tax-expressing group ( p = 0 . 034; logistic regression model using a 500 kb window around each integration site , data not shown ) . There is evidence that the Tax protein is expressed at a significantly higher frequency in patients with HAM/TSP than in ACs with a similar proviral load [24] . Since the data presented above showed that the proviral integration site distribution was associated with the level of Tax protein expression , we wished to test the hypothesis that the integration site distribution differed between HAM/TSP patients and ACs . The results show that the HAM/TSP patients had significantly more integration sites located in RefSeq genes ( 41% ) compared to the ACs ( 30% ) ( Figure 3 , p = 0 . 049; logistic regression model ) . These observations suggested that the integration site is associated with the rate of expression of the provirus in vivo , i . e . that integration in gene-dense regions is concomitant with higher provirus expression and consequently with a higher risk of the disease HAM/TSP . We tested the validity of this conclusion in an independent series of experiments by comparing the genomic distribution of proviral integration sites between HTLV-1-infected T cells that expressed Tax spontaneously within 18 hrs and those from the same individual that did not express Tax . In independent experiments , CD8+ cell-depleted PBMCs from each of three patients with HAM/TSP were incubated for 18 hrs in vitro to allow spontaneous Tax expression . Each of the three individuals had approximately 5% Tax+CD4+ T-cells after culture . After intracellular staining with a fluorescent-labelled monoclonal antibody for the Tax protein , the Tax-expressing cells were separated from Tax non-expressing cells by flow cytometric sorting . The integration sites in each respective cell population were then cloned . The integration sites from the Tax-expressing ( Tax+ ) fraction for the three individuals were pooled and the distribution compared to that of the Tax non-expressing fraction ( Tax− ) . The results ( Figure 4 ) corroborated the hypothesis that the distribution of integration sites in vivo is associated with the rate of proviral ( Tax ) expression . The Tax+ fraction had a significantly higher proportion of HTLV-1 integration sites in genes compared to the Tax− fraction ( 50% and 23% in RefSeq genes respectively ) ( Figure 4A ) . This difference was significant at the one-tailed level ( p = 0 . 04 ) . The Tax+ cell fraction also had a significantly higher proportion of HTLV-1 integration sites in the vicinity of a CpG island ( with 35 . 7% of integration sites lying within ±10 kb from a CpG island in Tax+ cells compared with 11 . 5% of integration sites in Tax− cells , p = 0 . 034 , χ-squared analysis , Figure 4B ) . Finally , the Tax+ fraction had a significantly higher gene density around each integration site than the Tax− fraction ( p = 0 . 019; logistic regression model using a 1 Mb window around each integration site , Figure 4C ) . In summary the results show that , both between and within patients , proviral integration into areas of transcriptional activity was associated with Tax expression .
The results presented here show that the integrated HTLV-1 provirus is associated with transcriptionally active regions of the human genome both in vitro in cell culture and in vivo in persistent infection . This was shown by an increased frequency of integration in gene-dense regions , near CpG islands and near transcriptional start sites compared to controls . However , we also found significant differences in the distribution of HTLV-1 integration sites when comparing the in vitro sites with those identified in patients' PBMCs , which had been subjected to years of immune and viral selection in vivo . We used 2 datasets of HTLV-1 sites generated in vitro: an existing set resulting from infection of HeLa cells with virus-like particles [22] and a novel set formed by the infection of Jurkat cells by short-term co-culture with MT2 cells . The presented results are in agreement with the analysis by Derse et al on the set of HeLa-derived HTLV-1 proviral integration sites [22] . In addition , the distribution of integration sites did not differ between these two in vitro datasets in any genomic parameter tested . This finding is consistent with previous evidence that the distribution of retroviral integration sites is largely independent of the cell type [1] , [4] , [6] , [22] , [36]–[39] . We then pooled these sites to compare the genomic distribution of 793 in vitro sites with 313 sites resulting from persistent infection in vivo . Two major differences were found between the distribution of in vivo and in vitro HTLV-1 integration sites . First , there was a significantly stronger association with integration in transcriptionally active regions in vivo than in vitro . That is , there were significantly more integrations in vivo into regions of high gene density , near RefSeq gene transcriptional start sites , and near CpG islands . Second , there were significantly fewer integrations into transcription units in vivo than in vitro: whereas HTLV-1 showed a significant bias towards integration in transcription units in vitro both in HeLa and in Jurkat cells , there was no deviation from expected frequencies in vivo . These results are consistent with the existence of two selection forces that act in vivo to shape the initial distribution of integration sites: positive selection for cells containing an integration within a transcriptionally active regions of the genome , and negative selection against gene disruption resulting from proviral integration within a transcription unit . Positive selection for proviruses integrated into transcriptionally active genomic regions is likely to result from the actions of the HTLV-1 Tax protein , which transactivates transcription of both viral and host genes and causes clonal T cell proliferation . This conclusion is consistent with two existing lines of evidence on the mechanism of HTLV-1 persistence in vivo . First , there is evidence that HTLV-1 persists in vivo during the chronic phase of infection chiefly by clonal proliferation of infected cells [29] , [40] . Second , there is an abnormally fast turnover rate of HTLV-1-infected T cells in vivo , especially of Tax-expressing cells [41] , [42] , indicating that the selective proliferation of HTLV-1-infected T cell clones in vivo depends on persistent expression of the provirus . It is possible that proviral integration in transcriptionally active regions also affects the expression of host genes lying in the vicinity of the integration site , so conferring a growth advantage on the infected clone . We observed a significantly lower frequency of HTLV-1 proviral integration in transcription units in vivo than in vitro . We postulate that the lower frequency in vivo results from the disruption of gene function . If correct , this conclusion implies the existence of haploinsufficiency in many human genes [43] . There are currently insufficient published data available to make possible an accurate estimate of the proportion of human genes that are subject to haploinsufficiency . However , there is an increasing number of reports of human diseases , including cancer and developmental disorders , that are associated with haploinsufficiency [44]–[46] . Haploinsufficiency has also shown to be a common phenomenon in yeast [47] , [48] . Negative selection against integration into transcription units was not observed in HIV infection , in which integration was biased towards transcription units both in vitro and in vivo [36] , [49] . Most cells productively infected by HIV have a short lifespan , perhaps too short to allow negative selection against integration in transcription units [50] . In contrast to HIV , HTLV-1 infection results in increased lymphocyte turnover via clonal proliferation [41] . We suggest that this T cell proliferation maintains the proviral load and the lymphocyte count in the face of negative selection against integrations in genes . We analysed the distribution of in vivo proviral integration sites to test the hypothesis that the distribution of integration sites is associated with proviral load , provirus expression and disease status . There were no statistically significant differences in the distribution of integration sites between the high-load and low-load subjects . However , there were significant differences in the distribution of integration sites between patients with high expression of the viral protein Tax and those with low expression , and between patients with HAM/TSP and ACs . Both the high Tax and the HAM/TSP group showed a greater frequency of proviral integrations in genes compared to the low Tax and AC group respectively . We conclude that integration in regions of higher gene density was concomitant with higher levels of the viral protein Tax . Since the level of Tax expression correlates with disease risk [24] , this integration distribution also manifests as a higher proportion of integrations in gene-dense regions in HAM/TSP patients compared to ACs . We propose that integration in transcriptionally active genomic regions favours HTLV-1 proviral expression and persistence , but that this is counter-selected by an efficient immune ( particularly CTL ) response . This conclusion is consistent with evidence that the efficiency of the CTL response correlates negatively with proviral load [26] and with recent evidence that HTLV-1 proviral expression contributes to the persistence of HTLV-1 in vivo [41] . As a more stringent test of the relationship between the integration site and the level of provirus expression , integration sites from Tax+ and Tax− provirus-positive cells isolated from cultured PBMCs were amplified by LM-PCR . We analysed these sites to test the hypothesis that the Tax+ cell fraction had significantly more integrations in transcriptionally active regions . The results showed that the Tax+ fraction indeed had significantly more integration sites lying in transcription units: 50% of the sites in the Tax+ fraction were located in RefSeq transcription units compared to only 23% in the Tax− fraction . Similar experiments with other retroviruses have also shown an association between integration in gene-rich regions and viral transcription in vitro [12] , [51] . The results of this within-patient experiment ( Figure 4 ) corroborated the relationship between integration site distribution and Tax expression . The results also revealed a stronger association between integration in genes and Tax expression within individuals than that observed in the between-individual experiment . We suggest that this stronger association results from the greater difference in Tax expression between Tax+ and Tax− cells ( an all-or-nothing effect ) than between the high-Tax and low-Tax patient groups , which represent two halves of a continuous distribution . In addition , other factors such as variation in the host immune response to HTLV-1 are likely to contribute to between-individual variation in the intensity of selection for proviral expression [26] . We propose that an individual's steady-state rate of proviral expression ( measured as Tax protein ) , and the accompanying risk of inflammatory diseases such as HAM/TSP , are the result of an equilibrium between HTLV-1 replication and the immune response in vivo ( Figure S2 ) . Selection for cells with high levels of Tax expression leads to an increased frequency of proviruses integrated in gene-dense regions of the genome . Clonal proliferation of an infected cell leads to dominance of Tax-expressing clones . The increased proviral load resulting from this proliferation will increase the rate of infectious spread of the virus and thereby increase the frequency of integration in genes , owing to the intrinsic preference of the virus . Balancing these positive forces , the specific CTLs kill HTLV-1-expressing cells and therefore select against integration in transcriptionally active areas of the genome . In addition , integration into genes , which may disrupt their function , reduces the steady state frequency of proviruses in gene-dense regions . These parameters act in concert with host genetic factors [52] , [53] to determine an individual's proviral load , level of Tax expression and disease status .
Samples from 24 individuals infected with Human T-cell Leukaemia Virus 1 ( HTLV-1 ) were analysed . These included 11 ACs and 13 patients with HAM/TSP . All patients attended the HTLV-1 clinic at St Marys Hospital , London and donated their blood having given written informed consent . HTLV-1 infection was confirmed by the presence of antibodies to HTLV-1 Gag and Env antigens in sera by Western blot ( HTLV blot 2 . 4; Genelabs ) . Diagnosis of HAM/TSP was made following World Health Organisation criteria . PBMCs were isolated from whole blood by density gradient centrifugation using Histopaque-1077 ( Sigma ) . Cells were washed twice in PBS and cryopreserved in fetal calf serum ( FCS , Sigma ) with 10% dimethyl sulphoxide ( DMSO , Sigma ) in liquid nitrogen until required . DNA was extracted as described in the manufacturer's protocol ( Qiagen , DNeasy Tissue Kit ) . The HTLV-1 producing cell line MT2 was labelled with CD4+ microbeads ( Miltenyi ) , stained with CFSE and gamma-irradiated ( 137Cs , 40 , 000 rads ) . The cells were then co-cultured with Jurkat cells for 3 hrs at 37°C at a 1∶1 ratio . The MT2 cells were then depleted and an aliquot of the Jurkat cells analysed to verify Gag transmission from the MT2 cells and to quantify MT2 contamination . Contamination was less than 5% . By one week co-culture , all remaining MT2 cells had died . Sixteen days post-culture , DNA was extracted from the Jurkat cells and amplified for HTLV-1 integration sites . To verify integrations were novel and not contaminating MT2 sites , DNA was also extracted from MT2 cells and the resulting integrations were used to search the set of novel integrations for MT2 sites . No contaminating MT2 sites were found . In this paper , we refer to integration sites generated by infection in cell culture as in vitro sites and those derived from patients' PBMCs as in vivo sites . Four µg of extracted DNA was digested with 10 U Nla III ( New England Biolabs ) in a total volume of 50 µl for 3 hours at 37°C . After purification with the PCR clean-up kit ( Qiagen ) following manufacturer's instructions , DNA was eluted into 50 µl elution buffer . For the linear linker-mediated PCR for amplification of in vivo sites , 20 µl of this DNA solution was incubated for 30 minutes at room temperature with 2 µl Quick Ligase ( New England Biolabs ) and 40 pmoles of the primer Bio1 [29] . After a further purification , the ligated DNA was eluted into 55 µl , of which 10 µl was used in each of four replicate linear PCR reactions: 100 cycles , using 20 pmoles of the primer Bio2 [29] and the following conditions: 94°C for 10 min; 100 cycles of 95°C for 45 sec , 60°C for 45 sec , 72°C for 2 min; and a final elongation step of 72°C for 10 min . Ten µl of a five-fold dilution of the linear PCR product was used in the classical ( bidirectional ) PCR reaction . In the bidirectional PCR , DNA was amplified with 20 pmoles each of Bio3 and Bio4 [29] with the following conditions: 94 °C for 10 min; 100 cycles of 95°C for 45 sec , 60°C for 45 sec , 72°C for 2 min; and a final elongation step of 72°C for 10 min . For amplification of the in vitro integration sites , the digested product was ligated to a longer double-stranded NlaIII linker overnight at 16°C . Bidirectional PCR was then carried out using primers in the viral LTR ( Bio2 ) and the linker ( AE2814 ) , followed by nested PCR using the primers Bio3 and AE2815 [36] . Two µl of amplified integration site products were combined with 0 . 5 µl TOPO-TA cloning sequencing vector PCR-4® and used to transform MachT1 cells according to manufacturer's instructions ( Invitrogen ) . Clones were picked and sequenced using the T3 primer ( Invitrogen ) . To check that the amplification was HTLV-1 specific , plasmids were digested with EcoRI , run on a 2% agarose gel and transferred to nylon membrane ( Roche ) by southern blotting . They were then probed using the Bio5 primer found in the viral LTR [29] conjugated to a single digoxigenin-labelled dideoxyuridine-triphosphate ( DIG-ddUTP ) ( Bio5-DIG ) according to manufacturer's instructions ( Roche Applied Science ) . Membranes were pre-hybridised for 30 min at 70°C using DIG EasyHyb ( Roche Applied Science ) before 4 hours of hybridisation with 10 pmol Bio5-DIG in 7 ml EasyHyb . Membranes were washed twice for 5 min in 2xSSC/ 0 . 1% SDS and twice for 15 min in 0 . 5xSSC/ 0 . 1% SDS at hybridisation temperature . Bound Bio5-DIG was detected using the Roche DIG nucleic acid detection kit following manufacturer's protocol . Results were documented by scanning the membrane once dry . The genomic integration site was located within the cloned sequence by identifying the terminal 5′ sequence of the viral LTR and the junction between the genomic sequence and Bio1 . The genomic sequence was mapped on to the hg17 assembly of the human genome using the BLAST-Like alignment tool ( BLAT ) [54] http://genome . ucsc . edu/cgi-bin/hgBlatcommandstart . A match was defined as a sequence having all 3 of the following: A total of 313 sequences from persistently infected individuals and 266 sequences from the in vitro co-culture could be mapped to the human genome obeying the above criteria . In addition , 527 sequences from a previously published in vitro study were mapped using the same protocol [22] . Integration sites were examined for the occurrence of various chromosomal features using tables available from the University of California , Santa Cruz database [55] . Proviral integration sites can be downloaded from http://bushmanlab . pbwiki . com/f/Meekings . fa . A set of randomised controls was generated . Ten control sites per integration site were generated , each equidistant from a genomic NlaIII site , to account for any possible bias introduced by a non-random genomic distribution of the restriction enzyme . These sites were analysed in the same manner as the experimental sites . The statistical method used to compare the experimental sites to MRC sites and the comparison between retroviruses is described in Protocol S1 . Software tools and annotated data files are available upon request . To stain for cell surface markers , treated cells were washed once , fixed with 2% paraformaldehyde ( Sigma ) for 20 mins at room temperature , washed again and resuspended in 100 µl PBS/ 10% FCS in a ‘V’ bottomed 96-well plate with appropriate antibodies ( 15 µg/ml phycoerythrin-cyanine 5 ( PC5 ) -conjugated anti-CD4 and energy coupled dye ( ECD ) -conjugated anti-CD8 antibodies ( Immunotech ) ) for 20 mins at room temperature . Cells were permeabilised at room temperature for 10 mins using PBS/ 0 . 1% Triton X-100 ( Sigma ) . The cells were then washed and resuspended in PBS/7% normal goat serum ( NGS , Sigma ) with 1∶200 dilution of Lt-4 anti-Tax antibody ( gift from Y . Tanaka ) for 25 mins at room temperature . Cells were resuspended in PBS ready for flow cytometric analysis . All flow cytometry was done on a Coulter Epics XL ( Beckman Coulter ) flow cytometer and data analysis done using Coulter Expo 32 software . DNA was amplified for HTLV-1 DNA and for β-actin ( as a measure of genomic DNA ) using the Tax sequence-specific primers SK43 and SK44 [56] and β-actin primers specific for the 5′ and 3′ ends . Three dilutions of neat eluted DNA ( 1∶4 , 1∶8 , 1∶16 ) were amplified by real time quantitative PCR in a Roche light cycler using SYBR® Green 1 Dye incorporation ( Roche ) and 1 µM of each primer . Incorporation was detected at 85°C at the end of each of the 45 cycles . Standard curves were generated using the rat cell line TARL2 which contains 1 copy per cell of the HTLV-1 provirus [23] . The sample copy number was estimated by interpolation from the standard curve , calculated as an average of the 3 dilutions , and expressed as percentage of PBMCs infected , assuming 1 proviral copy per cell .
|
The human leukaemia virus HTLV-1 causes a lifelong infection that cannot be cleared by the immune system . By integrating into the host's DNA , the virus can lie dormant within the cell . The virus can then be reactivated , by processes that are only partly understood , causing the infected cell to multiply and leading to an increase in the quantity of virus in the infected person . In some infected people , the virus is reactivated much faster than in others , and such people are more likely to develop HTLV-1-associated inflammatory diseases such as HAM/TSP , which results in paralysis of the legs . It is not understood what determines this rate of viral reactivation in each person . In this study , we found that integration of HTLV-1 in the host's DNA close to other genes was associated with faster viral reactivation and a higher probability of HAM/TSP . By comparing the viral integration site positions in samples from patients and in cells infected with HTLV-1 in the laboratory , we can identify some of the major forces that allow the virus to persist lifelong whilst avoiding eradication by the immune response .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"virology/persistence",
"and",
"latency",
"virology/viral",
"replication",
"and",
"gene",
"regulation",
"virology",
"virology/viruses",
"and",
"cancer",
"virology/effects",
"of",
"virus",
"infection",
"on",
"host",
"gene",
"expression"
] |
2008
|
HTLV-1 Integration into Transcriptionally Active Genomic Regions Is Associated with Proviral Expression and with HAM/TSP
|
Apicomplexans facilitate host cell invasion through formation of a tight-junction interface between parasite and host plasma membranes called the moving junction ( MJ ) . A complex of the rhoptry neck proteins RONs 2/4/5/8 localize to the MJ during invasion where they are believed to provide a stable anchoring point for host penetration . During the initiation of invasion , the preformed MJ RON complex is injected into the host cell where RON2 spans the host plasma membrane while RONs 4/5/8 localize to its cytosolic face . While much attention has been directed toward an AMA1-RON2 interaction supposed to occur outside the cell , little is known about the functions of the MJ RONs positioned inside the host cell . Here we provide a detailed analysis of RON5 to resolve outstanding questions about MJ complex organization , assembly and function during invasion . Using a conditional knockdown approach , we show loss of RON5 results in complete degradation of RON2 and mistargeting of RON4 within the parasite secretory pathway , demonstrating that RON5 plays a key role in organization of the MJ RON complex . While RON8 is unaffected by knockdown of RON5 , these parasites are unable to invade new host cells , providing the first genetic demonstration that RON5 plays a critical role in host cell penetration . Although invasion is not required for injection of rhoptry effectors into the host cytosol , parasites lacking RON5 also fail to form evacuoles suggesting an intact MJ complex is a prerequisite for secretion of rhoptry bulb contents . Additionally , while the MJ has been suggested to function in egress , disruption of the MJ complex by RON5 depletion does not impact this process . Finally , functional complementation of our conditional RON5 mutant reveals that while proteolytic separation of RON5 N- and C-terminal fragments is dispensable , a portion of the C-terminal domain is critical for RON2 stability and function in invasion .
The Apicomplexa are a large phylum of eukaryotic pathogens comprised of ∼6 , 000 described species which cause extensive disease in humans and other animals [1] , [2] . Species of particular interest include Toxoplasma gondii , which chronically infects approximately one-third of all humans and causes neurological disorders in immunocompromised individuals as well as the human malarial agent , Plasmodium falciparum , which is the cause of nearly a million deaths annually [3] , [4] . The disease caused by these obligate intracellular parasites is dependent upon their ability to penetrate , form a specialized vacuole , and replicate within their host cells [5] . Thus , a better understanding of the parasite molecules and processes that facilitate host cell invasion is needed to aid in development of better therapeutics and control strategies . Invasion in apicomplexans is a highly coordinated process of attachment and penetration that depends on sequential protein secretion events from two different organelles , the micronemes and rhoptries [6] . Initially , secretion from the micronemes releases molecular adhesins onto the parasite's plasma membrane , facilitating attachment to the host cell surface [7] . Translocation of these adhesins in an apical to posterior direction via an actin-myosin motor within the parasite pellicle generates a unique gliding motility which is thought to provide the force for host cell penetration . Intriguingly , the recent disruption of MIC2 and myosin A , key components of the gliding motility machinery previously considered essential to invasion , suggests the existence of alternative forces that can drive parasite penetration [8] , [9] . After initial attachment , the parasite apex is oriented toward the host cell , followed by discharge of the rhoptry contents [10] . Rhoptry secretion corresponds with the formation of a ring-shaped tight-junction interface between parasite and host plasma membranes called the moving junction ( MJ ) through which the parasite passes to enter the host cell . A complex of the rhoptry neck proteins RONs 2/4/5/8 localizes to the MJ during invasion where it is thought to provide a stable anchoring point for host cell penetration , possibly through interaction with the host cell cytoskeleton as host cytoskeletal components localize to the MJ and are important for invasion [11]–[15] . The MJ is also the site of a molecular sieve that restricts access of host plasma membrane proteins to the nascent parasitophorous vacuole , rendering the vacuole non-fusogenic and protecting the parasite from lysosomal destruction , a function that may be performed by the MJ RON complex [16] . The micronemal adhesin AMA1 tightly binds RON2 in both Toxoplasma and P . falciparum extracts and peptides or antibodies which block this interaction interfere with invasion [11] , [17]–[22] . These findings led to a model in which binding of RON2 ( a transmembrane protein injected from the rhoptries into the host plasma membrane ) to the ectodomain of AMA1 ( a transmembrane protein secreted from the micronemes into the parasite plasma membrane ) mediates tight-junction formation to bridge the invading parasite and host cell surfaces . However , the actual importance of the RON2-AMA1 interaction for tight-junction formation and invasion is now in question following recent reports showing that disruption of AMA1 has no detectible function in the MJ-mediated penetration step of invasion but instead plays a key role in adhesion [23] , [24] . In contrast to RON2 and AMA1 , RONs 4/5/8 are soluble proteins that are positioned in the host cytoplasm during invasion [13] , [14] . A knockout of the coccidia-restricted MJ component RON8 shows that while not essential , this protein is important for efficient invasion [25] . Taken together with the conservation of the other MJ RONs across the Apicomplexa , these data suggest a core complex of RON2/4/5 comprises the critical invasion machinery . However , current genetic evidence for the role of this core complex in invasion is limited to a conditional RON4 mutant in P . berghei that inhibits invasion by sporozoites , highlighting the need for direct functional analysis of RON5 and RON2 by reverse genetics approaches [23] . Here we provide a comprehensive analysis of Toxoplasma RON5 to evaluate its role in assembly of the MJ complex and function in invasion . Using a conditional knockdown approach , we show that depletion of RON5 results in the complete loss of RON2 and mistargeting of RON4 , indicating RON5 is critical for organization of the MJ complex . In contrast , targeting of RON8 is unaffected by disruption of the RON2/4/5 complex , in keeping with it being a coccidial-specific addition to the core complex . Parasites lacking RON5 egress efficiently but cannot invade new host cells or inject rhoptry effectors into the host cytosol , demonstrating the key importance of the MJ RON core complex in host cell penetration . Complementation of RON5 knockdown parasites with a series of mutants reveals that while proteolytic separation of RON5 N- and C-terminal fragments is dispensable , the C-terminal domain is critical for RON2 stability and MJ function . Together , this work demonstrates that RON5 is crucial for the organization of the MJ complex and provides the first genetic demonstration that the MJ RON core complex is critical for host cell invasion by the Toxoplasma parasite .
During maturation in transit to the rhoptries , RON5 is processed at least twice to separate the protein into three fragments . Antibodies raised against the RON5-N or -C fragments demonstrated that both were incorporated into the mature MJ complex and secreted into the MJ during invasion [13] , [14] . However , the fate of the fragment removed by a more N-terminal processing event ( predicted to be residues 34–314 following removal of the signal peptide ) remains to be characterized . Several rhoptry proteins contain N-terminal pro-domains that are critical for organelle targeting and thus this region may constitute a pro-domain for trafficking of this component of the MJ complex [26]–[28] . Alternatively , this region may be incorporated into the mature complex and function in the MJ during invasion . A stretch of hydrophobic residues that could form a transmembrane domain is present in this region , and thus two models have been proposed for topology during invasion with RON5 either spanning the host plasma membrane similar to RON2 or soluble within the host cytosol [13] , [14] , [29] . To resolve this point , we generated a double epitope tagged version of RON5 with a FLAG tag at the C-terminus and an internal HA tag just downstream of the predicted signal peptide cleavage site ( Figure 1A ) . This version of RON5 was placed under control of the RON5 promoter and the resulting expression cassette was targeted to the UPRT locus to enable stable expression of this double-tagged second copy of RON5 . As expected , the C-terminal FLAG tag was readily detected in the rhoptry necks , as assessed by colocalization with the non-MJ rhoptry neck marker RON11 ( Figure 1B ) . We then monitored rhoptry maturation in parasites expressing this cassette using an antibody against the pro-domain of ROP4 that specifically labels pro-rhoptry compartments [30] . While RON5C-FLAG was present in both pro and mature rhoptries , HA signal was only detected in proROP4-positive compartments , demonstrating that the N-terminal portion of RON5 is a pro-domain ( hereafter referred to as proRON5 ) that is not present in mature rhoptries and thus is not incorporated into the mature MJ complex ( Figure 1C ) . RONs 2/4/5/8 are the only rhoptry proteins known to localize to the moving junction and are believed to play an important role in host cell invasion . We have shown that the Coccidia-restricted RON8 is important but not absolutely required for invasion , suggesting that the remaining MJ RONs 2/4/5 compose an apicomplexan MJ core complex that constitutes the key invasion machinery employed across the phylum [25] . While the Toxoplasma genome encodes a RON4 paralog and two RON2 paralogs , RON5 appears to be a single copy gene with no isoforms [31] . Thus , we reasoned that disruption of RON5 was likely to yield unambiguous functional insight into the MJ core complex . Repeated attempts to disrupt RON5 in the RHΔku80 parasite strain were unsuccessful , further suggesting a critical role in parasite biology . To directly assess the function of RON5 using a conditional approach , we first generated a parasite strain containing a C-terminal 3xMYC epitope tag at the endogenous RON5 locus to improve detection of the protein and then replaced the endogenous RON5 promoter with a tetracycline-regulatable promoter element ( TRE , which is composed of seven tandem TetO sequences immediately upstream of a constitutive , truncated SAG4 promoter , Figure 2A ) . To allow for various epitope tag combinations in downstream experiments , we similarly constructed a RON5-3xHA version of this strain ( designated as RON5MYCcKD or RON5HAcKD ) . This second tagged conditional knockdown line also provided an independent confirmation of our results . As expected from the truncated promoter contained within the TRE , parasites having undergone the desired recombination event show a lower level of RON5 expression compared to the parental line ( Figure 2B , note also the upshift in migration of RON5C due to the presence of the epitope tag ) . Rhoptries are assembled de novo during each round of parasite division and protein traffic to the organelle is restricted to a narrow window during biosynthesis [32] . In agreement with this , we observe some mistargeting of RON5C under the control of the constitutive TRE promoter in the RON5cKD parasites ( Figure 2C , -Atc ) . Similar mistargeting was previously observed when expression of other rhoptry proteins were placed under the control of the TRE and likely corresponds to protein synthesized outside of the rhoptry biosynthesis timeframe [33] . Importantly , a focus of RON5C signal is present in the rhoptry necks of each cell , as assessed by co-localization with RON11 ( arrow , Figure 2C ) and RON5C is clearly detectible in MJ rings during invasion ( arrows , Figure 2D ) . Treatment with anhydrotetracycline ( Atc ) to repress expression results in a steady loss of RON5 with protein levels dropping below detectability by 72 hours ( Figure 2E ) . No gross effect on rhoptries was observed in parasites lacking RON5 as assessed by IFA with rhoptry body markers ROP2/3/4 ( not shown ) and the non-MJ rhoptry neck marker RON11 ( Figure 2E , +Atc ) . Additionally , no defect in intracellular replication was detected in parasites lacking RON5 ( data not shown ) . To test the importance of RON5 in host cell entry , we performed invasion assays on RON5cKD parasites depleted of RON5 . In the absence of Atc treatment , a minor decrease in the invasive capacity of these parasites is observed relative to the parental line , likely corresponding to the lower levels of RON5 produced in the knockdown strain ( Figure 3A , red bars ) . In contrast , a major block in invasion is observed following depletion of RON5 ( Figure 3A , red bars ) , indicating that RON5 is critical for this process . The low level of invasion observed following Atc treatment may be the result of residual levels of RON5 present in some parasites or could indicate that parasites lacking RON5 are able to invade but only at very low levels . In an attempt to distinguish between these possibilities , we performed pulse invasion assays in order to observe Atc-treated RON5cKD parasites in the process of host cell penetration . While invasion events were rare in these assays , all penetrating parasites observed displayed detectable RON5 in the moving junction and/or rhoptry necks ( Figure S2 ) suggesting invasion events in Atc-treated parasites are the result of residual RON5 . We next performed plaque assays to better assess the invasion defect over the course of several lytic cycles . While wild-type parasites readily formed plaques in the presence or absence of drug treatment , no plaques were formed by the RON5cKD parasites in the presence of Atc , even with a hundred-fold higher parasite load ( Figure 3B ) . Together , these results show the critical importance of RON5 for Toxoplasma invasion and suggest that RON5 may be essential for this process . The block in parasite invasion in the absence of RON5 is not accompanied by a simultaneous increase in attached parasites ( Figure 3A , blue bars ) . Thus , parasites depleted of RON5 either exhibit an attachment defect , or more likely , initially attach normally but then detach following a failure to invade as previously observed with disruption of RON8 and with knockdown of TgDHHC7 or TgARO [25] , [33] . To distinguish between these two possibilities , we treated parasites with cytochalasin D to inhibit actin polymerization , arresting the invasion process just after apical reorientation but prior to penetration by disabling gliding motility [34] . Under these conditions , parasites depleted of RON5 were found to attach to host cells with the same efficiency as untreated or parental line parasites . This data indicates that initial attachment is not impaired and suggests that a failure to invade in the absence of RON5 is followed by gliding motility-based detachment and that these parasites are then washed away during invasion assay processing ( Figure 3C ) . During invasion , parasites inject a number of key effectors from the rhoptry body into the host cytosol to modulate host signaling and innate immunity [35] . Rhoptry secretion can be visualized independent of invasion by arresting the invasion process with cytochalasin D . Under these circumstances , an early stage MJ is still formed at the point of apical contact between the parasite and host cell surface and several rhoptry body proteins ( ROPs ) can be visualized entering the cell in membranous structures called evacuoles [36] . To determine the importance of RON5 for secretion of rhoptry body contents , we monitored the formation of evacuoles in parasites depleted of RON5 . Although cytochalasin D-treated parasites lacking RON5 still attach normally ( Figure 3C ) , a nearly complete loss of evacuole formation was observed indicating RON5 is critical for injection of rhoptry contents into the host cell ( Figure 3D and Figure S1 ) . In addition to the roles in invasion and rhoptry secretion highlighted above , the MJ has also been suggested to play a role in host cell exit as RON4-positive MJ rings have been reported to form during egress [8] , [11] , [37] . To assess the importance of RON5 in this process , we induced egress using the calcium ionophore A23187 . Under these conditions , we observed no defect in host cell exit as parasites with or without RON5 egressed with the same efficiency ( Figure 3E ) . Collectively , these results demonstrate that RON5 plays a critical role in host cell invasion but is dispensable for egress . Our results with the RON5 knockdown are in agreement with the recent finding that ablation of rhoptry tethering ( via knockdown of the rhoptry-localized palmitoyl acyl transferase TgDHHC7 and its putative substrate TgARO ) similarly blocks invasion but not egress and highlights RON5 as a critical player in rhoptry mediated invasion [33] , [38] . To determine the impact of the loss of RON5 on the rest of the MJ complex , we examined the remaining MJ RON components by Western blot analysis . Interestingly , following RON5 knockdown , the RON2 signal is also eliminated , indicating that RON5 is critical for maintaining the stability of RON2 ( Figure 4A ) . This complete loss of signal is specific to RON2 as the protein levels of RON4 and RON8 are not similarly impacted under these conditions . While RON2 protein levels closely mimic those of RON5 over a series of time points during RON5 knockdown , qPCR analysis showed no decrease in the transcription of RON2 ( instead we surprisingly observe an approximately two-fold increase in RON2 transcripts ) , indicating that loss of RON2 occurs at the protein level ( Figure S3 ) . The dependence of RON2 upon RON5 was also clearly observed by IFA as parasites lacking RON5 also lack RON2 ( Figure 4B ) . In contrast to the destabilization of RON2 , RON8 is intact and properly targeted to the rhoptry necks in the absence of RON5 ( Figure 4B ) . While Western blot analysis of RON4 indicates that it is largely stable in the absence of RONs 5 and 2 , IFA revealed a targeting defect with RON4 signal often observed throughout the rhoptry bodies but not in the necks ( Figure 4C ) . While the interactions of individual RONs in the MJ complex are unknown , this loss of colocalization between RON4 and RON8 strongly suggests that these proteins do not directly interact in the absence of RONs 2 and 5 . Collectively , these results demonstrate that RON5 is required for the stability of RON2 and proper targeting of RON4 and show that RON5 knockdown effectively constitutes a RON5/2 double knockdown . The effect of RON5 knockdown on the integrity of the MJ complex and on invasion raises the question as to what regions of RON5 are necessary for maintaining stability of RON2 as well as whether any RON5-specific roles during invasion exist . To explore these questions , we established a functional complementation system in our RON5MYCcKD strain by targeting a full-length RON5 expression cassette under the control of its endogenous promoter to the UPRT locus . To distinguish this copy of RON5 from the MYC-tagged , regulatable copy transcribed from the endogenous locus , we engineered an HA epitope tag at the C-terminus . As expected , this HA-tagged version of RON5 targets properly to the rhoptry necks , co-localizing with RON11 ( Figure 5A ) . Expression of this second copy of RON5 , which is insensitive to Atc as it is driven from the RON5 promoter , fully rescues the stability of RON2 upon knockdown of endogenous RON5 ( Figure 5B ) . In addition , complementation with full-length RON5 rescues invasion to wild-type levels and restores the ability of these parasites to plaque in the presence of Atc ( Figure 5C–D ) . We next employed this system to assess the importance of processing of RON5 into RON5N and RON5C . To determine the site of processing , we scanned the RON5 sequence to identify candidate sites that match the consensus P1–P4 sequence characterized in other rhoptry protein processing events ( SΦXE , where Φ is a hydrophobic residue and X is any residue ) [39] . A single match was identified ( SFVE , residues 1258–1262 ) within the region where processing is expected to occur based on SDS-PAGE migration of the mature N- and C-terminal fragments and peptide coverage generated from mass spectrometric analysis of RON5N and RON5C ( Figure 6A ) [11] , [13] . However , mutagenesis of all four residues of this site ( SFVE>AGDR , expected to completely block processing ) in a second copy of RON5 did not affect migration of the C-terminal fragment by SDS-PAGE relative to the wild-type protein , demonstrating that this mutant was still processed ( Figure 6B , SFVE>AGDR ) . We have recently shown that processing of the rhoptry protein TLN1 occurs at a similar sequence containing a glutamine instead of aspartic acid ( SFVQ ) [28] . An SFVQ site is also present within the region where RON5 N/C processing is expected to occur ( residues 1288–1291 , Figure 6A ) . Similar mutagenesis of this site ( SFVQ>AGDR ) results in a modest upshift of RON5C that does not agree with a block in processing to separate RON5N and C , but is consistent with processing upstream at the SFVE site ( Figure 6B , SFVQ>AGDR ) . To test if this was the case , we generated a double mutant at both sites and observe a large upshift in this mutant to the approximate size expected for uncleaved RON5N/C minus its N-terminal pro domain , indicating a block in RON5N/C processing ( Figure 6B , SFVE+SFVQ ) . These results indicate either that processing of RON5 is favored at SFVQ and shifted to SFVE upon ablation of this site , or that processing occurs at both sites in the endogenous protein . To assess the functional impact of the failure to separate RON5N/C , we complemented the RON5MYCcKD strain with the double processing mutant . Despite the block in RON5N/C processing , this mutant was found to target to the rhoptry necks and MJ ring in an indistinguishable manner from the wild-type protein . This was also true following Atc depletion of endogenous RON5 , ruling out the possibility that endogenous RON5 supports proper trafficking of heterogeneous complexes containing both processed and unprocessed forms of RON5 ( Figure 6C–D ) . Surprisingly , the processing mutant also fully rescued the stability of RON2 ( not shown ) , invasion and the ability to form plaques upon depletion of endogenous RON5 ( Figure 6E–F ) . Collectively , these results demonstrate that proteolytic separation of RON5N and RON5C is not important for MJ complex integrity , trafficking or function , a surprising result considering that multiple processing sites are maintained within this region of the protein . Since RON5N/C processing is not required for function , we tested the possibility that RON5C is dispensable all together . To guide the design of addition mutants , we generated an alignment between Toxoplasma and P . falciparum RON5 sequences to determine conservation hot spots that might encode key regions for interaction with other complex members and function ( Figure S4 ) . The alignment reveals three general regions of varying conservation between the two species , the highest of these being the C-terminal half of RON5N ( residues 897–1257 ) . An intermediate level of conservation is seen for the N-terminal half of RON5N ( residues 315–896 ) while the lowest level of conservation is observed in an area that roughly corresponds to Toxoplasma RON5C ( residues 1258–1702 ) . Using this information together with secondary structure prediction , we designed a series of C-terminal truncations and expressed each of these mutants from the UPRT locus in the RON5MYCcKD strain ( Figure 7A ) . Three of these truncations ( Δ618-1702 , Δ898-1702 , and Δ1084-1702 ) , each of which removes the entire RON5C region as well as portions of RON5N , were found to grossly mistarget ( Figure 7B ) . For each of these mutants the mistargeted signal was sometimes absent in individual parasites within a clonal line , suggesting cell cycle variance . Indeed , co-staining for the IMC apical cap marker ISP1 showed that HA signal was only observed in parasites in the process of assembling daughter buds ( and thus new rhoptries ) , indicating that these RON5 mutants are likely degraded following a failure to target to the rhoptry neck as previously observed for other RON targeting mutants ( Figure S5A ) [40] . In contrast , truncations which remove half ( Δ1476-1702 ) or all ( Δ1258-1702 ) of RON5C continue to target to the rhoptry necks ( Figure 7B ) , although some cell-cycle dependent mistargeting was still observed ( not shown ) . Interestingly , the majority of RON5Δ1476-1702 signal localized slightly posterior to non-MJ complex markers for the rhoptry neck , although the significance of this slight shift in localization is unclear . These results indicate that RON5C is dispensable for trafficking while the C-terminal region of RON5N is necessary for localization to the rhoptry necks . To further explore trafficking determinants , we targeted proRON5 for deletion . Although the site ( s ) of proRON5 cleavage is not known , a candidate SFVE is found at residues 311–314 , which agrees with the N-terminal boundary of RON5N suggested by previous proteomic analyses [11] , [12] . To determine if proRON5 is important for RON5 targeting , we created an inframe deletion removing the region between the signal peptide and putative pro cleavage site ( residues 36 to 314 ) in the HA/FLAG double-tagged RON5 expression cassette . Detection with both HA ( Figure S5B ) and FLAG ( data not shown ) epitopes showed gross mistargeting of this protein . While this result demonstrates that proRON5 is necessary for RON5 targeting , our C-terminal truncation analysis indicates that proRON5 is not sufficient for this process . Together , these results suggest that the pro region as well as the C-terminal portion of RON5N play a role in proper targeting , possibly through ensuring proper RON5 folding and/or facilitating interaction with other members of the MJ RON complex . We next evaluated the ability of our C-terminal truncation mutants to rescue the stability of RON2 upon knockdown of endogenous RON5 . As expected , mutants which failed to traffic to the rhoptry neck ( Δ618-1702 , Δ898-1702 and Δ1084-1702 ) also failed to stabilize RON2 in the absence of endogenous RON5 ( Figure 7C ) . While the Δ1258-1702 truncation mutant lacking RON5C does target to the rhoptry neck , it also fails to rescue RON2 stability , demonstrating that although RON5N/C processing is dispensable , RON5C is required for RON2 integrity . In contrast , the Δ1476-1702 truncation lacking the C-terminal 227 residues of RON5C completely rescues RON2 stability ( Figure 7C ) . To monitor both the impact on penetration and downstream intracellular survival , we performed invasion and plaque assays using the RON5MYCcKD strain complemented with the Δ1258-1702 or ΔA1476-1702 mutants . We found that the Δ1258-1702 mutant was unable to rescue invasion or form plaques upon knockdown of endogenous RON5 while the Δ1476-1702 mutant restored both of these phenotypes ( Figure 7D–E ) . As expected , RON5Δ1476-1702 localized to the MJ of invading parasites following knockdown of endogenous RON5 ( Figure 7F ) . Taken together with our analysis of N/C processing , these results identify residues 1292–1475 of RON5 as critical for maintaining RON2 stability and suggest this domain may directly interact with RON2 .
The establishment of a tight-junction interface between invading apicomplexan parasites and their host cells was first observed by electron microscopy over 30 years ago [41] . More recently , the exciting discovery that a complex of rhoptry neck proteins is secreted into this tight-junction provided candidates for understanding the molecular basis for this unique mechanism of host cell penetration [11] . While a relatively thorough characterization of RON protein topology within the MJ has been carried out , a hydrophobic stretch of residues in the N-terminus of RON5 has been noted as a potential transmembrane region , which would impact the positioning of RON5 in this model ( see [14] , [29] ) . We show here that the RON5 N-terminal domain in which this hydrophobic region is contained is not a part of the mature MJ complex . Instead , this pro domain likely plays roles in RON5 folding or trafficking as deletion of proRON5 resulted in gross mistargeting of the remainder of the protein . While the RON5 pro region is necessary for trafficking , it does not appear to be sufficient as C-terminal truncations of RON5N also result in mistargeting . The C-terminal region of RON5N is the most highly conserved portion of the protein , potentially suggesting that this region is critical for complex assembly in addition to trafficking . Importantly , a version of RON5 lacking the entire RON5C domain ( RON5Δ1258-1702 ) targets to the rhoptry necks but cannot rescue RON2 stability ( see below ) , showing that RON5 contains the necessary rhoptry neck targeting information independent of RON2 . Knockdown of RON5 demonstrates a critical importance in organizing the MJ RON complex . The specific impact on RONs 2 and 4 following depletion of RON5 provides experimental support for the idea that RONs 2/4/5 constitute a MJ core complex , consistent with their conservation across the phylum . In contrast , RON8 appears to represent a coccidial innovation that contains its own targeting information to facilitate sorting to the rhoptry necks . The simultaneous loss of RON2 upon RON5 knockdown may be due to Endoplasmic Reticulum-Associated Degradation ( ERAD ) quality control systems that sense misfolded proteins , extract them from the ER and target them to the proteasome [42] , [43] . This specific degradation of RON2 but not other MJ complex components in the absence of RON5 suggests that RON5 may directly bind RON2 ( although this could also be achieved indirectly through other complex components ) and ensure its proper folding or mask a RON2-encoded signal for ER retention and degradation similar to characterized protein complexes in other systems [42] . In contrast to RON2 , the soluble MJ component RON4 is not degraded but fails to target to the rhoptry necks in the absence of RON2/5 , indicating that RON4 contains targeting information to enter the rhoptries , but requires interaction with RON2/5 to ultimately reach the necks of the organelle . Little is known about the determinants for sub-domain trafficking within the rhoptries . Interestingly , a reverse scenario was observed for the Toxoplasma rhoptry body protein ROP1 and the P . falciparum rhoptry body protein RAP1 , each of which mistarget to the rhoptry neck following truncation of C-terminal residues [26] , [44] . Loss of RON2 and mistargeting of RON4 following RON5 knockdown indicates that RON5 serves an escorter role that is required for MJ core complex trafficking and integrity . A somewhat similar scenario was reported for a recently identified RON complex in Toxoplasma consisting of RON9 ( a predicted transmembrane protein ) and RON10 ( a predicted soluble protein ) , although in this case both partners are required as gross mistargeting ( but not degradation ) and total loss of rhoptry localization of each protein occurs in the absence of the other [40] . Thus , the stability and trafficking of protein complexes targeted to the rhoptry neck appears to be commonly achieved through the presence of escorters , as has been observed for certain micronemal proteins [45] , [46] . Blocking proteolytic separation of RON5 N and C requires mutation of two sites within RON5 , suggesting there is some selective pressure to maintain this processing event . Despite this fact and to our surprise , we find that N/C processing during maturation is not important for RON5 functions in MJ complex organization or parasite invasion . Therefore , while RON5C is necessary for RON2 stability , processing is not required for some structural rearrangement of these domains as one might infer from the presence of multiple sites at which N/C processing can occur . While a number of processing events have been characterized in rhoptry proteins and the suspected maturase TgSUB2 is thought to be essential , the known functional importance of processing is limited to the removal of N-terminal pro-domains which are involved in trafficking and no longer needed upon reaching their destination [26] , [28] , [47] . While the key importance of RON5 for invasion seemed to provide an excellent opportunity for determining the role of rhoptry protein processing beyond such trafficking functions , the lack of any effect on invasion when separation of RON5N/C is blocked suggests that cleavage may play an extremely subtle role in MJ complex function or that parasites can rapidly adapt when processing is blocked . Previously , peptides that interfere with the interaction of RON2 and AMA1 were found to block invasion but not evacuole formation [19] . Interestingly , we find that knockdown of RON5 ( and RON2 ) results in a marked decrease in evacuole formation . This indicates a critical role for the MJ RON complex in facilitating rhoptry secretion and suggests that rhoptry secretion proceeds in a stepwise fashion with deployment of the MJ RONs from the rhoptry neck occurring prior to secretion of rhoptry body contents . The fact that RON5 is present in the MJ suggests that in addition to its importance as an escorter ensuring stability and proper targeting of RONs 2 and 4 , RON5 may also serve direct roles in host cell penetration . Indeed , parasites that have been depleted of RON5 fail to establish an observable MJ and the rare penetration events that are observed appear to be supported by residual RON5 in these individual cells . However , at this point we cannot distinguish between invasion defects resulting directly from loss of RON5 , indirectly from loss of RON2 , or both . Furthermore , mistargeting of RON4 suggests that RON4-specific functions are likely also impaired in these conditions . Future work aimed at characterizing the potential direct interaction between RON5 and RON2 may provide insight to design new RON5 mutants that can stabilize RON2 and allow investigation of any RON5-specific roles in invasion . Additionally , RON2-specific knockdowns are needed that will allow RON2 function to be directly probed at the genetic level . In conclusion , our results highlight the importance of the MJ RON core complex for Toxoplasma invasion . These results are particularly significant given the recent finding that AMA1 disruption impacts adhesion but not penetration [24] . Taken together with the importance of RON4 for invasion by P . berghei sporozoites , these findings indicate a key role for the RON 2/4/5 complex in establishing the apicomplexan moving junction and facilitating host penetration [23] .
Antibodies were raised in rats under the guidelines of the Animal Welfare Act and the PHS Policy on Humane Care and Use of Laboratory Animals . Specific details of our protocol were approved by the UCLA Institutional Animal Care and Use Committee , known as the Chancellor's Animal Research Committee ( protocol # 2004-055-31C ) . T . gondii RHΔhpt ( parental ) strain and modified strains were maintained in confluent monolayers of human foreskin fibroblast ( HFF ) host cells as previously described [48] . The following Toxoplasma primary antibodies were used in IFA or Western blot: mouse polyclonal anti-RON5C [13] , polyclonal rat anti-RON11 [33] , rabbit anti-RON2 [25] , rat polyclonal anti-RON4 ( see below ) , rabbit anti-RON4 [11] , mouse polyclonal anti-RON8 [13] , anti-ROP2/3/4 mAb TA7 1A11 [49] , rabbit anti-ROP13 [50] , rabbit anti-SAG1 [51] , anti-ISP1 mAb 7E8 [52] , mouse polyclonal anti-ISP3 [52] , monoclonal mouse anti-F1-ATPase beta subunit 5F4 ( Bradley , unpublished ) , rabbit anti-proROP4 UVT70 [30] . Hemagglutinin ( HA ) epitope tags were detected with mouse mAb HA . 11 ( Covance ) , rabbit polyclonal anti-HA ( Invitrogen ) or rat mAb 3F10 ( Roche ) . MYC epitope tags were detected with mouse mAb 9E10 ( Neomarkers ) . FLAG epitope tags were detected with mouse anti-FLAG mAb M2 ( Sigma ) . For generation of rat anti-RON4 sera , a portion of the RON4 coding sequence comprising residues 85–983 was recombinantly expressed in E . coli BL-21DE3 cells and purified over Ni-NTA agarose ( Qiagen ) as previously described [53] . The resulting protein was injected into a rat for anti-sera production . Freshly lysed parasites were collected for Western blots . In time course analysis of protein levels , time points were designed to correspond with monolayer lysis . All parasite samples were counted on a hemocytometer to ensure equivalent loading between lanes . Fixation and immunofluorescence staining of T . gondii were carried out as previously described [52] . Image stacks were collected at z-increments of 0 . 2 µm with an AxioCam MRm CCD camera and AxioVision software on an Axio Imager . Z1 microscope ( Zeiss ) using a 100× oil immersion objective . Deconvolved images were generated using manufacturer specified point-spread functions and displayed as maximum intensity projections . The endogenous tagging vector p3xHA . LIC . DHFR [54] was first modified to replace the DHFR selectable marker cassette with a chloramphenicol acetyl-transferase ( CAT ) selectable marker between the restriction sites HindIII/XbaI resulting in the plasmid p3xHA . LIC . CAT . A portion of the genomic locus of RON5 up to but not including the stop codon was PCR amplified from Toxoplasma genomic DNA ( primers P1/P2 , Table S1 ) and inserted into p3xHA . LIC . CAT or p3xMYC . LIC . CAT by ligation-independent cloning [55] to generate the vectors pRON5-3xHA . LIC . CAT and pRON5-3xMYC . LIC . CAT . These plasmids were linearized with PstI and transfected into the TATiΔku80 parasite line [56] . Following selection with chloramphenicol , parasites were cloned by limiting dilution and a clone expressing the tagged protein of interest was isolated and designated RON5-3xHA or RON5-3xMYC . For direct replacement of the RON5 promoter with the conditional TetOSAG4 promoter by homologous recombination , 5′ ( primers P3/P4 ) and 3′ ( primers P5/P6 ) regions flanking the RON5 promoter were PCR amplified from Toxoplasma genomic DNA and cloned into the vector pDT7S4myc [56] between NdeI and BglII/AvrII sites , respectively . The resulting vector , pTS4-RON5-DHFR , was linearized with ApaI and transfected into RON5-3xMYC or RON5-3xHA parasites . Following selection with 1 µM pyrimethamine , parasites were cloned by limiting dilution and genomic DNA from individual clones was analyzed for RON5 promoter replacement ( primers P7/P8 ) . A clone that had undergone the intended recombination event was designated RON5MYCcKD or RON5HAcKD . For expression of a complementing second copy of RON5 , the RON5 promoter was PCR amplified from Toxoplasma genomic DNA ( primers P9/P10 ) and inserted into the UPRT targeting vector pUPRTKO-HA [57] between SpeI and BamHI by blunting both the digested vector and PCR amplicons , resulting in the vector pUPRTKO-RON5-promoter-HA . The full length RON5 coding sequence was PCR amplified from a Toxoplasma cDNA library ( primers P11/P12 ) and inserted into this vector between BglII/NotI sites to generate the vector pUPRTKO-RON5-HA . This vector was linearized with NsiI and transfected into RON5MYCcKD parasites followed by selection with 5 µg/ml 5-fluorodeoxyuridine to facilitate targeted replacement of the UPRT locus [58] . For site directed mutagenesis , a portion of the RON5 coding sequence between the restriction sites SmaI and NotI was digested from the vector pUPRTKO-RON5-HA and inserted into the cloning vector pJet1 . 2 ( Fermentas ) . Site-directed mutants were generated by Quick Change Mutagenesis ( Stratagene ) with mutagenesis primers as follows ( forward primer given , reverse complement was also used ) : SFVE>AGDR ( P13 ) and SFVQ>AGDR ( P14 ) . For expression of double tagged RON5 to monitor proRON5 , a FLAG epitope tag version of the vector pUPRTKO-RON5-HA was first generated by PCR amplifying the 3′ UTR with a forward primer encoding the FLAG epitope sequence ( primers P15/P16 ) and inserting this amplicon between NotI/EcoRV , replacing the inframe fusion to a C-terminal HA tag with a FLAG tag ( pUPRTKO-RON5-FLAG ) . A portion of the 5′ RON5 coding sequence was PCR amplified ( primers P11/P17 ) and inserted into the cloning vector pJet1 . 2 ( Fermentas ) . An HA epitope was then inserted into the RON5 coding sequence between residues 35 and 36 using Quick Change Mutagenesis ( P18 ) and this modified coding sequence was inserted into the vector pUPRTKO-RON5-FLAG between BglII/RsrII resulting in the vector pUPRTKO-RON5-PRO-HA-C-FLAG . For generation of an inframe deletion of proRON5 , a portion of the RON5 promoter and 5′ coding sequence was PCR amplified from the vector pUPRTKO-RON5-PRO-HA-C-FLAG with a reverse primer encoding a KasI site ( primer P19/P20 ) and re-inserted into this vector between NheI/AscI . A portion of the RON5 coding sequence was then PCR amplified ( primers P21/P22 ) and inserted between KasI/AscI , resulting in the vector pUPRTKO-RON5Δpro-N-HA-C-FLAG . For generation of C-terminal truncations of RON5 , truncated portions of the RON5 coding sequence were amplified ( Δ618-1702: P11/P23; Δ898-1702: P11/P24; Δ1019-1702: P11/P25; Δ1258-1702: P11/P26; Δ1476-1702: P11/P27 ) and inserted into the vector pUPRTKO-RON5-promoter-HA between BglII/NotI to generate the indicated C-terminal truncations . Parasites were grown 48 hrs −/+1 . 5 µg/ml Atc , syringe lysed and infected into 6-well dishes containing fresh , confluent HFF monolayers −/+ Atc . Cultures were allowed to grow nine days before fixation with methanol followed by staining with crystal violet . Invasion assays were performed as previously described [59] . Briefly , parasites were grown 72 hrs −/+1 . 5 µg/ml Atc , monolayers were washed with PBS and intracellular parasites were collected by scraping and passage through a 27-gauge needle . Equivalent parasite numbers were resuspended in pre-warmed media and allowed to infect HFF monolayers on coverslips for one hour . Monolayers were then washed , fixed with EM-grade 3 . 7% formaldehyde/PBS ( Biosciences , Inc . ) , blocked with PBS/3% BSA for 30 min and incubated with rabbit anti-SAG1 diluted in PBS/3%BSA for 1 hr . After washing , samples were permeabilized in PBS/3% BSA/0 . 1% Triton X-100 for 30 min and then incubated with mAb 5F4 diluted in PBS/3% BSA for one hour . Following incubation with secondary antibodies , samples were examined by fluorescence microscopy and parasites were scored as invaded ( SAG1− , 5F4+ ) or attached ( SAG1+ , 5F4+ ) . Invasion assays were performed in triplicate , five fields were counted on each replicate coverslip and the average number of invaded and attached parasites per field was calculated . Synchronized pulse invasion assays were performed as previously described with parasites that had been pre-treated 72 hrs −/+ Atc [60] . For evacuole assays , parasites were grown −/+1 . 5 µg/ml Atc for 24 hours , then infected into fresh HFF monolayers and allowed to grow an additional 36 hours −/+1 . 5 µg/ml Atc to allow large vacuoles to form . Intracellular parasites were collected by scraping and passage through a 27-gauge needle . Evacuole assays were then performed as previously described [61] . The number of evacuoles was counted blind across five fields per coverslip on three independent coverslips per sample and the average number per field was calculated . Egress assays were performed as previously described [62] . Briefly , parasites were grown −/+1 . 5 µg/ml Atc for 24 hours , then infected into fresh HFF monolayers on coverslips and allowed to grow an additional 36 hours −/+1 . 5 µg/ml Atc . Coverslips were then washed with PBS and incubated in 1 µM calcium ionophore A23187 ( Sigma ) diluted in Hank's Balances Salts Solution at 37°C before being fixed in methanol and processed for IFA with rabbit anti-SAG1 . At least 100 vacuoles per coverslip were counted across five fields on three independent coverslips per sample and scored as egressed or not egressed . For each of the above assays , experiments were repeated at least twice and values from a representative experiment are shown as the mean ± SD . RON5HAcKD parasites were grown 72 hrs −/+1 . 5 µg/ml Atc . Total parasite RNA was harvested with TRiZol ( Invitrogen ) , purified using a RNaEASY column ( Qiagen ) and used as to generate cDNA with the iSCRIPT kit ( BioRad ) . Relative amounts of RON5 ( P28/P29 ) and RON2 ( P30/P31 ) mRNA were quantified by qPCR using iQ Sybr Green ( kapaBiosystems ) and normalized to actin ( P32/P33 ) using ΔΔCt statistical analysis [63] .
|
Toxoplasma and related apicomplexan parasites are obligate intracellular pathogens that actively invade their host cells , creating a specialized vacuole within which the parasite is able to replicate . Invasion involves the establishment of a tight-junction interface between host and parasite membranes called the moving junction ( MJ ) through which the parasite actively penetrates the host . At the onset of invasion , a protein complex composed of RONs 2/4/5/8 is injected from specialized parasite secretory organelles called rhoptries into the host membrane . Following secretion , this RON complex localizes to the MJ throughout the invasion event and is thought to be the basis for this tight-junction . In this study , we utilize a conditional knockdown of RON5 to show that this MJ component , present at the cytosolic face of the host membrane during penetration , is crucial for invasion and for MJ complex organization . In particular , loss of RON5 results in degradation of RON2 and mistargeting of RON4 in the parasite , effectively ablating the MJ complex . We exploit this knockdown strain to evaluate RON5 processing and identify regions of the protein that are necessary for organizing the complex . Our findings demonstrate the key role of RON5 in facilitating apicomplexan host invasion and disease .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"parastic",
"protozoans",
"toxoplasma",
"gondii",
"microbial",
"pathogens",
"protozoology",
"host-pathogen",
"interaction",
"biology",
"microbiology",
"molecular",
"cell",
"biology",
"parasitology"
] |
2014
|
RON5 Is Critical for Organization and Function of the Toxoplasma Moving Junction Complex
|
Viruses interfere with and usurp host machinery and circumvent defense responses to create a suitable cellular environment for successful infection . This is usually achieved through interactions between viral proteins and host factors . Geminiviruses are a group of plant-infecting DNA viruses , of which some contain a betasatellite , known as DNAβ . Here , we report that Cotton leaf curl Multan virus ( CLCuMuV ) uses its sole satellite-encoded protein βC1 to regulate the plant ubiquitination pathway for effective infection . We found that CLCuMu betasatellite ( CLCuMuB ) βC1 interacts with NbSKP1 , and interrupts the interaction of NbSKP1s with NbCUL1 . Silencing of either NbSKP1s or NbCUL1 enhances the accumulation of CLCuMuV genomic DNA and results in severe disease symptoms in plants . βC1 impairs the integrity of SCFCOI1 and the stabilization of GAI , a substrate of the SCFSYL1 to hinder responses to jasmonates ( JA ) and gibberellins ( GA ) . Moreover , JA treatment reduces viral accumulation and symptoms . These results suggest that CLCuMuB βC1 inhibits the ubiquitination function of SCF E3 ligases through interacting with NbSKP1s to enhance CLCuMuV infection and symptom induction in plants .
Monopartite begomoviruses often possess an essential disease-specific betasatellite and are responsible for devastating diseases in many crops [1] . For example , at least six distinct begomoviruses that are associated with a single betasatellite , Cotton leaf curl Multan betasatellite ( CLCuMuB ) , cause Cotton leaf curl disease ( CLCuD ) , which is a major constraint to cotton production in Asia [2] . Cotton leaf curl Multan virus ( CLCuMuV ) is one of these begomoviruses and can infect cotton and many other plants including Nicotiana benthamiana . CLCuMuV consists of a circular single-stranded DNA genome that encodes only 6 proteins ( V1 and V2 in virion-sense strand whilst C1 , C2 , C3 and C4 in virion complementary-sense strand ) . CLCuMuB is a small circular single-stranded DNA molecule that is essential for CLCuMuV to induce disease symptoms in plants [3] . Betasatellites , such as CLCuMuB , are approximately half the size of the begomovirus DNA genomes . They require the helper begomoviruses for replication and movement in plants and only encode a single multifunctional pathogenicity protein βC1 [1] . βC1 can up-regulate the proliferation of its cognate helper virus [4] , and complement the movement function encoded by the DNA B component of some bipartite begomoviruses [5] . βC1 is essential for producing viral disease symptoms [4 , 6–12] and plays important roles in suppression of transcriptional ( TGS ) [13] and posttranscriptional gene silencing ( PTGS ) [14–18] . Furthermore , βC1 can also promote the performance of the whitefly and impair plant development [19–22] . More details about the multiple functions of βC1 can be found in recently published reviews [1 , 23] . However , how geminiviruses exploit βC1 to perform these diverse functions needs further investigations . Ubiquitination is a highly dynamic posttranslational modification process that is a major protein degradation and rapid regulatory mechanism in plants [24] . Through the action of a sequential cascade of three enzymes consisting of E1 , E2 , and E3 , ubiquitin is covalently attached to substrate proteins , and then , in most cases , the polyubiquitinated proteins will be degraded by the 26S proteosome . As the most abundant member of the E3 family , the SKP1/CUL1/F-box ( SCF ) complex is the best characterized multi-subunit ubiquitin ligase . In the SCF complex , SKP1/ASK1 ( S-phase kinase-associated protein ) acts as a bridge between CUL1 ( Cullin1 ) and F-box proteins . CUL1 is the major structural scaffold and F-box proteins are responsible for recognizing target substrates . RBX1 is the fourth subunit that is heterodimerized with CUL1 , and binds E2 through its RING Finger domain . More than 700 predicted F-box proteins are encoded by the Arabidopsis thaliana genome , suggesting these F-box proteins have highly targeting potentials for extensive regulatory functions [25 , 26] . The SCF complex-based E3 ubiquitin ligases have been known to regulate plant hormone signaling . Several phytohormone receptors are F-box proteins in SCF complexes , such as SCFTIR1 for auxin , SCFCOI1 for jasmonates , SCFSLY1/GID2 for gibberellins and SCFMAX2 for strigolactones [27–30] . In addition , SCF complexes regulate ethylene ( ET ) signal transduction at multiple points ( SCFETP1 and SCFETP2 for EIN2 , SCFEBF1 and SCFEBF2 for EIN3 ) [31 , 32] . Since phytohormones have pivot functions in vegetative growth , compromising of these pathways is usually accompanied by abnormal developmental phenotype . Among them , JA plays a crucial role in defense against pathogens and insects . Recently , JA pathway was reported to be involved in plant defense against geminivirus infection [33] . In this study , we report that a geminivirus uses its satellite-encoded βC1 to interfere with the ubiquitination function of SCF E3 ligases to enhance viral infection and symptom development in plants .
CLCuMuB was reported to enhance DNA accumulation of the helper virus and be necessary for producing viral disease symptoms [4] . To see whether βC1 is responsible for these functions , we constructed a null mutant betasatellite for the βC1 gene [34] with a ATG-TGA transition in the start codon , hereafter called βM1 ( S1 Fig ) . Different from N . benthamiana plants infected with CLCuMuV and β ( CA+β ) causing severe downward leaf curling and darkening as well as swollen veins , plants infected with CLCuMuV and βM1 ( CA+βM1 ) grew taller , developed much milder symptoms and accumulated much less CLCuMuV genomic DNA ( S2A and S2B Fig ) . Further , we generated transgenic N . benthamiana plants expressing non-tagged or tagged βC1 . However , most transgenic plants have very severe symptoms and are infertile or dead finally . Nevertheless , we were able to obtain five lines expressing non-tagged βC1 under control of its native promoter ( βC1pro:βC1 ) , 2 lines expressing GFP-tagged βC1 driven by CaMV 35S promoter ( 35Spro:GFP-βC1 ) and 4 lines expressing HA-tagged βC1 driven by CaMV 35S promoter ( 35Spro:HA-βC1 ) . All these transgenic plants showed aberrant development phenotype ( S3 Fig ) . Taken together , these results suggest that CLCuMuB βC1 is required for development of typical disease symptoms and enhancement of CLCuMuV DNA accumulation . To understand how CLCuMuB βC1 facilitates virus infection , we used CLCuMuB βC1 as bait in a yeast two-hybrid ( Y2H ) system [35] to identify host CLCuMuB βC1-interacting proteins . From screening the Solanum lycopersicum cDNA library , we characterized a full-length SKP1-like protein ( designated as SlSKP1 ) that interacted with βC1 . Furthermore , 12 putative NbSKP1 homologues identified in the N . benthamiana genome through bioinformatics analysis ( http://solgenomics . net ) , encode proteins with more than 44% amino-acid identity to SlSKP1 . However , we obtained only 4 predicted cDNAs by RT-PCR . Indeed , RNA-seq results ( ftp://ftp . solgenomics . net/transcript_sequences/by_species/Nicotiana_benthamiana/ ) indicates that other 8 putative homologues are not or rarely expressed in leaf tissues . Three of the 4 NbSKP1 homologues NbSKP1 . 1 , NbSKP1 . 2 and NbSKP1 . 3 , collectively called NbSKP1s , interact with CLCuMuB βC1 , whilst the other do not or interact very weakly with βC1 in yeasts and it is named as NbSKP1L1 ( NbSKP1-like 1 ) ( Fig 1A ) . NbSKP1 . 1 shares 95 . 5% , 91 . 7% and 44 . 9% amino-acid identity to NbSKP1 . 2 , NbSKP1 . 3 and NbSKP1L1 , respectively ( S4 Fig ) . To examine whether CLCuMuB βC1 directly interacts with NbSKP1 . 1 , in vitro GST pull-down assay was performed . His-HA double-tagged NbSKP1 . 1 ( His-HA-NbSKP1 . 1 ) was expressed in E . coli BL21 ( DE3 ) and then purified by Ni-NTA Agarose ( Qiagen , Netherlands ) column . After elution , His-HA-NbSKP1 . 1 was incubated with Glutathione Sepharose 4B ( GE , American ) bonded with E . coli-expressed GST , GST-tagged CLCuMuB βC1 ( GST-βC1 ) or its mutant with the deletion of C-terminal 43 amino acids ( GST-βC1ΔC43 ) . His-HA-NbSKP1 . 1 was pulled down by GST-βC1 but not GST and GST-βC1ΔC43 ( Fig 1B ) , indicating that NbSKP1 . 1 can directly interact with βC1 . To our surprise , His-HA double-tagged NbSKP1L1 ( His-HA-NbSKP1L1 ) was also pulled down by GST-βC1 but not GST and GST-βC1ΔC43 . ( Fig 1B ) . We also demonstrated in planta interaction of CLCuMuB βC1 with NbSKP1 . 1 using co-immunoprecipitation ( Co-IP ) assay . In this assay , HA-tagged NbSKP1 . 1 ( HA-NbSKP1 . 1 ) was co-expressed transiently with GFP or GFP-tagged CLCuMuB βC1 ( GFP-βC1 ) in N . benthamiana by agroinfiltration . GFP-βC1 transgenic N . benthamiana exhibits leaf curl symptoms , which indicates GFP-βC1 is a functional protein ( S3C Fig ) . Total protein extracts were immunoprecipitated by GFP-Trap beads ( ChromoTek , German ) . The resulting precipitates were analyzed by western blot assays using an anti-HA antibody ( CST , USA ) . We found that HA-NbSKP1 . 1 was co-immunoprecipitated by GFP-βC1 but not GFP ( Fig 1C ) . Similarly , we also found that HA-tagged NbSKP1L1 ( HA-NbSKP1L1 ) was co-immunoprecipitated by GFP-βC1 but not GFP ( Fig 1C ) . To confirm these Co-IP results , we performed the reverse IP . GFP-βC1 was co-expressed transiently with HA-tagged GUS ( HA-GUS ) , HA-NbSKP1 . 1 or HA-NbSKP1L1 in N . benthamiana by agroinfiltration . Total protein extracts were immunoprecipitated by HA-beads ( Abmart , China ) . The resulting precipitates were analyzed by western blot assays using an anti-GFP antibody ( ChromoTek , German ) . GFP-βC1 was pulled down by HA-NbSKP1 . 1 and HA-NbSKP1L1 but not HA-GUS ( S5A Fig ) . To find where CLCuMuB βC1 interacts with NbSKP1 . 1 and NbSKP1L1 in plant cells , citrine yellow fluorescent protein ( YFP ) -based bimolecular fluorescence complementation ( BiFC ) assays [36] were performed . HA-tagged βC1 or βC1ΔC43 was fused to the N-terminal domain of YFP ( nYFP ) to generate HA-βC1-nYFP or HA-βC1ΔC43-nYFP . NbSKP1 . 1 , NbSKP1L1 and the N-terminal fragment of firefly luciferase ( nLUC ) as a negative control were fused to HA-tagged C-terminal domain of YFP ( HA-cYFP ) to generate HA-cYFP-NbSKP1 . 1 , HA-cYFP-NbSKP1L1 and HA-cYFP-nLUC . Western blot assays using an anti-HA antibody showed that all chimeric proteins can be expressed correctly ( S5B Fig ) . HA-βC1-nYFP or HA-βC1ΔC43-nYFP was transiently co-expressed with HA-cYFP-NbSKP1 . 1 , HA-cYFP-NbSKP1L1 or HA-cYFP-nLUC respectively in N . benthamiana . No such interaction between HA-βC1-nYFP and HA-cYFP-nLUC was found . However , positive interactions between HA-βC1-nYFP and HA-cYFP-NbSKP1 . 1 or HA-cYFP-NbSKP1L1 were observed in both nucleus and cell periphery , as indicated by occurrence of yellow fluorescence ( Fig 1D ) . As a control , HA-βC1ΔC43-nYFP didn’t interact with HA-cYFP-NbSKP1 . 1 or HA-cYFP-NbSKP1L1 ( Fig 1D ) . Taken together , these results demonstrate that NbSKP1s and NbSKP1L1 interact with CLCuMuB βC1 both in vitro and in vivo , and the interaction of the two proteins occurs in nucleus and cell periphery of plant cells . The crystal structures of human SKP1 [37] and Arabidopsis ASK1 [38] suggest that SKP1 can be divided into N-terminal and C-terminal domains . The N-terminal BTB-POZ domain of SKP1 is responsible for its binding to CUL1 whilst its C-terminal domain is thought to be essential for SKP1 to interact with F-box proteins . The Y2H assays showed that CLCuMuB βC1 interacted with the first 98 amino-acid N-terminal region of NbSKP1 . 1 ( N98aa ) , but not with the C-terminal region ( aa 99–155 ) of NbSKP1 . 1 ( C57aa ) , as indicated by growth of yeast on Leu− plates containing galactose ( Gal ) and raffinose ( Raf ) ( Fig 2 ) . In human and Arabidopsis , SKP1/ASK1 interacts with CUL1 to assemble into SCF complexes through its N-terminal domain [37 , 38] . We found that CLCuMuB βC1 interacts with N-terminal domain of NbSKP1 . 1 ( Fig 2 ) . This prompted us to investigate whether CLCuMuB βC1 interferes with the assembly of NbSKP1 . 1 into the SCF complex . To test this hypothesis , GFP competitive pull-down assay was performed . Because E . coli-expressed NbCUL1 was insoluble , GFP and GFP-tagged NbCUL1 ( GFP-NbCUL1 ) were expressed in N . benthamiana , then precipitated by GFP-Trap beads . To eliminate the influence from endogenous NbSKP1s and NbSKP1L1 , an excessive amount of E . coli-expressed His-HA-NbSKP1 . 1 was used to saturate the beads and endogenous NbSKP1s and NbSKP1L1 were crowded out from GFP-NbCUL1 , then the supernatant was removed . After an increasing amount of E . coli-expressed His-tagged βC1 ( His-βC1 ) was added , more and more His-HA-NbSKP1 . 1 was pulled off from GFP-NbCUL1 . and levels of His-HA-NbSKP1 . 1 released into the supernatant were increased ( Fig 3A ) . Further , we confirmed CLCuMuB βC1 interfering with the interaction between NbSKP1 . 1 and NbCUL1 by BiFC assays . We generated nYFP-NbSKP1 . 1 and cYFP-NbCUL1 fusion constructs and co-expressed them with HA-tagged nLUC ( HA-nLUC ) or HA-tagged CLCuMuB βC1 ( HA-βC1 ) in N . benthamiana . HA-βC1 is a functional protein ( S3D–S3F Fig ) . Stronger signals were detected for the combination of nYFP-NbSKP1 . 1 and cYFP-NbCUL1 in the presence of HA-nLUC than in the presence of HA-βC1 ( Fig 3B and 3C ) . Meanwhile , the protein level of nYFP-NbSKP1 . 1 and cYFP-NbCUL1 seem similar between the two groups ( Fig 3D ) . These data suggest that CLCuMuB βC1 interferes with the interaction between NbSKP1 . 1 and NbCUL1 via binding to NbSKP1 . 1 . βC1 but not βC1ΔC43 interacts with NbSKP1s and NbSKP1L1 . Meanwhile βC1 but not βC1ΔC43 induces viral symptoms ( S6 Fig ) . These results promote us to check whether silencing NbSKP1s can produce some viral symptoms . We constructed a deletion mutant betasatellite by replacing the entire βC1 gene from CLCuMuB with sites of two restriction enzymes AscI and XbaI to generate CLCuMuB ( ΔβC1 ) , hereafter called βM2 ( S1 Fig ) . We guessed that our CLCuMuB-based vector βM2 may be used as a VIGS vector . To confirm this , we cloned a N . benthamiana phytoene desaturase ( NbPDS ) gene fragment into βM2 to generate βM2-PDS . Photo-bleach phenotype was observed around the leaf veins of N . benthamiana plants agroinoculated with βM2-PDS in the presence of helper virus CLCuMuV ( S7 Fig ) . This result demonstrates that βM2 can be used as a CLCuMuB-based VIGS vector to effectively silence genes , and CLCuMuV may exhibit a phloem limitation . To investigate the role of NbSKP1s in CLCuMuV infection , we silenced NbSKP1s using our CLCuMuB-based VIGS vector , βM2 . To exclude the effect from size , three cDNA fragments corresponding to the 176-bp , 184-bp and 345-bp NbSKP1 . 1 sequences were fused with 169-bp , 161-bp and 0-bp βC1 sequences respectively and then were cloned into βM2 to generate βM2-SKP1F1 , βM2-SKP1F2 and βM2-SKP1F3 ( Fig 4A1–4A3 ) . A 345-bp fragment of βC1 was inserted into βM2 to generate βM2-βC1F as the control . The position relationship among 176-bp , 184-bp and 345-bp NbSKP1 . 1 fragments was shown in S8 Fig . N . benthamiana plants were agroinfiltrated with CLCuMuV ( CA ) and βM2-βC1F , βM2-SKP1F1 , βM2-SKP1F2 or βM2-SKP1F3 . Silencing of NbSKP1s resulted in an increasing accumulation of CLCuMuV DNA at 14 dpi ( Fig 4B1–4B3 ) . Since the mRNA level of NbSKP1L1 was very low in normal plants ( S9 Fig ) , and similar results can be found in the RNA-seq data of N . benthamiana in Sol Genomics Network ( ftp://ftp . solgenomics . net/transcript_sequences/by_species/Nicotiana_benthamiana/ ) , we gave up to check the mRNA level of NbSKP1L1 . Silencing of NbSKP1s ( NbSKP1 . 1 , NbSKP1 . 2 and NbSKP1 . 3 ) was triggered by all three constructs , and the levels of NbSKP1s mRNA were significantly reduced when compared to the βM2-βC1F control ( Fig 4C1–4C3 ) . βM2-SKP1F3 was more effective than βM2-SKP1F1 and βM2-SKP1F2 to cause silencing of NbSKP1s ( Fig 4C1–4C3 ) . At 21 dpi , 50% plants infected with CA+βM2-SKP1F1 , 50% plants infected with CA+βM2-SKP1F2 and 100% plants infected with CA+βM2-SKP1F3 exhibited severe downward leaf curling and darkening as well as swollen veins , typical symptoms in plants infected by CA+β ( Fig 4D1–4D3 ) . If we continue to observe the symptom development , growth retardation will also be found ( S10 Fig ) . We also confirmed the effect of silencing NbSKP1s on CLCuMuV accumulation and symptoms using another control βM2-GFPF , which 345-bp GFP fragment was cloned into βM2 . N . benthamiana plants were agroinfiltrated with CLCuMuV ( CA ) and βM2-GFPF or βM2-SKP1F3 . We found again that silencing of NbSKP1s enhances CLCuMuV DNA accumulation and results in viral symptoms ( S11 Fig ) . TYLCCNB-based VIGS works mainly in vascular tissues [39] , the tissues which CLCuMV tends to be limited to [40] . We further confirmed the effect of silencing NbSKP1s on CLCuMuV infection by TYLCCNB-based VIGS system [39] . We inserted the 345-bp GFP fragment and the 345-bp SKP1F3 fragment into pBinPLUS-2mβ of TYLCCNB-based VIGS system [39] , then agroinoculated them respectively with TYLCCNV for silencing . Similarly , silencing of NbSKP1s enhanced CLCuMuV DNA accumulation and 100% NbSKP1s silenced plants exhibited viral symptoms ( S12 Fig ) . We have demonstrated that βC1 is able to interfere with the interaction between NbSKP1s and NbCUL1 ( Fig 3 ) . Moreover , silencing of NbSKP1s has a dramatic influence on viral DNA accumulation and symptom development ( Fig 4 ) . We therefore investigated whether silencing of NbCUL1 could also enhance CLCuMuV DNA accumulation and cause severe viral symptoms . Two cDNA fragments corresponding to the 268-bp and 345-bp sequences of NbCUL1 were fused with 77-bp and 0-bp βC1 sequences respectively and then were cloned into βM2 to generate βM2-CUL1F1 and βM2-CUL1F2 respectively ( Fig 5A1 and 5A2 ) . The position relationship among 268-bp , and 345-bp NbCUL1 fragments were shown in S8 Fig . These two VIGS vectors along with CLCuMuV were then agroinfiltrated respectively into N . benthamiana plants . Silencing of NbCUL1 by either CA+βM2-CUL1F1 or CA+βM2-CUL1F2 resulted in an higher accumulation of CLCuMuV DNA ( Fig 5B1 and 5B2 ) and severer viral symptoms ( Fig 5D1 and 5D2 ) . Taken together , these results suggest that βC1 may enhance its helper geminivirus’ accumulation and viral symptom induction by interfering with the interaction between SKP1 and CUL1 through its binding to SKP1 . Because βC1 interferes with the interaction between SKP1 and CUL1 , and cul1 mutants are altered in JA responses [41 , 42] , we tested whether βC1 can interfere with JA pathways . First , we evaluated root growth rate in HA-βC1 transgenic plants , the root length of 6-day-old seedlings was measured every 24 h for 5 days . Data showed that HA-βC1 transgenic roots grow more slowly than wild-type roots ( Fig 6A ) . Meanwhile , we measured inhibition of primary root elongation caused by treatment with methyl-jasmonate ( MeJA ) , and HA-βC1 transgenic plants showed less sensitivity than wild-type plants to 50 μM MeJA ( Fig 6B ) . Further , quantitative real-time PCR was used to quantify the mRNA level of marker genes for JA responses . Three genes: Defensin-like protein 1 , Defensin-like protein 2 and Pathogen like protein were chosen for JA responses . Compared to wild-type plants , all three markers genes showed lower mRNA expression level in two independent HA-βC1 transgenic lines ( #2 HA-βC1 and #3 HA-βC1 ) ( Fig 6C ) . Auxin and gibberellins signalings are also regulated by CUL1-based SCF ubiquitin E3 ligases [27 , 29] . Real-time PCR assays showed lower mRNA expression level of their marker genes ( Gibberellin-regulated protein 14 and Gibberellin-regulated protein 6 for gibberellins , SAUR14 and PID for auxin ) in HA-βC1 transgenic lines than in wild-type controls ( S13A and S13B Fig ) . Taken together , CLCuMuB βC1 can really cause deficient function in SCF complexes and interfere with hormone signaling pathways . SCFCOI1 is the receptor for JA , and some geminiviruses interfere with JA pathway [20 , 21 , 33 , 43 , 44] . Meanwhile CLCuMuB βC1 seems to have no inhibition on jasmonates biosynthesis according to JA level data measured by mass spectrum and HPLC . Regardless of being wounded or not , plants infected with CA+β showed higher JA level compared to plants infected with CA+βM1 or healthy plants ( S14 Fig ) . These results imply that CLCuMuB βC1 doesn’t impair JA biosynthesis . Higher JA level in plants infected with CA+β may be derived from the feedback due to the impaired JA signaling . The stability of JA receptor COI1 , a F-box protein , is dependent on an intact SCFCOI1 complex [45] . Because βC1 can interfere with the interaction between SKP1 and CUL1 , we assumed that it may reduce the stability of COI1 in vitro . Co-IP analysis indicated that GFP-CUL1 associated with both Myc-COI1 and HA-NbSKP1 . 1 ( S15 Fig ) , suggesting that Myc-COI1 can be integrated within SCF complexes . After Myc-COI1 was transiently expressed in N . benthamiana and purified with anti-Myc affinity beads . Myc-COI1 protein was then mixed with total protein extracts prepared from N . benthamiana which was transiently expressed HA-βC1 or HA-nLUC . The stability of Myc-COI1 was assessed by western blot assays after the treatment at 25°C for various periods of time up to 8 h . The Myc-COI1 protein degraded more rapidly in HA-βC1 extracts compared to in HA-nLUC extracts ( Fig 6D and 6E ) . Moreover , the accumulation of Myc-COI1 in HA-βC1 transgenic lines was reduced 84–92% compared to that in wild-type plants ( WT ) ( S16 Fig ) , whilst the accumulation of GFP ( as an expression control ) in HA-βC1 transgenic lines was reduced by 26–41% in WT plant ( S16 Fig ) . Taken together , these data implied that CLCuMuB βC1 damages the integrity of SCFCOI1 complex to hinder JA responses . GA releases the brakes of plant growth . During this process , DELLA protein GAI is ubiquitinated by the SCFSLY1 and eventually degradated by the 26S proteasome [46] . Mutant plants that are deficient in GA pathways exhibit a dwarf phenotype [46] . Further , plants infected with CA+β is dwarf compared to plants infected with CA+βM1 ( S2 Fig ) . To check whether the function of SCFSLY1 is hindered by CLCuMuB βC1 , we co-expressed YFP-GAI with either HA-βC1 or HA-nLUC to investigate its degradation as described [33] . At 48 hpi , YFP-GAI fluorescence was observed in the nuclei 48 hpi ( Fig 7A ) , indicating YFP-GAI can be co-expressed with HA-βC1 or HA-nLUC normally in N . benthamiana leaves . However , whether plants were treated with 100 μM GA3 or not , YFP-GAI fluorescence was enhanced when co-expressed with HA-βC1 ( Fig 7A ) . Western blot assays using an anti-GFP antibody indicated that YFP-GAI accumulation was less in plants co-expressed with HA-nLUC than those co-expressed with HA-βC1 ( Fig 7A ) . Meanwhile , co-expression with HA-βC1 or HA-nLUC did not significantly affect mRNA level of YFP-GAI at this time point ( Fig 7B ) . Moreover , co-expression of HA-βC1ΔC43 did not enhance YFP-GAI accumulation ( S17 Fig ) . As an internal control , a GFP expression construct was coinfiltrated with HA-βC1 or HA-nLUC expression construct . No significant differences in GFP fluorescence or GFP protein accumulation were detected between them ( Fig 7C ) . Taken together , these results indicate that CLCuMuB βC1 can increase the accumulation of GAI by hindering its degradation to hinder GA responses . βC1 interferes with SCF function to enhance geminivirus DNA accumulation and damages the integrity of SCFCOI1 complex to hinder JA responses . This would suggest that JA is likely to be involved in plant defense against CLCuMuV . To test this hypothesis , we inoculated CLCuMuV along with CLCuMuB into MeJA or mock-treated N . benthamiana plants . Symptoms were daily monitored from 9 to 14 dpi . We found that application of exogenous MeJA resulted in milder symptoms ( Fig 8A–8E ) and lower viral DNA accumulation ( Fig 8F ) . These results demonstrate that MeJA could compromise viral pathogenicity . We also inoculated CLCuMuV along with βM1 into MeJA or mock-treated N . benthamiana plants . Real-time results show no difference on viral DNA accumulation between the two kinds of treatment ( Fig 8G ) . Thus , βC1 may enhance geminivirus infection , at least partially by inhibiting JA pathway through interfering with the function of SCFCOI1 .
Betasatellites are indispensable for some monopartite geminiviruses to induce viral symptoms in host plants . The sole protein βC1 encoded by several betasatellites , has been reported to be responsible for this phenomenon [1] . However , how βC1 induces viral symptoms remain obscure . CLCuMuB βC1 was previously reported to interact with a tomato ubiquitin conjugating enzyme ( UBC ) , SlUBC3 , by its C-terminal myristoylation-like motif [22] . The myristoylation-like motif only exists in CLCuMuB βC1 and its close relative okra leaf curl betasatellite ( OLCB βC1 ) . However , OLCB βC1 does not interact with SlUBC3 [22] . Further , silencing of UBC3 in N . benthamiana did not cause any obvious phenotype and enhanced viral DNA accumulation in this study ( S18 Fig ) . Thus , it is possible that symptoms induced by CLCuMuB might not be mediated by interaction between βC1 proteins and host UBC3 enzyme . Here , we demonstrate that CLCuMuB βC1 is also indispensable for symptom production ( S2 Fig ) . Through a series of interaction assays , we found that CLCuMuB βC1 interacts with NbSKP1s , important components of SCF complexes ( Fig 1 ) . Further , CLCuMuB βC1 interferes with the interaction between SKP1 and CUL1 ( Fig 3 ) to impair the function of SCF complexes , such as SCFCOI1 and SCFSYL1 ( Figs 6 and 7 ) , which is consistent with the previous observation that overexpression of CLCuMuB βC1 in tobacco causes a global reduction of polyubiquitinated proteins [22] . We found that disrupting the function of SCF complexes by silencing of either SKP1 or CUL1 leads to some typical virus symptoms , such as severe leaf curling , crimping , leaf darkening and growth retardation ( Figs 4 and 5 ) . Indeed , perturbation of the ubiquitin system can cause leaf curling and vascular tissue abnormalities [47] . Further , overexpression of CLCuMuB βC1 blocked the degradation of GAI ( Fig 8 ) , the target of the SCFSLY1 , repressed plant responses to GA , which may explained why the presence of CLCuMuB make plant dwarf phenotype . These results suggest that some geminiviral βC1 proteins can elicit viral symptoms by disrupting the plant ubiquitination pathway by interfering with SKP1-CUL1 interaction through its interaction with SKP1 . Although NbSKP1s silencing is in fact causing higher accumulation of viral DNA ( Fig 4B1–4B3 ) , the symptoms seem simply due to NbSKP1s silencing but not higher accumulation of virus , because we found higher accumulation of CLCuMuV DNA , but no symptom in plants infected with CLCuMuV and βM2-SKP1-176 which is generated though inserting the 176-bp NbSKP1 . 1 fragment directly into βM2 , without fused with the 169-bp βC1 fragment ( S19 Fig ) . We noticed that silencing of either SKP1 or CUL1 did not produce all symptoms caused by CLCuMuB βC1 . Besides leaf curling , crimping , darkening and growth retardation caused by silencing of either SKP1 or CUL1 , the viral symptoms elicited by CLCuMuB βC1 also include bending shoot and enations from abaxial side of leaves . Tomato yellow leaf curl China virus ( TYLCCNV ) βC1 was reported previously to elicit leaf morphological changes in Arabidopsis by mimicking the functions of ASYMMETRIC LEAVES 2 through its interaction with ASYMMETRIC LEAVES 1 and by repressing the accumulation of miR165/166 to subvert leaf polarity [20] . Meanwhile , suppression of miR165/166 can cause enations from abaxial side of leaves [48] . It is possible that CLCuMuB βC1 induces enations by suppression of miR165/166 . Further , TYLCCNV βC1 may also induce viral symptoms by up-regulating the expression of a calmodulin-like protein ( rgsCaM ) [16] . Considering that geminivirus βC1 is a multiple functional protein , CLCuMuB βC1 may contribute to the viral symptoms by multiple mechanisms including disrupting the plant ubiquitination pathway . In this study , we demonstrate that CLCuMuB βC1 impairs the interaction between NbSKP1s and NbCUL1 by interacting with NbSKP1s and silencing of either NbSKP1s or NbCUL1 enhances CLCuMuV DNA accumulation . Deletion of CLCuMuB βC1 reduced CLCuMuV titer ( S2 Fig ) . Silencing of either NbSKP1s or NbCUL1 caused enhanced virus accumulation ( Figs 4 and 5 ) . Geminiviruses may interfere with plant ubiquitination to suppress plant defense against geminivirus infection [49] . It has been reported that V2 protein of Tomato yellow leaf curl Sardinia virus ( TYLCSV ) interacts with UBA1 , a ubiquitin-activating enzyme , which is a positive regulator of plant defense [50 , 51] , and silencing of either UBA1 or RHF2a ( RING-type E3 ubiquitin ligase ) in N . benthamiana enhances TYLCSV infection [50 , 52] . Geminiviral C4 activates expression of host RING E3 ligase RKP to ubiquitinate cell cycle inhibitors ICK/KRPs to help the replication of Beet severe curly top virus ( BSCTV ) via promoting cell division [53 , 54] . However , how geminivirus βC1 proteins interfere with plant ubiquitination pathway to enhance viral accumulation is still obscure . In this study , we found that CLCuMuB βC1 disrupted the integrity of SCFCOI1 ( Fig 6D and 6E ) . Meanwhile CLCuMuB βC1 does not inhibit JA biosynthesis ( S14 Fig ) . More importantly , JA treatment reduces the plant susceptibility to CLCuMuV ( Fig 8 ) , which is consistent with the previous observation that JA treatment attenuates the infection of plant with Beet curly top virus ( BCTV ) [33] . TYLCCNB βC1 was reported to suppress JA-related host defenses for increasing population densities of their whitefly vectors [19 , 21] . Further , Cabbage leaf curl virus ( CaLCuV ) infection can also repress JA response [21 , 44] . The C2 proteins of TYLCSV , Tomato yellow leaf curl virus ( TYLCV ) and BCTV were reported to impair derubylation of SCF E3 ligase complexes and inhibit jasmonate signaling by interacting with CSN5 [20 , 33] . Thus , CLCuMuB βC1 could enhance CLCuMuV accumulation , at least partially by repressing JA responses through interfering with plant ubiquitination . We observed that the levels of CLCuMuV DNA in SKP1- or CUL1-silenced plants were lower than that in the presence of CLCuMuB with functional βC1 although silencing of either SKP1 or CUL1 resulted in a higher accumulation of CLCuMuV DNA ( Figs 4 and 5 and S2 ) . It has been reported that knock-down of either CSN5A or CSN3 , two components of protein degradation-related CSN complexes , hinders BCTV infection although knockout of Arabidopsis csn5a mutant can partially complement BCTV C2 mutant [50 , 52 , 55] . Further , overexpression of a given F-box protein can circumvent the general SCF malfunction [56 , 57] . These observations suggest that begomoviruses might not only hamper , but also redirect the activity of SCF complexes for begomoviruses propagation [33] . Very recently , ubiquitination is reported to regulate the stability of TYLCCNV βC1 [58] . Thus , host plants , geminiviruses and their satellites may have evolved to exploit the dual roles of the ubiquitination pathway in plant defense and viral pathogenesis to co-survive in their long-term arm races .
The full-length infectious CLCuMuV clone contains 1 . 7-mer CLCuMuV DNA genome . Two separate DNA fragments were PCR amplified using primer pairs HindIII-A-F/XbaI-A-R , or XbaI-A-F/KpnI-A-R respectively and total DNA extracted from cotton leaf tissues with CLCuD [34] as the template , double-digested with HindIII and XbaI or XbaI and KpnI , and then inserted into pBinplus ARS digested with HindIII and KpnI . The βDNA infectious clone contains 2-mer CLCuMuB genomes . Two DNA fragments were PCR amplified using primer pairs KpnI-β-F/HindIII-β-R or HindIII-β-F/SacI-β-R respectively and total DNA from cotton samples with CLCuD [34] as the template , digested with KpnI and HindIII or HindIII and SacI , and then inserted into pCAMBIA-2300 digested with KpnI and SacI to generate βDNA . The null mutant betasatellite vector βM1 was constructed by introducing a ATG-TGA transition in the start codon . βDNA was used as the template . Two DNA fragments were PCR amplified using primer pairs βM1-R/SacI-β-R or HindIII-β-F/βM1-F respectively , then were fused to obtain SacI-βM1-HindIII with ATG-TGA mutation . the other two DNA fragments were PCR amplified using primer pairs HindIII-β-F/βM1-R and βM1-F/KpnI-β-F , then were fused to obtain HindIII-βM1-KpnI with ATG-TGA mutation . digested with SacI and HindIII or HindIII and KpnI , SacI-βM1-HindIII and HindIII-βM1-KpnI were inserted into pCAMBIA-2300 digested with KpnI and SacI to generate βM1 . The T-DNA silencing vector βM2 was constructed by introducing a multiple cloning site to replace the βC1 ORF in CLCuMuB . Two DNA fragments were PCR amplified using primer pairs KpnI-βMF/XbaI-βM2-R or XbaI-βM2-F/SacI-βM2-R respectively using βDNA as the template , digested by KpnI and XbaI or XbaI and SacI , and then inserted into pCAMBIA-2300 digested by KpnI and SacI to generate vector βM2 . DNA fragments of HA-βC1-nYFP , HA-βC1ΔC43-nYFP , HA-cYFP-NbSKP1 . 1 , HA-cYFP-NbSKP1L1 , HA-cYFP-nLUC , GFP-βC1 , HA-βC1 , HA-βC1ΔC43 , HA-NbSKP1 . 1 , GFP-NbCUL1 , nYFP-SKP1 , cYFP-NbCUL1 , Myc-COI1 and YFP-GAI were obtained by overlapping PCR . The resulting PCR products were cloned between the duplicated Cauliflower mosaic virus 35S promoter and Nos terminator of pJG045 , a pCAMBIA1300-based T-DNA vector [59] . βC1pro:βC1 , a βC1expression vector with its native promoter , was generated by inserting 1–1346 nt of CLCuMuB genome ( GQ906588 ) into pCAMBIA-2300 . Among these vectors , βC1pro:βC1 , 35Spro:GFP-βC1 and 35Spro:HA-βC1 were used to generate transgenic plants respectively . PVX-cLUC , PVX-βC1 and PVX-βC1ΔC43 were constructed by introducing DNA fragments of cLUC , βC1 and βC1ΔC43 into a PVX vector [60] . pBinPLUS-TA and pBinPLUS-2mβ were kindly provided by Professor Xueping Zhou [61] . All constructs were confirmed by DNA sequencing . Primers used in this study were listed in S1 Table . Total DNA was extracted from apical developing leaves using the DNAsecure Plant Kit ( TIANGEN , China ) . DNA concentration of each sample was calculated through OD260 via Epoch Multi-Volume Spectrophotometer System ( Bio-Tek , USA ) and then diluted to around 60ng/ul for PCR amplication . A single copy of CLCuMuV genome was amplified by PCR and then was ligased into pMD19-T ( TaKaRa , Japan ) to generate a CLCuMuV-positive plasmid . A 10-fold serial dilution of the plasmid DNA from 2×108 to 200 copy was prepared and used as the standard . A CLCuMuV-specific primer set ( qCLCuMuV V1-F and qCLCuMuV V1-R ) was used to amplify a 198-bp amplicon . For SYBR Green-based real-time PCR performed in a 10 μL reaction mixture containing 5 μl Power SYBR Green PCR Master Mix ( 2× ) ( Life , USA ) , primer concentration was optimized by running the assay using the plasmid DNA dilution series with two different primer concentration ( 10 and 20 μM ) . 0 . 1 μL of each 20 μM primer and 0 . 3 μL 60 ng/μL templet were finally chosen to amplify viral DNA in samples for following assays . Because the standard curves generated were linear in the whole range tested with a coefficient of regression R2:0 . 99 and calculated slope around -3 . 5 for SYBR Green assay . The copy number of viral DNA can be calculated via Ct value of each sample and the standard curve . To obtain the ratio of viral DNA: plant genome DNA , Plant genome DNA can also be calculated via internal reference method . The genome DNA of healthy N . benthamiana was extracted and a 2-fold serial dilution of the genome DNA from 94 . 5ng to 1 . 48ng was prepared and used as the standard . An eIF4a-specific primer set ( qeIF4a-F and qeIF4a-R ) was used to amplify a 60-bp amplicon . Primer concentration was optimized by using the plant genome DNA dilution series with three different primer concentrations ( 10 , 15 and 20 μM ) . 0 . 1 μL of each 15 μM primer was finally chosen because the standard curves generated were linear in the whole range tested with a coefficient of regression R2:0 . 99 and calculated slope around -3 . 3 for SYBR Green assay . The plant genome DNA can be calculated via Ct value of each sample and the standard curve . The full-length CLCuMuB βC1 was PCR amplified and cloned into yeast vector pYL302 to generate the LexA DNA binding domain ( BD ) containing bait vectors BD-CLCuMuB βC1 . The full-length NbSKP1 . 1 , NbSKP1 . 2 , NbSKP1 . 3 , NbSKP1L1 and NbSKP1 . 1 deletion derivatives were PCR amplified and cloned into the B42 activation domain ( AD ) -containing vector pJG4-5 . The yeast two-hybrid prey library containing tomato cDNAs was used to screen CLCuMuB βC1-binding proteins . The yeast two-hybrid screen and interaction assays were performed as described [35] . N . benthamiana plants were grown in pots at 25°C in growth rooms under 16 h light/8 h dark cycle with 60% humidity . Light intensity is 4000 lx . Solt mixed with vermiculite at a 1:1 ratio was used as the substrate for plants to grow . the plants were watered with a nutrient solution . For CLCuMuB-based VIGS assays , CLCuMuV or βM2 and its derivatives were introduced into Agrobacterium strain GV2260 . Agrobacterium cultures containing CLCuMuV or βM2 derivative plasmids were grown overnight at 28°C until OD600 = 2 . 0 , then CLCuMuV with corresponding βM2 derivative vector were mixed at 1: 1 ratio , pelleted , resuspended in infiltration buffer ( 10 mM MgCl2 , 10 mM MES , and 200 μM acetosyringone , pH 5 . 6 ) to OD600 = 1 . 0 , kept at room temperature for 4 h and infiltrated into the lower leaf of 6-leaf stage plants using a 1-ml needleless syringe . For Agrobacterium tumefaciens-mediated transient expression studies , GV2260 strains containing the relevant expression vectors were cultured and prepared as described above , then were infiltrated into N . benthamiana leaves . The infiltrated leaves were detached at 48 to 60 hpi for the corresponding assays . For coexpression , equal amounts of A . tumefaciens cultures were mixed and used for infiltration . MeJA treatments: a 50 μM MeJA solution or mock solution ( ethanol ) were applied to 6-week-old N . benthamiana plants by spray every other day from 1 day before the inoculation to 14 dpi . Citrine YFP-based BiFC was performed as described [36] . The experimental group and corresponding control group should be inoculated in a same leaf to reduce the difference of expression condition . Live plant imaging was performed on a Zeiss LSM710 confocal microscope . Enhanced citrine YFP-derived fluorescence was acquired using 514-nm laser and emission 519- to 587-nm filters . 8-bit confocal images were acquired with an EC Plan-Neofluar 103/0 . 30 M27 objective for 103 magnification and a Plan-Apochromat 403/0 . 95 Korr M27 objective for 403 magnification . Images were analyzed with ZEN 2012 Light Edition . The experimental group and corresponding control group were inoculated in a same leaf . At 48 dpi , images of live plant samples from experimental and corresponding control groups were taken under the same parameters via a Zeiss LSM710 confocal microscope . Software ZEN 2012 was used to measure the fluorescence intensity mean value of an image . 4 independent images for each group were measured and values were analyzed via t-test . Three biological repeats were needed . Because βC1 protein was reported not stable in vivo and may be degraded through ubiquitin 26S proteasome system ( UPS ) [20] , so in this assay we added MG132 , an inhibitor against the 26S proteasome , to improve the accumulation of GFP-βC1 . For Co-IP assays , 50 μM MG132 ( Sigma , USA ) was inoculated into N . benthamiana leaves 12 h before being detached . total proteins from leaves were extracted with a ratio of 1:2 of native extraction buffer 1 [NB1; 50 mM TRIS-MES pH 8 . 0 , 0 . 5 M sucrose , 1 mM MgCl2 , 10 mM EDTA , 5 mM DTT , 50 μM MG132 , protease inhibitor cocktail CompleteMini tablets ( Roche , http://www . roche . com/ ) ] [62] . Protein extracts were incubated with the GFP-Trap beads ( ChromoTek , German ) for 2 hours at 4°C , The beads were washed three times with ice-cold NB1 at 4°C . IP samples were analyzed by SDS-PAGE , immunoblotted using anti-HA ( CST , USA ) and anti-GFP antibodies ( Abmart , China ) and detected using Pierce ECL western blotting substrate ( Thermo , USA ) . GST-CLCuMuB βC1 and HA-His-NbSKP1 . 1 fusion proteins were produced in BL21 ( DE3 ) codon plus RIL cells . HA-His-NbSKP1 . 1 was purified using Ni-NTA Agarose ( Qiagen , Netherlands ) column . GST-CLCuMuB βC1 was purified using Glutathione Sepharose 4B ( GE , USA ) and then used to pull down HA-His-NbSKP1 . 1 in vitro for 2 hours at 4°C . The beads were washed three times with ice-cold elution buffer ( 300 mM NaCl , 50 mM Tric-HCl , pH 8 . 0 , 0 . 1% Triton-X 100 ) at 4°C . The washed beads were boiled in SDS sample buffer , and proteins were separated by SDS-PAGE and detected by western blot using an anti-HA antibody ( CST , USA ) . His-CLCuMuB βC1 and HA-His-NbSKP1 . 1 fusion proteins were produced in BL21 ( DE3 ) codon plus RIL cells . E . coli cells harboring the corresponding clones were cultured in LB medium ( 5 mL ) containing kanamycin ( 50 μg/mL ) at 37°C , till the O . D . at 600 nm reached 0 . 6 . Then the cells were inoculated for large scale expression . The expression of corresponding genes were induced by the addition of isopropyl-β-D-thiogalactopyranoside ( IPTG , Sigma ) to the final concentration of 0 . 2 mM and cells were further allowed to grow for 20 hours at 16°C . The cells were spun down at 4 , 000 rpm , resuspended in the ice-cold lysis buffer ( 50 mM Tris-HCl , 300 mM NaCl , 1 mM PMSF , 50 mM DTT , pH 8 . 5 ) . Resuspended cells were sonicated till suspension became optically clear . HA-His-NbSKP1 . 1 was soluble and purified using Ni-NTA Agarose ( Qiagen , Netherlands ) column . His-CLCuMuB βC1 was in inclusion bodies and was dissolved by 8 M Urea ( 50 mM DTT , 8 M Urea ) with a ratio of 0 . 1g: 1ml . Insoluble substance were removed by centrifugation at 14 , 000 rpm , 30 min , 4°C . Supernatant was dripped slowly using a 1-ml syringe with needle into 200 mL ice-cold refolding buffer ( 50 mM Tris-HCl , 300 mM NaCl , 500 mM Arginine , 2 M Urea , 1 mM PMSF , pH 8 . 5 ) agitated by a magnetic stirring apparatus . Then this His-CLCuMuB βC1 solution was dialyzed against the dialysis buffer ( 50 mM Tris-HCl , 300 mM NaCl , pH 8 . 5 ) . The protein obtained by this method was enriched by Ni-NTA Agarose ( Qiagen , Netherlands ) column and eluted for further experiments . 1 mL GFP-CUL1 or GFP extracts were prepared and immunoprecipitated by 20 μL GFP-Trap beads ( ChromoTek , German ) for each sample as described in the Co-Immunoprecipitation ( Co-IP ) part . After two washes with wash buffer ( 50 mM Tris-HCl , 300 mM NaCl , 1 mM PMSF , 50 mM DTT , pH 8 . 5 ) , 1 mL 100 μg/mL E . coli-expressed His-HA-NbSKP1 . 1 was added and incubated at 4°C for 1 hour . After two washes with wash buffer , 80 μg , 40 μg , 20 μg His-βC1 or 80 μg BSA was added in 1 mL corresponding samples and incubated at 4°C for 1 hour . After three washes with wash buffer , samples were separated by SDS-PAGE , transferred to PVDF membrane , and detected with corresponding antibodies . The experiments were performed as described by Lozano-Duran [33] . Seeds of wild-type or HA-βC1 transgenic N . benthamiana used in this study were surface sterilized and sown on Murashige and Skoog ( MS ) agar plates with 30 g/L sucrose and 0 . 6% Agar . Seedlings were grown at 25°C under 4000 lx white light with a 16-h-light/8-h-dark photoperiod . MS plates were placed in a vertical orientation for 6 d , and seedlings were then transferred to MS plates containing no or 50 μM MeJA ( Sigma , USA ) . Root length was scanned every day until 5 days later . 14–15 days Nicotiana benthamiana plants were inoculated with CA+β or CA+βM1 . Leaves in three replicate plants for each treatment were sampled . The leaf materials from each plant were flash-frozen in liquid nitrogen , weighed and stored at -80°C until JA analysis . Sample preparation was performed as described by Glauser and Wolfender , [63] . Except methanol–water , 40:60 ( v/v ) was used to resolubilize the final residue and do subsequent UHPLC-Q-TOFMS Analysis . Drug Discovery Facility , Center of Biomedical Analysis , Tsinghua University provided the service for sample determination . GAI was cloned from cDNA of N . benthamiana and the experiments were performed as described by Lozano-Duran [33] . At 48 h past inoculation , the agroinfiltrated leaves were sprayed with a 100 μM GA3 solution or with mock solution ( ethanol ) . Fluorescence was visualized 1 to 2 hours later using a Zeiss LSM710 confocal microscope . Leaf samples were grind by liquid nitrogen , Then total proteins were extracted with a ratio of 1:4 of extraction buffer ( 50 mM Tris-HCl , 100 mM NaCl , 25 mM imidazole , 10% glycerol , 0 . 1% Tween-20 , 20 mM β-mercaptoethanol ) [45] . Samples were separated by SDS-PAGE , transferred to PVDF membrane , and detected with the anti-GFP ( ChromoTek , German ) . Myc-COI1 was expressed in N . benthamiana and purified as described [45] . 60 μL of purified protein was added to 540 μL of total crude protein extracts ( 1 mg/mL ) from N . benthamiana which was transiently expressed HA-βC1 or HA-nLUC , and then were incubated at 25°C for indicated time periods , separated by SDS-PAGE , transferred to PVDF membrane , and detected with the anti-Myc ( Abmart , China ) . Total DNA was extracted from apical developing leaves using the DNAsecure Plant Kit ( TIANGEN , China ) . Total RNA was extracted from apical developing leaves using the Trizol reagent ( TIANGEN , China ) and treated with RNase-free DNase I ( Sigma-Aldrich ) . First strand cDNA was synthesized using 2–5 μg of total RNA with oligo-d ( T ) primer and M-MLV reverse transcriptase ( TIANGEN , China ) . Real time RT-PCR was performed using Power SYBR Green PCR master mix ( Life , USA ) . EIF4a and Actin were used as internal control for N . benthamiana for normalization . Primers were designed with Primer3web ( http://primer3 . ut . ee/ ) and listed in Supplemental Table S1 . The values were calculated using the comparative normalized Ct method and all the experiments were repeated at least two times . Data were analyzed and plotted with Origin 8 . 1 . Sequence data from this article can be found in the GenBank data libraries under accession numbers: CLCuMuV ( GQ924756 ) ; CLCuMuB ( GQ906588 ) ; SlSKP1 ( XM_004250675 ) ; NbSKP1 . 1 ( KP017273 ) ; NbSKP1 . 2 ( KP017274 ) ; NbSKP1 . 3 ( KP017275 ) ; NbSKP1L1 ( KP017276 ) ; NbCUL1 ( KP017277 ) ; UBC3 ( KR296788 ) ; eIF4a ( KX247369 ) ; Actin ( JQ256516 ) ; PID ( KR082145 ) ; COI1 ( AF036340 ) ; GAI ( KR082148 ) ; GFP ( U87973 ) ; Defensin-like protein 1 ( KX139060 ) ; Defensin-like protein 2 ( KX139061 ) ; Pathogen like protein ( KX139062 ) ; Gibberellin- regulated protein 14 ( KX139063 ) ; Gibberellin-regulated protein 6 ( KX139064 ) ; SAUR14 ( KX139065 ) .
|
Viruses pose a serious threat to field crops worldwide; therefore , understanding the mechanisms of viral disease can help crop improvements . Here , we investigate how Cotton leaf curl Multan virus ( CLCuMuV ) interacts with plant to cause viral disease . We found that CLCuMuV uses its sole satellite-encoded protein βC1 to regulate the plant ubiquitination pathway for effective infection . By interrupting the interaction of NbSKP1 with NbCUL1 through its interaction of SKP1 , βC1 interferes with the plant ubiquitination pathway and impairs plant hormone signallings to enhance viral accumulation and symptoms . These new insight into the mechanisms of viral disease may help crop improvements in the future .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"biotechnology",
"plant",
"anatomy",
"membrane",
"staining",
"rna",
"extraction",
"organisms",
"plant",
"science",
"genetically",
"modified",
"plants",
"molecular",
"biology",
"techniques",
"plants",
"extraction",
"techniques",
"genetic",
"engineering",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"genetically",
"modified",
"organisms",
"protein",
"extraction",
"artificial",
"gene",
"amplification",
"and",
"extension",
"proteins",
"ubiquitination",
"leaves",
"molecular",
"biology",
"agriculture",
"biochemistry",
"plant",
"biochemistry",
"post-translational",
"modification",
"biology",
"and",
"life",
"sciences",
"agricultural",
"biotechnology",
"plant",
"biotechnology",
"polymerase",
"chain",
"reaction"
] |
2016
|
CLCuMuB βC1 Subverts Ubiquitination by Interacting with NbSKP1s to Enhance Geminivirus Infection in Nicotiana benthamiana
|
Mastermind-like 1 ( MAML1 ) is a transcriptional co-activator in the Notch signaling pathway . Recently , however , several reports revealed novel and unique roles for MAML1 that are independent of the Notch signaling pathway . We found that MAML1 enhances the transcriptional activity of runt-related transcription factor 2 ( Runx2 ) , a transcription factor essential for osteoblastic differentiation and chondrocyte proliferation and maturation . MAML1 significantly enhanced the Runx2-mediated transcription of the p6OSE2-Luc reporter , in which luciferase expression was controlled by six copies of the osteoblast specific element 2 ( OSE2 ) from the Runx2-regulated osteocalcin gene promoter . Interestingly , a deletion mutant of MAML1 lacking the N-terminal Notch-binding domain also enhanced Runx2-mediated transcription . Moreover , inhibition of Notch signaling did not affect the action of MAML1 on Runx2 , suggesting that the activation of Runx2 by MAML1 may be caused in a Notch-independent manner . Overexpression of MAML1 transiently enhanced the Runx2-mediated expression of alkaline phosphatase , an early marker of osteoblast differentiation , in the murine pluripotent mesenchymal cell line C3H10T1/2 . MAML1−/− embryos at embryonic day 16 . 5 ( E16 . 5 ) had shorter bone lengths than wild-type embryos . The area of primary spongiosa of the femoral diaphysis was narrowed . At E14 . 5 , extended zone of collagen type II alpha 1 ( Col2a1 ) and Sox9 expression , markers of chondrocyte differentiation , and decreased zone of collagen type X alpha 1 ( Col10a1 ) expression , a marker of hypertrophic chondrocyte , were observed . These observations suggest that chondrocyte maturation was impaired in MAML1−/− mice . MAML1 enhances the transcriptional activity of Runx2 and plays a role in bone development .
Runt-related transcription factor 2 ( Runx2 ) is a transcription factor belonging to the Runx gene family , which is homologous to Drosophila runt , a pair-rule gene involved in somitogenesis [1] . Runx2 is an essential factor for bone and hypertrophic cartilage formation that is expressed very early in bone development and continues to be present through the later phase of development [2] . Runx2 promotes the differentiation of pluripotent mesenchymal progenitor cells into the osteogenic lineage , but its role in terminal differentiation to mature osteoblasts and the production of bone matrix remains unclear . To date , it has been reported that several transcription factors and cofactors , such as TAZ [3] , Grg5 [4] , Rb [5] , and HDAC4 [6] , interact with Runx2 and positively or negatively regulate its function . However , in many cases , the physiological significance of the interaction is unclear . To further elucidate the function of Runx2 , we performed luciferase assay-based screening of additional factors regulating the transcriptional activity of Runx2 using a full-length cDNA library containing approximately 10 , 000 clones . The screening system identified the mastermind-like ( MAML ) family of proteins showed especially strong potential for regulating Runx2 transcriptional activity . Overexpression of MAML1 enhanced the Runx2-mediated expression of alkaline phosphatase , an early marker of osteoblast differentiation , in C3H10T1/2 cells . Furthermore , MAML1−/− embryos at E14 . 5 and 16 . 5 had shorter bone lengths than wild type embryos . The area of primary spongiosa of the femoral diaphysis was narrowed , indicated that chondrocyte maturation was impaired . These data suggest that MAML1 enhanced the transcriptional activity of Runx2 and plays a role in bone development .
We used a full-length cDNA library containing approximately 10 , 000 clones ( FLJ clones , established by New Energy and Industrial Technology Development Organization [NEDO] , Japan ) and p6OSE2-Luc reporter assay system ( Figure 1A ) . We identified a few novel factors that enhance Runx2 transcriptional activity . Among them , AK123604 ( Homo sapiens cDNA FLJ41610 ) , which is highly similar to Mastermind-like protein 3 ( MAML3 ) , showed especially strong activity . MAML is a human homolog of Drosophila mastermind , a protein that plays a role in the Notch signaling . MAML family members consist of MAML1 , MAML2 and MAML3 . We found that MAML1 and MAML2 also enhanced Runx2 transcriptional activity as well ( Figure 1B ) . Because the establishment of knockout mice of MAML1 preceded MAML2 and MAML3 , we primarily analyzed MAML1 . MAML1 consists of 1016 amino acids and contains a conserved basic region and two acidic regions . To investigate which region is essential for regulating Runx2 , we assessed each deletion mutant by p6OSE2-Luc reporter assay ( Figure 2 ) . The N-terminal basic region , which is essential for the interaction with Notch , and the C-terminal acidic region of MAML1 were dispensable for Runx2 transcriptional activity . On the other hand , the center region ( residues 343–711 ) , whose function is not well known , was essential . Because MAML1 is a coactivator of Notch signaling , we investigated whether or not the action of MAML1 on Runx2 was dependent on Notch . Notch1ΔE is cleaved by γ-secretase to produce the Notch intracellular domain ( NICD ) , which translocates into the nucleus and transactivates the target gene . A Γ-secretase inhibitor DAPT inhibited the Notch1ΔE-mediated activation of pTP1-Luc , in which luciferase expression was controlled by Notch signaling ( Figure 2C ) . On the other hand , DAPT did not affect the action of MAML1 on Runx2 in the presence of Notch1ΔE ( Figure 2D ) . This suggests that MAML1 possibly enhance the transcriptional activity of Runx2 in a Notch-independent manner . 293T cell used in the luciferase assay is derived from human embryonic kidney and does not express Runx2 . Therefore , we next investigated whether MAML1 controls osteoblastic differentiation through Runx2 in the murine pluripotent mesenchymal cell line C3H10T1/2 ( Figure 3 ) . Overexpression of Runx2 promoted the expression of the ALP gene , an early osteoblast marker . Co-overexpression of MAML1 rapidly augmented the Runx2-mediated expression of ALP , whereas MAML1 alone did not induce ALP expression . However , this effect was observed only in early phase of osteoblast differentiation and later phase markers such as bone sialoprotein and osteocalcin were not changed ( data not shown ) . We analyzed MAML1 knockout ( MAML1−/− ) mice [7] . Normal Mendelian ratios are observed up to E18 . 5 , but MAML1−/− mice with C57BL/6 background die during the perinatal period . In the original paper , however , MAML1−/− mice die within 14 days after birth [7] . The difference of lethality in the mice is thought to be due to the difference of the background . At E16 . 5 , MAML1−/− mice were smaller than wild type mice ( Figure 4A ) . Whole mounted embryos at E16 . 5 stained with Alcian Blue and Alizarin Red showed that the mineralized region in the long bones of MAML1−/− mice was relatively short compared with wild type mice ( Figure 4A ) . Histological analysis revealed that the area of primary spongiosa of the femoral diaphysis was reduced in MAML1−/− mice compared to wild type mice ( Figure 4B , 4C ) . At E14 . 5 , extended zone of Col2a1 and Sox9 expression , markers of chondrocyte differentiation , and decreased zone of Col10a1 expression , a marker of hypertrophic chondrocyte , were observed ( Figure 4D ) . These observations indicated the impairment of chondrocyte maturation in MAML1−/− mice .
We utilized approximately 10 , 000 arrayed and addressable cDNA clones , which allowed systematic , efficient , and unbiased screening of cDNAs encoding factors that could activate Runx2-mediated expression of the p6OSE2-Luc reporter construct ( Figure 1 ) . This revealed that MAML was a potential activator of Runx2-mediated luciferase expression . MAML is a coactivator of Notch signaling . Upon ligand stimulation from neighboring cells , Notch is cleaved by γ-secretase and its intracellular domain ( NICD ) translocates into the nucleus [8] . NICD interacts with CSL through a RAM domain at the N-terminus that has high affinity for the β-trefoil domain of CSL . Then , the ankyrin repeats domain of NICD docks with the Rel-homology domain of CSL and creates a high-affinity binding site for MAML . MAML associates with the CSL-NICD complex through the N-terminal basic region , recruits p300 , RNA polymerase II and other unknown factors , and activates the transcription of target genes such as Hes1 [9] , [10] . On the other hand , Notch-independent action of MAML1 on p53 [11] , beta-catenin [12] , MEF2C [13] and NF-kappaB [14] has been previously reported . Recently , two groups have published studies using genetically modified mice [15] , [16] . Hilton and colleagues showed that Notch signaling inhibits osteoblast differentiation through Hes or Hey proteins , which diminish Runx2 transcriptional activity via physical interaction , and acts to maintain a pool of mesenchymal progenitors . Engin and colleagues showed that pathological gain of Notch function in established osteoblastic lineages activates expansion of the immature osteoblastic pool by increasing transcription of the genes encoding osterix , cyclin D and cyclin E and by repressing the function of Runx2 by direct interaction and inhibition of its binding . These findings suggest that Notch signaling negatively regulate the function of Runx2 . We indicated that the N-terminal Notch-binding region of MAML1 is dispensable for the action of MAML1 on Runx2 and Notch signaling inhibitor does not affected the action of MAML1 on Runx2 . Furthermore , knockdown of p300 , a coactivator [9] , [10] , did not affected the activation of Runx2 transcriptional activity by MAML1 ( data not shown ) . These data suppose that the action of MAML1 on Runx2 is Notch-independent . To elucidate how MAML regulates Runx2-mediated transcription , we investigated the physical interaction of MAML1 with Runx2 , but we could not demonstrate the interaction between Runx2 and MAML1 ( data not shown ) , suggesting that this interaction is very weak and possibly indirect . We showed the impairment of chondrocyte maturation in MAML1−/− mice . Because Runx2 facilitates chondrocyte maturation , the phenotype of MAML1−/− mice may be caused by the dysfunction of Runx2 . On the other hand , the expression of Sox9 , a transcription activator of collagen type II , was upregulated by Notch activation and this activation of Notch signaling thereby promoted differentiation of proliferative and prehypertrophic chondrocytes [17] . Therefore , from this current findings , it is not clear yet whether or not the phenotype of MAML1−/− mice is due to the dysfunction of Runx2 or Notch signaling . Other possibility to explain MAML1−/− mice bone phenotype is that other cell signaling cascades and molecules could be involved into MAML dependent gene regulation and thus bone development . For example , MEF2C , a transcription factor that regulates muscle and cardiovascular development , was reported to control bone development by activating the gene program for chondrocyte hypertrophy [18] . Taken together , our analysis revealed novel function of MAML1 , Notch independent promotion of Runx2 activity and its role in bone development . Further elucidation of the precise molecular mechanisms responsible for the initiation and termination of this functional association during bone development may provide us with a new basis for understanding the molecular network in osteoblasts and potential therapeutic targets for bone diseases .
All animal experiments were performed according to protocols approved by the Institutional Animal Care and Use Committee at National Institute for Child Health and Development ( Protocol 2004-003 ) . The p6OSE2-Luc and p6OSE2-mut-Luc reporter construct were previously reported [19] . The pEF-BOS hMam-1 ( MAML1 ) plasmid , its truncated forms [20] , hMam-2 ( MAML3 ) , hMam-3 ( MAML2 ) [21] and pCS2+Notch1ΔE [22] were previously reported . The pCG mRunx2 plasmid by Dr . Nakashima ( Tokyo Medical and Dental University , Tokyo ) , and the p3xFLAG mRunx2 plasmid by Dr . Hikata ( Keio University , Tokyo ) . The pTP1-Luc ( pGa981-6 ) construct was provided by Dr . Ursula Strobl ( Institute of Clinical Molecular Biology and Tumor Genetics , Germany ) . For the primary screening , we diluted approximately 10 , 000 FLJ clones ( Full-length human cDNA sequencing project , NEDO ) to 10 ng/µL in 10 mM Tris-HCl ( pH 8 . 5 ) , and dispensed 5 µL to each well in 384-well plates . We then added 10 ng of p6OSE2-Luc , 2 ng of pCG , 0 . 1 µL of Fugene6 ( Roche Diagnostics ) , and 5 µL of Opti-MEM I Reduced-Serum Medium ( Invitrogen ) to each well . 293T cells were diluted to 1 . 25×105 cells/mL with Dulbecco's modified Eagle medium ( DMEM ) containing 10% heat-inactivated fetal bovine serum ( FBS ) , 50 units/mL penicillin , and 50 µg/mL streptomycin , and seeded at 40 µL ( 5 , 000 cells ) per well . After 48 hours of culture , we removed the supernatant and added 40 µL of Steady Glo Luciferase Assay Reagent ( Promega ) diluted 2-fold with phosphate buffered saline ( PBS ) to each well . After 10 minutes at room temperature , luminescence was measured using a plate reader ( ARVO , Perkin Elmer ) . After the second screening , the assay was performed in 96-well plates . We added γ-secretase inhibitor IX , N-[N- ( 3 , 5-difluorophenylacetyl-L-alanyl ) ]-S-phenylglycine t-butylester ( DAPT; Calbiochem ) , to the medium 2 hours before transfection . We purchased the C3H10T1/2 murine pluripotent mesenchymal cell line from ATCC and maintained it in DMEM containing 10% heat-inactivated FBS , 50 units/mL penicillin , and 50 µg/mL streptomycin . For differentiation assays , we seeded cells in a multi-well plate at a density of 2 , 000 cells/cm2 and cultured them for 3 days . The medium was then changed to the osteoblastic medium ( MEM-alpha containing heat-inactivated FBS , 50 units/mL penicillin , 50 µg/mL streptomycin , 50 µM ascorbic acid 2-phosphate , 10 mM β-glycerophosphate , and 0 . 1 µM dexamethasone ) , transfected with p3xFLAG-Runx2 and/or pEF BOS-hMam1 by FugeneHD ( Roche diagnostics ) , and cultured . We isolated total RNA from the cultured cells using the RNeasy mini kit ( QIAGEN ) and reverse transcribed 2 µg of total RNA using Ready-To-Go You-Prime First-Strand Beads ( GE Healthcare ) and oligo-dT primer . The products were diluted 10-fold with distilled water and used as a template for real-time PCR . Real-time PCR was performed using a TaqMan Gene Expression Assay , TaqMan Universal PCR Mix and the 7900HT Fast Real-Time PCR System ( Applied Biosystems ) . We backcrossed MAML1 null mice [7] at least 10 times onto a C57BL/6 background . We fixed mouse embryos on embryonic day 16 . 5 ( E16 . 5 ) in ethanol overnight and then stained them overnight with Alcian blue solution ( 0 . 15 mg/mL Alcian blue 8GX in 20% acetic acid and 80% ethanol ) . The embryos were washed briefly with ethanol twice , treated with 2% potassium hydroxide overnight , and then stained overnight with Alizarin red solution ( 0 . 075 mg/mL Alizarin red S in 1% potassium hydroxide ) . Tissues were fixed in 4% paraformaldehyde-PBS overnight at 4°C , processed , embedded in paraffin , and sectioned . Slides were deparaffinized , washed with water , treated with 3% acetic acid , and then with 1% Alcian blue 8GX for 60 minutes . After staining , we washed the slides briefly with 3% acetic acid , then with water for 5 minutes , counterstained with Kernechtrot Stain Solution ( Muto Pure Chemicals , Tokyo ) for 5 minutes , washed with water for 3 minutes , and dehydrated the slides . Tissues were fixed in 4% paraformaldehyde-PBS overnight at 4°C , processed , embedded in paraffin , and sectioned . Slides were deparaffinized , treated with proteinase K ( 8 µg/mL ) for 10 minutes at RT , and then with 0 . 2% glycine in PBS for 10 minutes at RT . Slides were refixed in 4% paraformaldehyde-PBS for 10 minutes at RT , washed with PBS for 5 minutes 3 times , acetylated with 0 . 1 M triethanolamin-HCl ( pH 8 . 0 ) for 10 minutes , washed with PBS for 30 minutes , and then prehybridized with prehybridization buffer ( 50% deionized formamide and 5× saline-sodium citrate ( SSC ) ) for 60 minutes at 65°C . We hybridized the slides with DIG-labeled antisense riboprobes in hybridization buffer ( 50% deionized formamide , 5× SSC , 0 . 25 mg/mL yeast tRNA , 10% dextran sulfate , and 5× Denhardt's solution ) in a humidified chamber at 65°C overnight . After hybridization , the slides were washed with 5× SSC ( 1× SSC: 0 . 15 M NaCl , 0 . 015 M sodium citrate ) at 65°C for 20 minutes , 0 . 2× SSC at 65°C for 3 hours , and NT buffer ( 0 . 1 M Tris-HCl [pH 7 . 5] , 0 . 15 M NaCl ) for 5 minutes at RT . We incubated the slides at 4°C overnight with alkaline phosphatase ( ALP ) -coupled anti-DIG antibody in NT buffer containing 0 . 1% sheep serum . The slides were washed with NT buffer for 15 minutes 3 times and equilibrated in NTM ( 0 . 1 M NaCl , 0 . 1 M Tris-HCl [pH 9 . 5] , and 0 . 05 M MgCl2 ) for 5 minutes at RT . The slides were then treated with BM Purple AP Substrate ( Roche ) for 3 hours at RT in a humid chamber protected from light . The two-tailed independent Student's t-test was used to calculate all P values .
|
To identify new molecules involved in bone and cartilage development and/or homeostasis , we utilized approximately 10 , 000 arrayed and addressable cDNA clones , which allowed systematic , efficient , and unbiased screening of cDNAs encoding factors that could activate critical bone differentiation activity via activation of Runx2 , master regulator of bone development . We analyzed MAML1−/− mice to investigate the role of MAML1 in bone development . MAML1−/− embryos at embryonic day 14 . 5 and 16 . 5 had shorter bone lengths than wild-type embryos . The area of primary spongiosa of the femoral diaphysis was narrowed , indicated that chondrocyte maturation was impaired . This revealed that MAML1 plays an important role in proper bone development and may provide us with a new basis for identifying potential therapeutic targets for bone diseases .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"biology"
] |
2013
|
MAML1 Enhances the Transcriptional Activity of Runx2 and Plays a Role in Bone Development
|
Human cytomegalovirus ( HCMV ) can infect many different cell types in vivo . Two gH/gL complexes are used for entry into cells . gH/gL/pUL ( 128 , 130 , 131A ) shows no selectivity for its host cell , whereas formation of a gH/gL/gO complex only restricts the tropism mainly to fibroblasts . Here , we describe that depending on the cell type in which virus replication takes place , virus carrying the gH/gL/pUL ( 128 , 130 , 131A ) complex is either released or retained cell-associated . We observed that virus spread in fibroblast cultures was predominantly supernatant-driven , whereas spread in endothelial cell ( EC ) cultures was predominantly focal . This was due to properties of virus released from fibroblasts and EC . Fibroblasts released virus which could infect both fibroblasts and EC . In contrast , EC released virus which readily infected fibroblasts , but was barely able to infect EC . The EC infection capacities of virus released from fibroblasts or EC correlated with respectively high or low amounts of gH/gL/pUL ( 128 , 130 , 131A ) in virus particles . Moreover , we found that focal spread in EC cultures could be attributed to EC-tropic virus tightly associated with EC and not released into the supernatant . Preincubation of fibroblast-derived virus progeny with EC or beads coated with pUL131A-specific antibodies depleted the fraction that could infect EC , and left a fraction that could predominantly infect fibroblasts . These data strongly suggest that HCMV progeny is composed of distinct virus populations . EC specifically retain the EC-tropic population , whereas fibroblasts release EC-tropic and non EC-tropic virus . Our findings offer completely new views on how HCMV spread may be controlled by its host cells .
Human cytomegalovirus ( HCMV ) is ubiquitously distributed in the human population . In immunocompetent adults infections are mainly asymptomatic , but in immunocompromised patients like transplant recipients or AIDS patients life threatening infections occur at a high rate . HCMV is also the leading cause of birth defects among congenitally transmitted viral infections . HCMV replicates in vivo and in vitro in many different host cells including epithelial cells , connective tissue cells , hepatocytes , various leukocyte populations and vascular endothelial cells ( reviewed in [1] ) . The broad host cell range implicates that either an ubiquitous cellular receptor , recognized by one protein or protein complex in the viral envelope , mediates entry , or that HCMV uses elaborate combinations of different viral envelope proteins to employ different cellular receptors . More than 10 glycoproteins have been identified in HCMV particles [2] , including the essential glycoproteins gB , gH , gL , gM and gN , which all play a role in the virus entry process [3]–[7] . Although a number of cellular surface proteins have been identified to bind these envelope proteins and play a role in virus particle attachment or promoting intracellular signaling after binding [8]–[13] , none of them is currently considered to be a functional entry receptor . The best candidates for binding to entry receptors are the HCMV gH/gL complexes . The gH/gL complex has been shown to promote fusion of cellular membranes [7] and can either form a gH/gL/gO [14] , [15] or a gH/gL/pUL ( 128 , 130 , 131A ) complex [16]–[18] . HCMV isolates from patients are consistently able to form both gH/gL complexes [19] , [20] . In contrast , many HCMV laboratory strains express only the gH/gL/gO complex , which restricts virus entry to few cell types like fibroblasts and neuronal cells [21] , [22] . Leukocytes , dendritic , epithelial and endothelial cells ( EC ) can only be infected by virus expressing the gH/gL/pUL ( 128 , 130 , 131A ) complex [16] , [17] , [22] , [23] , which can also promote infection of fibroblasts [24] . Virus strains expressing only gH/gL/gO enter fibroblasts through fusion at the plasma membrane [25] . When fibroblast infection is promoted by gH/gL/pUL ( 128 , 130 , 131A ) only , then entry is through pH-sensitive endocytosis [26] . It is currently not clear whether gH/gL/gO complexes exert their function by directly initiating entry [27] . gO has been shown to be incorporated in the virus envelope of the HCMV strain AD169 , a laboratory strain which does not express the gH/gL/pUL ( 128 , 130 , 131A ) complex [2] , [27] , but not in the envelope of the clinical isolate TR [27] . Deletion of gO in a virus background , which still allows formation of the gH/gL/pUL ( 128 , 130 , 131A ) complex , strongly impairs release of infectious virus particles from infected cells . Virus spread becomes focal and dependent on the gH/gL/pUL ( 128 , 130 , 131A ) complex [24] , [26] , [28] . In contrast to the gH/gL/gO complex , the gH/gL/pUL ( 128 , 130 , 131A ) complex has been found to be consistently incorporated into virions [16]–[18] , [29] . The exact roles of the individual proteins of the gH/gL/pUL ( 128 , 130 , 131A ) complex are not known , but pUL128 , pUL130 and pUL131A are all needed to form a functional complex with gH/gL and to have this complex incorporated into virions [16]–[18] . Although the data are controversal , the gH/gL/pUL ( 128 , 130 , 131A ) complex very likely promotes entry into endothelial and epithelial cells through an endocytotic pathway [30]–[33] . There is also good evidence for epithelial cells that binding and uptake of virus is promoted through a cell type-specific receptor for the gH/gL/pUL ( 128 , 130 , 131A ) complex [34] . Viruses lacking both , gO and pUL ( 128 , 130 , 131A ) , are not viable , indicating that at least one of the two gH/gL complexes is needed for infection [24] . It is not known whether both gH/gL complexes are incorporated in one particle or whether they are incorporated into distinct particles , and how the usage of the complexes for entry is regulated . The formation of distinct gH/gL complexes is not restricted to HCMV and has also been described for EBV and HHV-6 [35] , [36] . For EBV , a gH/gL/gp42 and a gp42-negative gH/gL complex have been described . The latter binds to integrins αvß6 and αvß8 and promotes entry into epithelial cells by fusion at the plasma membrane [37]–[39] . The gH/gL/gp42 complex binds to HLA-DR ß and promotes entry into B-cells by an endocytotic route [38]–[40] . During virus production in B-cells , gp42 is intracellularly targeted to HLA-DR ß , where it is vulnerable for degradation . Consequently , B-cells release virus particles , which are low in gH/gL/gp42 . This virus is directed towards epithelial cells . Epithelial cells on the other hand do not express HLA-DR ß and produce virus which is high in gH/gL/gp42 and is directed to B-cells [35] . Thus , the EBV host cell tropism is switched by alternate replication in B- or epithelial cells . For HHV-6 a gH/gL/gO and a gH/gL/Q1/Q2 complex have been identified [36] , [41] , [42] . The latter has a high affinity for the HHV-6 cellular receptor CD46 [41] , whereas the gH/gL/gO complex does not bind CD46 [36] . Here , we show that , similar to EBV , also HCMV progenies derived from different cell types differ in their cell tropism . Fibroblast-derived virus progeny could readily infect fibroblasts and EC , whereas EC-derived virus progeny was barely able to infect EC , and this difference in tropism was reflected by a respectively high or low content of the gH/gL/pUL ( 128 , 130 , 131A ) complex in virus particles . EC-tropism could be depleted from fibroblast-derived virus progeny , indicating that this progeny is composed of distinct populations of virus particles with different EC infection capacities . Spread patterns in culture and cell disruption experiments indicated that fibroblasts readily released EC-tropic and non EC-tropic virus particles , whereas EC selectively retained the EC-tropic population .
When fibroblasts and EC are infected with HCMV in vitro , virus homogeneously spreads in fibroblast cultures whereas spread in endothelial cell cultures stays focal [17] , [43] . Here , we infected fibroblasts and EC with the HCMV strains VR1814 and TB40/E , two clinical isolates passaged on endothelial cells , and vTB40-BAC4 , a virus derived from TB40/E and cloned as a bacterial artificial chromosome ( BAC ) . Infections were performed at a low multiplicity of infection ( m . o . i . ) , and 2 or 8 days after infection cells were stained for HCMV immediate early 1 ( ie1 ) protein expression . Numbers of initially infected HFF or EC were comparable ( Fig . 1A , VR1814 , day 2 and data not shown ) . When fibroblasts were infected , HCMV homogeneously spread throughout the culture indicating release of virus from infected cells and infection via free supernatant virus ( Fig . 1A , day 8 ) . In contrast , EC infection remained focal indicating virus transmission which delivers virus particles from cell-to-cell , without releasing it . This spread pattern in EC cultures was comparable for all HCMV strains tested and independent of whether a microvascular cell line ( TIME ) or primary macrovascular endothelial cells ( HUVEC ) were infected ( Fig . 1A , day 8 , lower panels ) . Focal spread in EC cultures could be completely inhibited by neutralizing anti-HCMV antibodies in human serum or anti-pUL131A antibodies ( Fig . S1 ) , indicating that virus spread in EC cultures was not due to direct cell-to-cell spread . Spread in fibroblast cultures was restricted from supernatant-driven spread to focal spread by human antiserum and not inhibited at all by anti-pUL131A antibodies ( Fig . S1 ) . To test whether the focal spread could be attributed to differences in release of infectious virus , we performed growth curves of vTB40-BAC4 on HFF , TIME cells and HUVEC , and measured virus release into the supernatants by titration on fibroblasts . HFF and EC equally released high amounts of virus into their supernatants ( Fig . 1B ) . As spread of infection in EC cultures was focal although EC released virus in abundance , focal spread might be due to the inability of EC supernatant virus to infect EC . Indeed , although HFF and EC supernatants comparably infected fibroblasts , the capacities of EC-derived supernatants to infect EC were very low ( Fig . 1C ) . In the experiment shown , spread in HUVEC cultures appeared more cell-associated than in TIME cell cultures , where also single cells in between foci were ie1-positive ( Fig . 1A ) . This correlated with the lower HUVEC infection capacity of EC-derived supernatants when compared to the TIME cell infection capacity ( Fig . 1C ) . Infection capacities on different cell types are often compared by methods , which depend on counting infected cells which are either stained for viral antigen or GFP- expression . These methods reach their technical limits , when the infection capacities strongly differ on the cell types to be tested . It is very difficult to obtain reliable cell counts on the less permissive cell type , without at the same time saturating infection on the more permissive cells . Saturation yet , would lead to an overestimation of the infection capacity on the less permissive cells , when related to the more permissive cells . To circumvent these problems and to simplify the analysis , we used a luciferase reporter virus to monitor infection . An SV40 promoter-driven luciferase expression cassette was inserted into BAC4-FRT5-9 , a TB40-BAC4-derived BACmid lacking the genes UL5 to UL9 and carrying an FRT site at the position of the deleted locus ( Fig . S2A ) . Virus was reconstituted from BAC4-FRT5-9 ( vBAC4-FRT5-9 ) and BAC4-luc ( vBAC4-luc ) . Virus growth of these mutants in HFF and EC was comparable to growth of the parental vTB40-BAC4 ( Fig . S2B ) . Both , vBAC4-FRT5-9 and vBAC4-luc , also showed comparable spread patterns in HFF and EC cultures ( Fig . S2C ) . We used vBAC4-luc to evaluate EC and fibroblast infection capacities of virus preparations on HFF and TIME cells . The luciferase signals obtained from HFF and TIME cell infections were related to each other and expressed as TIME/HFF infection ratios , and thus , represent relative EC infection capacities . After infection , phosphono acetic acid ( PAA ) was added to block the viral DNA replication and the further amplification of the luciferase signal . Thus , the luciferase activity evaluates infection of cells in a fashion analogous to staining cells for HCMV ie1 protein expression . Indeed , when infection with one and the same virus preparation was evaluated either by counting ie1-positive cells or by measuring the luciferase activity in cell lysates , both methods always gave comparable results ( Fig . 2A and data not shown ) . The assay proved to be linear over a wide range of m . o . i and highly sensitive ( Fig . 2B ) . It is a standard observation in the field that one and the same HCMV preparation yields variable results , when repeatedly titrated on different target cell batches . When we tested HFF- and TIME cell-derived supernatants in independent luciferase assays , the results strongly depended on the quality of the cells used and varied with e . g . passage number and time after passage ( Fig . 2C ) . Virus preparations derived from infected HFF and TIME cells ( virus source ) were tested twice , using different batches of HFF and TIME cells ( assay 1 and 2 ) . In assay one , the infection capacities of the supernatants on HFF and TIME cell differed much more than in assay two ( Fig . 2C , upper panel ) , and consequently , the TIME/HFF infection ratios were in the range of 8 and 1 . 5% in assay one and in the range of 40 and 6% in assay two . ( Fig . 2C , lower panel ) . Yet , when the TIME/HFF infection ratios of the HFF supernatants were divided by the TIME/HFF infection ratios of the TIME cell supernatants , the quotients were comparable in both assays ( assay one: 6 . 9 , assay two: 7 . 5 ) . Therefore , the properties of virus preparations to be compared to each other were always tested in parallel . EC infection strictly depends on the gH/gL/pUL ( 128 , 130 , 131A ) complex [16] , [17] . The mutant vBAC4-luc/UL131Astop does not express pUL131A . The gH/gL/pUL ( 128 , 130 , 131A ) complex is not formed , and the mutant cannot infect EC . The mutant vBAC4-luc/ΔgO does not express gO and promotes entry into EC and also HFF via the gH/gL/pUL ( 128 , 130 , 131A ) complex [24] , [26]–[28] . We compared both mutants and the parental vBAC4-luc in the luciferase assay . Confirming the data from Figure 1C , supernatant from a vBAC4-luc infection of HFF showed a lower capacity to infect EC , when compared to the capacity to infect HFF ( Fig . 2D ) . vBAC4-luc/UL131Astop infected HFF , whereas the luciferase signals obtained from infected HUVEC and TIME cells remained below the detection limit . vBAC4-luc/ΔgO equally well infected HFF , TIME cells and HUVEC and thus showed an infection pattern clearly different from the parental vBAC4-luc . Taken together , the luciferase assay proved to be highly sensitive , to allow quantitative measurements over a wide range of m . o . i . , and to reflect what is seen , when infection is detected by staining cells for ie1 protein expression . With the luciferase assay described above , we could compare the properties of virus progenies from HFF and EC . Supernatants of infected HFF and EC were harvested 6 days after infection , titrated on HFF , and the viral DNA content determined by real-time PCR . The ratios of infectious virus to viral DNA copy numbers were comparable for HFF- and EC-derived supernatants ( data not shown ) . These supernatants were then used to infect HFF and TIME cells . Forty-eight hours later , infection was monitored by the luciferase assay . Although virus derived from all three cell types showed a comparable infection of HFF ( data not shown ) , EC-derived virus was significantly less capable in infecting EC than fibroblast-derived virus ( Fig . 3 ) . On average , the TIME/HFF infection ratios were about fourfold lower for virus released from EC than for virus released from HFF . Incorporation of gH/gL/pUL ( 128 , 130 , 131A ) glycoprotein complexes into virions [16] , [17] is a prerequisite to infect endothelial cells . As virus released from EC was less capable in infecting EC than virus released from fibroblasts , we asked whether this difference can be associated with the abundance of gH/gL/pUL ( 128 , 130 , 131A ) complexes incorporated into virions . We determined the gB and gH levels , and , representative for the presence of the gH/gL/pUL ( 128 , 130 , 131A ) complex , the pUL128 content in EC- and HFF-derived virions . Virus particles were pelleted from EC- or HFF-derived supernatants , lysed , and their gB , gH and pUL128 protein content determined by Western blot analysis . The amounts of gB and gH in virus pellets from HFF and EC supernatants always showed a constant relation ( data not shown ) . Yet , HFF-derived virus particles contained more pUL128 protein than EC-derived virus particles ( Fig . 4A ) . This could be quantitatively analysed by measuring the gB band intensities of the lysates , which reflect the particle amounts loaded , and then , relating the pUL128 band intensities to the respective gB bands ( Fig . 4A , middle panel ) . Remarkably , the pUL128/gB ratios mirror the TIME/HFF infection ratios ( Fig . 4B , lower panel ) . Thus , a low EC infection capacity correlated with a low level of gH/gL/pUL ( 128 , 130 , 131A ) complexes in virions . Interestingly , total cell lysates of the respective infected cells showed that EC and HFF expressed comparable amounts of pUL128 ( Fig . 4B ) . This indicated that the differences in EC infection capacities observed are created at a late stage during maturation or release of virus progeny . To exclude that the observed differences between EC- and HFF-derived supernatants are due to non-infectious particles or contaminations with cell membrane components , gradient-purified virus from infections of HFF and HUVEC were analysed in the Western blot as described above ( Fig . 4C ) . The pUL128/gB ratios again mirrored the TIME/HFF infection ratios of the supernatants , the virus was purified from . Whereas the virus released by EC is low in gH/gL/pUL ( 128 , 130 , 13A ) complexes , focal spread in EC cultures was highly efficient . Like EC infection by supernatant virus , it can be blocked by anti-pUL128 , anti-pUL130 and anti-pUL131A antibodies [16] , [17] , [44] , [45] . This indicates that pUL ( 128 , 130 , 131A ) are accessible to antibodies and promote infection of neighboring cells . We tested different cellular preparations for the presence of cell-associated EC-tropic virus . HUVEC and as a control HFF were infected with vBAC4-luc , and 6 days after infection supernatants were harvested . Cells were washed to remove loosely bound virus and then homogenized using cell douncers . Aliquots of the total homogenates , containing the disrupted cells and virus freed by cell disruption , were saved . Homogenates were then cleared by centrifugation at 3 , 500×g to separate supernatants containing virus , which can be released by physical disruption . The pellets of cell debris , containing virus which is not released from cells , were also resuspended . These four preparations were then tested on HFF and TIME cells by the luciferase assay . Virus supernatants from HFF and HUVEC showed a high and a low EC infection capacity , respectively ( Fig . 5A ) . All three homogenate preparations from HFF showed a reduced EC infection capacity , when compared to HFF supernatant virus . Notably , the two HUVEC preparations , which contained cell debris showed an about tenfold higher EC infection capacity than the HUVEC supernatants ( Fig . 5A ) . Thus , the progeny able to infect EC is released by HFF , but remained tightly associated with cellular structures in the case of EC . The differences observed are not due to different quantities of virus in the different preparations , because all HFF- and HUVEC-derived preparations showed high luciferase values , when tested on HFF ( Fig . 5B ) . The tenfold differences between HUVEC-derived preparations , containing broken cells , and those without cells are due to high and low luciferase values on TIME cells , respectively ( Fig . 5B ) . Highly EC-tropic virus could neither be released from EC by sonication nor by several rounds of freezing and thawing ( data not shown ) . Taken together , the data show that HFF readily release , whereas HUVEC tightly retain EC-tropic virus . Virus released from EC is only poorly tropic for EC , whereas virus associated with EC shows a much higher tropism for EC . Virus released from HFF is highly EC-tropic and virus found in the particulate fraction of disrupted HFF rather shows a lower tropism for EC . One explanation would be that HCMV progeny is heterogeneous and consists of distinct virus populations with regard to their EC-tropism . EC show a propensity to retain EC-tropic virus and release non EC-tropic virus , whereas HFF readily release both , EC-tropic and non EC-tropic virus . The hypothesis , that HFF progeny is a mixture of EC-tropic and non EC-tropic virus is testable by separation of EC-tropic and non EC-tropic virus . As HUVEC strongly retain EC-tropic virus , they might serve to specifically bind EC-tropic virus and deplete HFF virus progeny of its EC-tropic fraction . We preincubated HFF-derived supernatant virus with HUVEC or with HFF , pelleted the cells , and analysed the HFF and TIME cell infectivity of virus remaining in the supernatants ( Fig . 6A ) . Preincubation with HUVEC removed about 30 to 90% and preincubation with HFF about 99% of the infectious virus from supernatants , when tested on HFF ( data not shown ) . The TIME/HFF infection ratios of the non-bound virus in supernatants preincubated with HUVEC was drastically and significantly reduced to the level observed in supernatants of HCMV-infected HUVEC ( Fig . 6A ) . Preincubation with HFF in contrast , although it removed the bulk of infectivity , only weakly reduced the EC infection propensity of the non-bound virus ( Fig . 6A ) . Thus , HUVEC , which retained EC-tropic virus in infection , were a good matrix for binding EC-tropic virus , whereas HFF , which readily released EC-tropic virus into the supernatant , were a weak matrix for EC-tropic virus . The depletion of EC-tropism strongly suggested that HFF virus progeny was heterogeneous and composed of distinct virus populations , which could be sorted . To find out whether the depletion of EC-tropism is based on removing virus particles expressing the gH/gL/pUL ( 128 , 130 , 131A ) complex , we coincubated HFF-derived virus progeny with protein G sepharose beads to which we had bound anti-pUL131A antibodies . Beads coated with antibodies specific for pUL131A , but not uncoated beads or beads coated with preimmune serum , depleted about 70% of the EC-tropism ( Fig . 6B ) . This strongly implied that depletion of EC-tropism is through retaining virions , expressing the gH/gL/pUL ( 128 , 130 , 131A ) complex .
The use of different receptor binding proteins to mediate entry into different cell types , and the use of different entry pathways even into one cell type is a common feature of herpesvirus entry . Herpesviruses have additionally developed strategies , which may route infection in vivo . For EBV , the group of L . Hutt-Fletcher has pioneered the paradigm that epithelial cells produce a virus progeny high in gH/gL/gp42 complexes , which promotes B-cell infection . B-cells in turn , produce virus progeny low in gH/gL/gp42 complexes which efficiently infect epithelial cells , but not B-cells . Although not absolute , this relative switch of cell tropism after alternate replication in epithelial and B-cells directs infection from one cell type to the other . Here , we propose that also different producer cells of HCMV may direct the infection . gH/gL/gO complex formation is needed for release of infectious virus from any infected cell type tested so far [24] , [28] . Incorporation of gH/gL/pUL ( 128 , 130 , 131A ) complexes into virions is essential for infection of e . g . endothelial , epithelial , and dendritic cells , and for leukocytes [16] , [17] , [22] , [23] . If gO is missing , then the infection spreads predominantly focal and depends on the gH/gL/pUL ( 128 , 130 , 131A ) complex , even in cell types , which usually do not depend on this complex for infection [24] , [26] . Our initial observation was that virus spread in fibroblast cultures differed from virus spread in EC cultures [17] . Spread in fibroblast cultures appeared supernatant- driven , whereas spread in EC cultures was focal . A strictly cell-associated virus spread in EC cultures had also been observed by the group of G . Gerna , who reported that propagation in HUVEC strictly depended on passage of cells and could not be achieved by supernatant virus [43] . We offer an explanation for the focal spread in EC cultures by showing that EC predominantly release virus , which is not EC-tropic , but at the same time tightly retain EC-tropic virus which may then be transferred to neighboring cells only . Virus transfer was accessible to neutralizing antibodies and dependent on the gH/gL/pUL ( 128 , 130 , 131A ) complex , as we could show in Figure S1 . Focal spread was completely blocked by a neutralizing human antiserum and by anti-pUL131A antibodies . Interestingly , when infected HFF cultures were treated in a similar way , a neutralizing human antiserum blocked infection by free virus , but left a focal spread of virus , indicating that for HFF a direct cell-to-cell spread mechanism may be possible [46] . Anti-pUL131A antibodies could not at all inhibit spread in HFF cultures . These data confirmed earlier studies by our group , which showed that only gH/gL/pUL ( 128 , 130 , 131A ) dependent virus spread , like spread in EC cultures , or spread of a delta gO mutant in fibroblast cultures could be inhibited by anti-pUL131A antibodies [17] , [26] . Similar to the EBV model , supernatants from infected HFF showed a higher capacity to infect EC than EC-derived supernatants , and we could show that the biochemical basis for that is a respectively high and low content of gH/gL/pUL ( 128 , 130 , 131A ) complex in virions . The question , which arose then , was , what causes the observed difference in gH/gL/pUL ( 128 , 130 , 131A ) content . For EBV it has been described that in infected B-cells HLA-DR ß binds the gp42 protein of the gH/gL/gp42 complex , which promotes B-cell infection , holds it back intracellularly , and thus , makes it vulnerable for degradation . As a consequence , B-cells release mainly virions containing a two-part gH/gL complex , which cannot infect B-cells . Epithelial cells , which do not express HLA-DR ß , do not retain gp42 and thus , release virus , which contains more of the three-part gH/gL/gp42 complex . The mechanisms , by which the differences in the released populations of virions in HCMV are achieved , appear to be different . For HCMV , we found that EC and fibroblasts produce heterogeneous virus progenies . EC release a virus progeny , which is not EC-tropic , and retain a progeny , which is highly EC-tropic . HFF release an EC-tropic progeny . which can be depleted of its EC-tropism by using HUVEC or protein G sepharose beads coated with antibodies directed against pUL131A . This strongly suggested that HFF progeny is composed of distinct EC-tropic and non EC-tropic virus populations , and that the EC-tropic population most likely is a population with a high gH/gL/pUL ( 128 , 130 , 131A ) content . If HFF-derived virus progeny was homogeneous , a specific depletion only of EC-tropic virus would not be possible . Interestingly , HUVEC , which retain EC-tropic virus in infection experiments , were a good matrix to retain EC-tropic virus in the test tube , whereas HFF , which readily release EC-tropic virus in infections , were a bad matrix . Thus , we propose that the difference in cell tropism of virus released from EC and fibroblasts is the result of a sorting process . EC strongly and specifically retain EC-tropic virus through the gH/gL/pUL ( 128 , 130 , 131A ) complex . HFF release EC-tropic and non EC-tropic virus . Thus , for HCMV , not protein components of gH/gL complexes are retained in a cell-type specific manner , but rather mature virions carrying the gH/gL/pUL ( 128 , 130 , 131A ) complex in their envelopes . Figure 7 depicts the EBV and HCMV models for virus spread side by side . Future experiments will have to show where and how EC-tropic virus is held back . It has recently been shown that overexpression of gH/gL/pUL ( 128 , 130 , 131A ) in epithelial cells interferes with HCMV infection . It has been postulated that this reflects binding of the gH/gL/pUL ( 128 , 130 , 131A ) complex to the respective entry receptor [34] . This was not observed for fibroblasts and thus , an HCMV entry receptor binding to gH/gL/pUL ( 128 , 130 , 131A ) and expressed on EC would be a good candidate also for retaining EC-tropic virus by infected EC . How virus is then transferred to neighboring cells , will also have to be investigated in the future . An attractive model would be a mechanism as described for MHV-68 , for which it has been shown that virus particles attached to and moving on plasma membrane fronds are directly transferred to neighboring cells [47] . Assuming that gH/gL/pUL ( 128 , 130 , 131A ) complexes are incorporated into virus progeny at random , EC-tropism of a virus particle might be defined by a threshold level of gH/gL/pUL ( 128 , 130 , 131A ) complexes . Accordingly , high levels of gH/gL/pUL ( 128 , 130 , 131A ) complexes in turn could also block release from EC . Thus , the levels of gH/gL/pUL ( 128 , 130 , 131A ) complexes could define whether a particle is EC-tropic or not , whether it is retained by EC during infection , and whether it can be depleted from supernatants by EC-preincubation . This could also explain , why progeny of a ΔgO virus , which expresses only the gH/gL/pUL ( 128 , 130 , 131A ) complex , readily spreads cell-associated and can barely be released from EC [24] , [28] . ΔgO virus progeny can equally well infect EC and HFF ( Fig 2D ) . Wildtype TB40-BAC4 virus progeny , in contrast , shows a higher propensity to infect HFF ( Fig . 2D ) , which could be explained by being a mixture of EC-tropic and non EC-tropic particles . For EBV , it has been observed that virus bound to the surface of resting B cells is 103-104 times more infectious for epithelial cells than cell-free virus [48] . For HCMV it has not yet been tested whether surface-bound virus could promote a switch of cell tropism . We restricted our experiments to endothelial cells and fibroblasts . Macrophages , dendritic cells , and epithelial cells also strictly depend on the gH/gL/pUL ( 128 , 130 , 131A ) complex for their infection . Whether their infection also follows the pattern of the EC infection shown here , will have to be investigated in the future . Recently , it has been published by Wang et al . [30] that HCMV progenies derived from epithelial cells and fibroblast also differ . They reported that both cell types release progenies which can readily infect epithelial cells and fibroblasts , but differ with respect to the pathway they use to enter epithelial cells . They found a twofold higher gH/gL/pUL ( 128 , 130 , 131A ) content in epithelial cell-derived particles , which they considered as marginal . As they used an AD169 mutant , in which UL131A had been repaired , it will have to be clarified , whether their findings reflect that epitheliotropic virus produced in epithelial cells is , in contrast to our findings in EC , not retained , or whether the observed differences are due to differences of the HCMV strains used . It has recently been shown that AD169 incorporates gO into virions , whereas HCMV strain TR does not [27] . This suggests that strain-specific differences may indeed affect gH/gL-dependent processes . Whether our observations made in cell culture , reflect features valid for all HCMV strains , and what role a switch in tropism and spread patterns may play in vivo , will be the subject of future research . It will be of particular interest to find out whether the relative propensity of different cell types to release virus plays a crucial role in establishment of infection and transfer of virus to new hosts or the fetus . For HCMV , it has been shown that primary isolation of EC-tropic virus depends on infected cells as a source of virus , whereas fibroblast infection can also be achieved with cell-free virus sources like throat washes and amniotic fluid [43] . This might already be an indication that cells lining the compartments , where these fluids are produced , do not release EC-tropic virus .
Primary human foreskin fibroblasts ( HFF ) ( PromoCell , Germany ) were used from passage 12 to 22 and maintained in Dulbecco's modified Eagle's medium ( DMEM ) supplemented with 10% fetal calf serum , 2 mM L-glutamine , 100 units/ml penicillin and 100 µg/ml streptomycin . Primary human umbilical vein endothelial cells ( HUVEC ) ( LONZA , USA ) were used from passage 1 to 6 . HUVEC and TIME ( telomerase-immortalized human microvascular endothelial ) cells [49] were maintained in an EGM-2 MV BulletKit medium system ( LONZA , USA ) . The HCMV strains used were VR1814 [19] , TB40/E [50] and TB40/E cloned as a BAC ( TB40-BAC4 ) [51] . The HCMV strain TB40/E cloned as a bacterial artificial chromosome ( BAC ) ( TB40-BAC4 ) [51] was used for HCMV BAC mutagenesis . A 48 bp FRT site was inserted into the TB40-BAC4 , thereby disrupting the open reading frames ( ORFs ) UL5 , UL6 , UL7 , UL8 and UL9 . Briefly , a linear PCR fragment containing a kanamycin-resistance gene flanked by two 48 bp FRT sites and sequences homologous to the HCMV UL5 and UL9 coding regions was generated using the primers UL5pcp15for ( 5′-ATGTTTCTAGGCTACTCTGACTGTGTAGATCCCGGCTTTGCTGTATATCGTGTATCTAGACGGGGGTGTCCAGGGTTTTCCC-3′ ) and UL9pcp15rev ( 5′-ATTGTTGTAACGATAACTAAGGGTATGATCCACATTGTATGTGGGGTGGCAGTATCGTGTCTTCCGGCTCGTATGTTGTGTGG-3 ) and pCP15 as template [52] . The PCR product was inserted into TB40-BAC4 by homologous recombination in E . coli , thereby deleting 3 , 066 kb . The kanamycin-resistance gene was subsequently excised by FLP-mediated site-directed recombination [53] , and the resulting BAC mutant called BAC4-FRT5-9 . To generate a luciferase reporter HCMV , the SV40-driven firefly luciferase expression cassette was excised from pGL3-promoter ( Promega ) with Sal I and Bgl II , filled in by Klenow polymerase and inserted into the pOriR6K-zeo plasmid linearized by EcoR V . The resulting plasmid pO6-Luc was inserted into BAC4-FRT5-9 via FLP-mediated FRT recombination mutagenesis using the temperature-sensitive expression plasmid pCP20 [54] . The resulting BAC mutant was called BAC4-Luc . The BAC mutants BAC4-Luc/ΔgO and BAC4-Luc/UL131Astop were cloned into the BAC4-Luc background as described previously [24] , [26] . Deletions and insertions were controlled by restriction pattern analysis and subsequent sequencing . BACmids were reconstituted to virus by transfection of BAC DNA into HFF using FugeneHD transfection reagent ( Roche Diagnostics ) according to the manufacturer's instructions . Transfected cells were propagated until viral plaques appeared and the supernatants from these cultures used for further propagation of virus . Virus stocks were prepared from supernatants of infected HFF , HUVEC or TIME cells . Supernatants were cleared of cellular debris by centrifugation for 15 min at 3 , 500×g and stored at −80°C . For Western Blot analysis of HCMV particles , virus was concentrated from cell culture supernatants . Briefly , 200 ml supernatant from infected cells showing about 90% CPE was cleared of cellular debris by centrifugation at 3 , 500×g for 15 min . Then , virus was pelleted from cleared supernatant by ultracentrifugation at 80 , 000×g for 70 min . Virus pellets were resuspended in 1 . 5 ml 0 . 04 mol/l sodium phosphate pH 7 . 4 . Virus titers of cleared supernatants were determined by a TCID50 assay performed on 96 well plates on HFF . To infect cells , medium was removed from 90% confluent cell monolayers and replaced by virus diluted in DMEM containing 5% FCS . For some experiments , virus infection was enhanced by a centrifugation step ( 30 min , 860×g at room temperature ) , followed by incubation at 37°C for 90 min . To compare infectivity of virus derived from fibroblasts and endothelial cells , subsequent infections were performed in DMEM 5% FCS/EGM-2 mixed at a ratio of 1∶1 , to exclude medium effects . During infections , medium was exchanged every second day in a way that supernatants harvested contained virus released during the preceding 48 hours . As HCMV in general more readily infects fibroblasts than EC , in all experiments , where infections of EC and fibroblasts ( spread patterns and growth curves ) were compared , the infections were adapted in a way that EC were infected with more virus than fibroblasts to achieve comparable numbers of ie1-positive cells after 48 hours . For gradient purification of virions , supernatants from infected cell cultures showing approximately 100% late-stage CPE were cleared of cell debris by centrifugation for 10 min at 2 , 800×g . Supernatants were then ultracentrifuged for 70 min at 80 , 000×g . Pellets containing virions were resuspended in 1 ml PBS and transferred onto a preformed , linear glycerol/tartrate gradient ( 15–35% sodium tartrate and 30–40% glycerol in 0 . 04 mol/l sodium phosphate pH 7 . 4 ) , which was ultracentrifuged for 45 min at 80 , 000×g . The virion-containing band was harvested with a syringe and the virions were washed and pelleted by an additional ultracentrifugation for 70 min at 80 , 000×g . The pellet was resuspended in 0 . 04 mol/l sodium phosphate . HCMV-infected cells were fixed in 50% acetone/50% methanol , stained using a mouse anti-ie1 antibody ( anti-ie1; Perkin Elmer ) and detected with a Cy3-coupled goat anti-mouse antibody ( Dianova ) . For counterstaining of cell nuclei , cells were incubated in PBS containing 5 µg/ml Hoechst 333258 ( Invitrogen ) for 1 min . HFF and TIME cells were grown in 96 well plates ( 20 , 000 cells/well ) and infected in triplicates at an m . o . i . between 0 . 02 and 0 . 5 for 90 min . Inoculi were then replaced by medium supplemented with 300 µg/ml phosphono acetic acid ( PAA ) . 48 h after infection cells were lysed in 50 µl lysis buffer ( 25 mM Tris/H3PO4 , 2 mM CDTA , 2 mM DTT , 10% glycerol , 5% Triton-X 100 ) and luciferase activity was determined for 20 µl of lysate with a luciferase assay system ( Promega ) according to the manufacturer's instructions . Virus particles or infected cells were lysed in 6× sample buffer ( 300 mM Tris-HCl ( pH 6 . 8 ) , 10% SDS , 30% glycerol , 5% ß-mercaptoethanol , 0 . 01% ( w/v ) bromphenolblue , 0 . 01% ( w/v ) phenolred ) , separated on 15% polyacrylamide gels and transferred onto nitrocellulose ( Amersham Biosciences ) . Membranes were blocked with 5% low-fat milk in TBS and stained for gB or pUL128 using mouse anti-gB antibody ( 2F12; Abcam ) or mouse anti-pUL128 antibody ( 4B10 , kindly provided by T . Shenk , University of Princeton , USA ) , respectively . The specific protein bands were detected by using an peroxidase-coupled anti-mouse antibody ( Dianova ) and the SuperSignal West Dura Extended Duration Kit ( Perbio ) . The intensities of protein bands were quantified using the Fujifilm Intelligent Light Box LAS-300 and the Image reader LAS-300 . Non-saturated light signals were analysed to determine the protein amounts using ImageQuant 5 . 0 software . The pUL128 levels were related to gB levels of the respective samples . Cells were infected in 6 cm dishes , and 6 days after infection the supernatants ( 4 ml ) harvested and cleared of cellular debris ( 3 , 500×g , 15 min ) . Cell monolayers were washed with cold PBS , scraped and cells dounced in 4 ml DMEM medium supplemented with 5% FCS using tight fit hand homogenizers ( Sartorius-Stedium ) . 3 . 5 ml of the total homogenized cells were pelleted ( 3 , 500×g , 15 min ) , the supernatant removed ( supernatant of homogenized cells ) and the pellets resuspended in fresh 3 . 5 ml DMEM medium supplemented with 5% FCS . GeneBank/EMBL/DDBJ accession number for TB40-BAC4 is EF999921 .
|
gH/gL complexes of herpesviruses are supposed to promote fusion of the viral envelope with cellular membranes . The gH/gL core complex associates with additional proteins which define the tropism for certain cell types by promoting binding to specific receptors . Two alternative gH/gL complexes of human cytomegalovirus ( HCMV ) define the cell tropism , the entry pathway and the spread of virus . Formation of a gH/gL/gO complex during infection determines release of infectious virus into the supernatant . The gH/gL/pUL ( 128 , 130 , 131A ) complex determines the tropism for endothelial cells ( EC ) and promotes focal spread . Here , we could show that HCMV-infected cells produce EC-tropic and non EC-tropic virus populations . While fibroblasts release both populations into the supernatant , EC predominantly release the non EC-tropic population . Different host cells of HCMV thus may direct the distribution of virus progeny .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"virology/virion",
"structure,",
"assembly,",
"and",
"egress",
"virology/host",
"invasion",
"and",
"cell",
"entry",
"virology"
] |
2011
|
HCMV Spread and Cell Tropism are Determined by Distinct Virus Populations
|
Rasa3 is a GTPase activating protein of the GAP1 family which targets Ras and Rap1 . Ubiquitous Rasa3 catalytic inactivation in mouse results in early embryonic lethality . Here , we show that Rasa3 catalytic inactivation in mouse hematopoietic cells results in a lethal syndrome characterized by severe defects during megakaryopoiesis , thrombocytopenia and a predisposition to develop preleukemia . The main objective of this study was to define the cellular and the molecular mechanisms of terminal megakaryopoiesis alterations . We found that Rasa3 catalytic inactivation altered megakaryocyte development , adherence , migration , actin cytoskeleton organization and differentiation into proplatelet forming megakaryocytes . These megakaryocyte alterations were associated with an increased active Rap1 level and a constitutive integrin activation . Thus , these mice presented a severe thrombocytopenia , bleeding and anemia associated with an increased percentage of megakaryocytes in the bone marrow , bone marrow fibrosis , extramedular hematopoiesis , splenomegaly and premature death . Altogether , our results indicate that Rasa3 catalytic activity controls Rap1 activation and integrin signaling during megakaryocyte differentiation in mouse .
Ras families GTPase-activating proteins ( GAP ) , like Ras GAPs , Rho GAPs and Arf GAPs , are tumor suppressors as the loss of their GAP activity allows uncontrolled Ras , Rho and Arf activities and promotes cancer . Rasa3 ( or GAP1IP4BP , R-Ras GAP ) is a member of the Ras GAP1 subfamily with Rasa2 ( or GAP1m ) , Rasa4 ( or Capri ) and Rasal ( or Rasal1 ) [1]–[5] . This Ras GAP subfamily is known to function as dual GAP for Ras an Rap-GTPases [6] , [7] . Rasa3 protein structure is characterized by a conserved basic domain structure comprising two N-terminal tandem C2 domains , a central GAP domain and a C-terminal pleckstrin homology ( PH ) domain that is associated with a Bruton's tyrosine kinase ( Btk ) motif [8] . Binding of the latter domain to phosphoinositides determines Rasa3 targeting to the cytosolic leaflet of the plasma membrane where it inactivates Ras and Rap1 [9]–[11] . Down-regulation of Rasal and Rasa4 induces cellular transformation in vitro [12] , [13] , and Rasal is down-regulated in multiple human tumors by epigenetic silencing [14] . Rasa4 inactivation in mouse leads to impaired macrophages Fcγ receptor-mediated phagocytosis and oxidative burst , as well as to increased bacterial infection [15] . No clear definition of Rasa2 function in vivo is currently available . Mutant mice expressing a catalytically-inactive Rasa3 protein have been reported to die at mid embryonic life [16] . Indeed , removal of exons 11 and 12 of the mouse Rasa3 gene , 2 exons which are essential for the Ras GAP activity , leads to the expression of a 88 amino acids-truncated but catalytically inactive Rasa3 protein [16] . Phenotypically , Rasa3 mutant embryos present massive subcutaneous and intraparenchymal hemorrhages probably consecutive to abnormal adherens junctions between capillary endothelial cells [16] . Multiple roles for Ras and Rap1 , the Rasa3 targets , have been defined in hematopoietic cells: these proteins control cellular proliferation , differentiation , migration and adhesion . In particular , Rap1 has been implicated in the maturation of megakaryocytes and the pathogenesis of chronic myelogenous leukemia [17] . Here , we found that catalytic inactivation of Rasa3 specifically in the hematopoietic system results in a lethal syndrome characterized by major alterations during megakaryopoiesis . These alterations were associated with increased active Rap1 level and constitutive integrin activation in megakaryocytes , a phenotype quite different clinically , biologically and mechanistically from that of recently published mice with a spontaneous missense mutation between the two N-terminal tandem C2 domains of Rasa3 [18] .
In order to study the specific effects of a catalytically-inactive Rasa3 mutant protein on the hematopoietic system and to circumvent the early embryonic lethality reported in Rasa3−/− mice , we used irradiated Severe Combined Immune Deficient ( SCID ) mice reconstituted with E12 . 5 liver cells derived from Rasa3+/+ , Rasa3+/− or Rasa3−/− embryos . SCID mice were first analyzed 6 weeks after irradiation/reconstitution: all Rasa3 genotypes were able to reconstitute the lymphoid compartment in irradiated SCID mice since no significant difference was detected between SCID-Rasa3+/+ , SCID-Rasa3+/− and SCID-Rasa3−/− mice in total numbers of splenic T and B cells ( Table S1 ) . No significant difference was observed in red blood cell , blood platelet and bone marrow megakaryocyte counts as well as spleen weight between SCID-Rasa3+/+ and SCID-Rasa3−/− mice at this stage ( Table S1 ) . More than 80% of SCID mice reconstituted with Rasa3−/− cells died within 14 months after reconstitution while , at the same time , about 95% of SCID-Rasa3+/+ and SCID-Rasa3+/− mice were still alive ( Fig . 1a ) . Pathological analysis revealed that 85% of SCID-Rasa3−/− mice presented with thoracic and/or peritoneal hemorrhages ( data not shown ) and that more than 80% had a splenomegaly ( Fig . 1b ) . Our results below present the analysis of a total of 24 moribund SCID-Rasa3−/− mice . Among these 24 mice , 20 had a megakaryocytic dysplasia associated with a severe thrombocytopenia , and the remaining 4 developed a preleukemia . The main objective of this study was to define the cellular and the molecular mechanisms of the megakaryocytic dysplasia . In ∼80% ( 20/24 ) of SCID-Rasa3−/− mice , blood analysis revealed a thrombocytopenia ( Fig . 1c ) . Thrombocytopenia in these mice was associated with megakaryocyte alterations . A significant increase in the percentage of CD41+ megakaryocytes was observed in the bone marrow of SCID-Rasa3−/− mice ( Fig . 1d ) . These megakaryocytes were morphologically abnormal and presented a marked increase in the intensity of von Willebrand factor ( vWF ) staining , as compared with SCID-Rasa3+/+ megakaryocytes ( Fig . 1e ) . vWf+ deposits were also abnormally detected along the SCID-Rasa3−/− diaphysis ( Fig . 1e , arrowheads ) . In addition , more megakaryocytes were present in the osteoblastic niche in SCID-Rasa3−/− mice , as compared with SCID-Rasa3+/+ mice ( Fig . 1f , left panel ) . Inversely , the vascular niche hosted less megakaryocytes in mutant mice ( Fig . 1f , right panel ) . The stem and megakaryocyte progenitor cell compartments were characterized in the bone marrow of SCID-Rasa3+/+ and SCID-Rasa3−/− mice by flow cytometry . A similar percentage of live c-Kit+/Lin− cells was detected in the bone marrow of these mice , and the proportion of Sca-1+ cells within this population was also not significantly different in SCID-Rasa3+/+ and SCID-Rasa3−/− mice ( Table S2 ) . Staining of these c-Kit+ Lin− Sca-1+ ( KLS ) cells with CD34 and Flk-2 antibodies defined the KLS-CD34− Flk-2− hematopoietic stem cell compartment; but again , no difference was detected in the percentage of these cells between SCID-Rasa3+/+ and SCID-Rasa3−/− mice ( Table S2 ) . It has been recently shown that bone marrow progenitors with megakaryocyte potential reside in the Lin− c-Kit+ FcγRII/IIIlo Sca-1− CD150+ cell population [19] . This cell population was decreased in the bone marrow of SCID-Rasa3−/− mice , as compared with SCID-Rasa3+/+ , but the difference did not reach statistical significance ( Table S2 ) . Femur sections revealed the presence of a fibrosis characterized by numerous collagen trabeculae in the cavity of SCID-Rasa3−/− femurs , while the cavity of SCID-Rasa3+/+ femurs was totally free of collagen trabeculae ( Fig . S1a ) . Consequently , the number of nucleated cells recovered on average from one femur of SCID-Rasa3−/− mice was significantly reduced , as compared with SCID-Rasa3+/+ mice ( SCID-Rasa3+/+: 20 . 9±2 . 8×106 nucleated cells , n = 12; SCID-Rasa3−/−: 6 . 4±1 . 1×106 nucleated cells , n = 19; P<0 . 001 , unpaired t test ) . Bone marrow cell density was similar in the cavity of SCID-Rasa3+/+ and SCID-Rasa3−/− femurs ( Fig . S1b ) , and similar percentages of CD117+/c-Kit+ progenitor cells and Ter119+ CD71+ erythroblasts were detected in SCID-Rasa3+/+ and SCID-Rasa3−/− bone marrow cells ( Table S2 and data not shown ) . An increased splenic hematopoiesis - including megakaryopoiesis - , associated with a disorganized pulp architecture , and foci of liver hematopoiesis were also observed in SCID-Rasa3−/− mice ( Table S3 and Fig . S2 ) . Thrombocytopenia and hemorrhages in SCID-Rasa3−/− mice were associated with a regenerative anemia , whereas normal counts were maintained for total white blood cell and circulating neutrophil , lymphocyte , monocyte and eosinophil ( Fig . S3 and Table S4 ) . Thrombopoietin ( TPO ) levels were significantly decreased in SCID-Rasa3−/− mice , as compared with SCID-Rasa3+/+ mice , a probable consequence of the markedly increased Mpl+/CD150+ megakaryocyte number in the spleen of these mice ( TPO level in SCID-Rasa3+/+ mice: 1732±211 pg/ml , n = 11; TPO level in SCID-Rasa3−/− mice: 653±74 pg/ml , n = 16; mean ± SEM; P<0 . 001 ) . Collectively , our results indicate that the loss of Rasa3 catalytic activity in 20/24 SCID-Rasa3−/− mice leads to megakaryocyte alterations , to thrombocytopenia , hemorrages and a regenerative anemia . Bone marrow was isolated from SCID-Rasa3 mice 2 months after irradiation/reconstitution and cultured under a confocal microscope . Despite a ∼2-fold increased percentage of megakaryocytes in the SCID-Rasa3−/− bone marrow , there was a trend for a decreased number of megakaryocytes released from SCID-Rasa3−/− bone marrow explants , as compared with SCID-Rasa3+/+ explants ( Fig . 2a , left panel ) . Released SCID-Rasa3−/− megakaryocytes were able to spread on the culture plate but never fully differentiated in proplatelet forming megakaryocytes ( Fig . 2a , centre and right panels , and Fig . 2b ) . The linear distance covered by the released megakaryocytes and their velocity were significantly lower in SCID-Rasa3−/− than in SCID-Rasa3+/+ explants ( Fig . 2c ) . In order to further analyze the role of Rasa3 in megakaryocyte adhesion and differentiation , we used megakaryocytes obtained from Rasa3+/+ and Rasa3−/− fetal liver cells ( FLC ) cultured in the presence of TPO . This cellular model is simpler and faster than the model of bone marrow explants isolated from SCID-Rasa3 mice , and it recapitulated the megakaryocyte defects previously observed in the later model . Indeed , after 2 days of TPO treatment , flow cytometry analysis detected a significant 1 . 42-fold increase in the percentage of CD41+ megakaryocytes in the Rasa3−/− FLC culture , as compared with Rasa3+/+ culture ( Fig . 3a ) . CD41+ megakaryocytes with 16N and 32N ploidy were significantly increased in these Rasa3−/− FLC cultures , as compared with Rasa3+/+ FLC cultures ( Fig . 3b ) . Colony-forming unit-megakaryocyte ( CFU-Mk ) assay using FLC revealed a significant decrease in the number of small immature megakaryocyte colonies in Rasa3−/− cell culture , as compared with Rasa3+/+ cell culture ( Fig . 3c ) . However , numerous large mature megakaryocytes were detected in the Rasa3−/− CFU-Mk assay , while not in the Rasa3+/+ CFU-Mk assay ( Fig . 3c , arrowheads ) . After 6 days of TPO treatment , many proplatelets were observed in the Rasa3+/+ FLC culture whereas , in the Rasa3−/− FLC culture , no proplatelets were detected ( Fig . 3d ) . Instead of proplatelets , many abnormal adherent cells were detected in the Rasa3−/− FLC culture that were CD41 positive , indicating their megakaryocyte origin ( Fig . 3d , lower panels and Fig . 3e ) . Actin cytoskeleton staining of these abnormal adherent Rasa3−/− megakaryocytes revealed a unique dotted actin pattern without stress fiber at the contact with the culture plate , significantly different from the expected actin stress fiber pattern observed in the few adherent Rasa3+/+ megakaryocytes present in the FLC culture at day 6 of TPO treatment ( Fig . 4a , bottom , and 4b ) . The actin cytoskeletal organization was also altered at the top of the adherent Rasa3−/− megakaryocytes: actin was decreased at the periphery and much more concentrated at the center of the cell , as compared with adherent Rasa3+/+ megakaryocytes ( Fig . 4a , top ) . Collectively , these results indicate that Rasa3−/− FLC abnormally develop into mature megakaryocytes , and that Rasa3−/− megakaryocytes derived from FLC culture have an altered actin cytoskeleton organization associated with an abnormal adherent phenotype , a reduced motility and an absence of normal terminal differentiation in proplatelets . Interestingly , this Rasa3−/− megakaryocyte phenotype ( i . e . defect in proplatelet formation , dotted actin cytoskeletal pattern with reduced stress fibers and abnormal adherent megakaryocytes ) resembles that of rare thrombocytopenic patients with a constitutive αIIbβ3 integrin activity caused by specific mutations in ITGA2B or ITGB3 genes [20]–[22] . Soluble fibrinogen binding to αIIbβ3 integrin present at the megakaryocyte surface is regulated by inside-out signaling which determines the affinity/avidity of the integrin for its ligand . In the absence of megakaryocyte stimulation , only little amount of soluble FITC-fibrinogen bound to day 3 FLC culture-derived Rasa3+/+ mature megakaryocytes ( Fig . 5a ) . By contrast , in this resting condition , a larger amount of soluble FITC-fibrinogen bound to Rasa3−/− mature megakaryocytes , reaching the binding level of Rasa3+/+ megakaryocytes when stimulated by TPO for 30 min ( Fig . 5a ) . Stimulation of Rasa3−/− mature megakaryocytes by TPO did not further increase soluble FITC-fibrinogen binding . Importantly , no difference in αIIb/CD41 surface expression was detected by flow cytometry between day 3 FLC-derived Rasa3+/+ and Rasa3−/− mature megakaryocytes ( Rasa3+/+: 1234±70 arbitrary units ( A . U . ) , Rasa3−/−: 1084±244 A . U . , n = 3 independent experiments , P = 0 . 11 ) , suggesting that Rasa3−/− megakaryocytes have a constitutively activated inside-out signaling leading to a constitutive binding of soluble fibrinogen to αIIbβ3 integrin . Staining of day 3 FLC culture-derived Rasa3+/+ and Rasa3−/− mature megakaryocytes with the JON/A antibody , which selectively binds to the high affinity conformation of integrin αIIbβ3 , confirmed this hypothesis: a significant increase of JON/A+ megakaryocyte percentage was detected in Rasa3−/− megakaryocytes in resting and TPO-stimulated conditions , as compared with Rasa3+/+ megakaryocytes ( percentage of JON/A+ megakaryocytes , mean ± SEM: non-stimulated Rasa3+/+ megakaryocytes: 14 . 0±1 . 9%; non-stimulated Rasa3−/− megakaryocytes: 70 . 1±3 . 4% , P = 0 . 002; TPO-stimulated Rasa3+/+: 23 . 3±0 . 5%; TPO-stimulated Rasa3−/− megakaryocytes: 73 . 1±3 . 4% , P = 0 . 02 ) . Integrin activation triggers megakaryocyte adhesion to immobilized integrin ligands like collagen-I or fibrinogen and an outside-in signaling , resulting in the reorganization of the actin filaments and the modification of the cell shape [23] . Megakaryocytes from FLC-Rasa3+/+ cultured on day 3 adhered to collagen-I- and fibrinogen-coated plates , but nearly not to Poly-D-Lysine-coated plates , as expected ( Fig . 5b ) . Adherence to immobilized collagen-I and fibrinogen resulted in cell spreading reaching diameters over 50 µm in a limited number of Rasa3+/+ megakaryocytes , as described ( Fig . 5c , red dots , and Fig . 5d ) ( 24 ) . Adherence to Poly-D-Lysine- , collagen-I- and fibrinogen-coated plates was significantly higher in Rasa3−/− than in Rasa3+/+ megakaryocytes ( Fig . 5b ) . The percentage of megakaryocytes with a diameter over 50 µm was significantly increased in the Rasa3−/− culture , as compared with the Rasa3+/+ culture ( Fig . 5d ) . Outside-in integrin activation triggers the binding of the cytoskeletal protein talin to membrane integrins [24] . In association with their abnormal adhesion properties , Poly-D-Lysine adherent Rasa3−/− megakaryocyte recruited more talin to their membrane , as compared with Rasa3+/+ megakaryocytes ( Fig . 5e ) Collectively , our results indicate that Rasa3−/− megakaryocytes have a constitutively activated inside-out αIIbβ3 integrin signaling associated with major alterations in outside-in integrin signaling leading to cell adherence and spreading independently of integrin ligands . Since the small GTPase Rap1 , a Rasa3 substrate , controls inside-out and outside-in integrin signaling in megakaryocytes and platelets [25]– , the level of active , GTP-bound Rap1 was analyzed in mature Rasa3+/+ and Rasa3−/− megakaryocytes . A significant 2 fold-increase in active Rap1 was observed in Rasa3−/− megakaryocytes , as compared with Rasa3+/+ megakaryocytes , providing a molecular mechanism for the link between Rasa3 , talin and integrin activation ( Fig . 6a; P = 0 . 019 ) . By contrast , no significant difference was detected in the level of active GTP-bound Ras between Rasa3+/+ and Rasa3−/− megakaryocytes ( GTP-Ras mean fluorescence intensity: Rasa3+/+ megakaryocytes: 670 . 1±117 . 9 arbitrary units ( A . U . ) ; Rasa3−/− megakaryocytes: 706 . 2±56 . 7 A . U . ; P = 0 . 48; 3 independent experiments , 50 megakaryocytes analyzed per field , 2 fields per FLC culture ) . In order to confirm the important role of Rap1 in the abnormal adherent phenotype of Rasa3−/− megakaryocyte , outside-in experiments were performed in the presence of the Rap1 inhibitor GGTI-298 , which is not active on Ras . Addition of GGTI-298 to the culture medium completely abolished the abnormal adhesion phenotype of Rasa3−/− megakaryocyte , but had no effect on the proplatelet phenotype ( Fig . 6b and 6c , and data not shown ) . Altogether , these results indicate that the absence of Rasa3 increases Rap1 activation , and that Rap1 rather than Ras is probably responsible for the abnormal Rasa3−/− adherent megakaryocyte phenotype . Thus , increased Rap1 activation in the absence of Rasa3 leads to constitutive activation of integrins and increased outside-in signaling . Platelet adhesion and activation were analyzed on platelets isolated from adult Rasa3+/+ and Rasa3+/− mice ( Fig . S4 ) . Rasa3+/− platelets adhesion to BSA-coated plates was significantly increased compared with Rasa3+/+ platelets ( Fig . S4a ) . On fibrinogen-coated plates , a trend for an increase adhesion was detected in Rasa3+/− platelets , but the difference with Rasa3+/+ platelets did not reach statistical significance ( platelet counts per field of view ( FOV ) ( means ± SEM ) ; Rasa3+/+ platelets: 71±18 platelets/FOV; Rasa3+/− platelets: 114±39 platelets/FOV; P = 0 . 12 ) . In resting condition , two platelet activation markers were found altered in Rasa3+/− platelets: the JON/A antibody binding to Rasa3+/− platelets and the percentage of CD62P P-selectin positive Rasa3+/− platelets were significantly increased , as compared with Rasa3+/+ platelets ( Fig . S4b and S4c ) . No difference in the percentage of CD62P+ platelets was detected after stimulation with ADP or CRP ( Fig . S4c ) . In resting condition , we found no difference in CD61 expression on Rasa+/+ and Rasa3+/− platelets , whereas CD41 expression was significantly reduced on Rasa+/− platelets; this data indicates that the increased JON/A binding to Rasa3+/− platelets is not simply a consequence of an increased αIIβ3 integrin expression ( Fig . S4d ) . Finally , platelet aggregation after ADP stimulation was similar in Rasa3+/+ and Rasa3+/− platelets ( Fig . S4e ) . Altogether , these results indicate that Rasa3+/− platelets present adhesion and activation defects in resting conditions , suggesting that a similar pathological mechanism is present both in megakaryocytes and platelets . In the ∼20% ( 4/24 ) remaining SCID-Rasa3−/− mice , a very different phenotype was observed: a massive and homogeneous cellular infiltration was detected in the bone marrow and spleen , suggestive of a leukemia ( Fig . 7a and data not shown ) . Adult naïve SCID mice intraperitoneally injected with 107 splenocytes isolated from these SCID-Rasa3−/− mice did not develop a similar proliferative disorder within 4 months after injection , suggesting the presence of a preleukemia rather than a leukemia in these 4 SCID-Rasa3−/− mice ( data not shown ) . No fibrosis was detected in the bone marrow of these 4 SCID-Rasa3−/− mice . Flow cytometry analysis with a panel of antibodies revealed that cells massively infiltrating the bone marrow and the spleen were positive for CD117/c-Kit , CD38 and Sca-1 , and negative for all other cell surface markers tested , including B220 , CD3 , MAR-1 , Gr1 , Mac1 , Ter119 , CD71 , CD4 , CD34 and F4 . 80 ( Fig . 7b , 7c and data not shown ) . As expected , the percentage of B220+ , CD3+ , Gr1int Mac1+ , Ter119+ CD71+ , CD41+ and F4 . 80+ cells was significantly decreased in the bone marrow and the spleen of these 4 mice ( data not shown ) . These 4 mice had a reduced survival ( survival range: 6–11 months after SCID mice irradiation/reconstitution ) and a splenomegaly ( spleen weight range: 0 . 185–1 . 062 g ) . Collectively , these results indicate that about 20% of SCID-Rasa3−/− mice develops a preleukemia with a massive infiltration of bone marrow and spleen with CD117+ Sca-1+ CD38+ cells , probably leading to bone marrow failure and premature death . They also suggest that Rasa3 is a potential tumor suppressor gene , acting may be on Ras , as proposed by Blanc et al . [18] . However , the level of active , GTP-bound Ras was similar in CD117+/c-Kit+ hematopoietic stem cells derived from Rasa3+/+ and Rasa3−/− FLC cultures ( Fig . 7d ) .
Using a Rasa3 catalytic mutant in FLC and irradiated/reconstituted SCID models , we show here that Rasa3 catalytic activity controls megakaryocyte development and differentiation into proplatelet forming megakaryocytes . In the irradiated/reconstituted SCID model , these megakaryocyte alterations are associated with thrombocytopenia , bleeding , regenerative anemia and decreased survival , as well as with bone marrow fibrosis , extramedular hematopoiesis and splenomegaly . An increased percentage of mature megakaryocytes with an abnormal morphology was detected in bone marrow cells from irradiated/reconstituted SCID mice when Rasa3 catalytic activity was inactivated . This increased percentage was associated with a slightly decreased percentage of progenitors with megakaryocyte potential , suggestive of a megakaryopoisis alteration . An obvious megakaryopoiesis alteration was also detected in Rasa3−/− FLC culture , where the number of CFU for immature megakaryocyte was significantly decreased and associated with the presence of numerous mature megakaryocytes . Ploidy in these Rasa3−/− abnormal megakaryocytes was also slightly altered . On the contrary to active Ras level , level of active GTP-bound Rap1 was significantly increased in Rasa3−/− megakaryocytes . Interestingly , the small GTPase Rap1 is both a Rasa3 substrate and a well known regulator of integrin signaling in megakaryocytes and platelets [2] , [25]–[29] . Both inside-out and outside-in integrin signaling are controlled by Rap1 , including αIIbβ3 signaling . Thus , the increased active GTP-bound Rap1 level detected in Rasa3−/− megakaryocytes represents a plausible molecular mechanism linking Rasa3 to integrin signaling and the altered megakaryocyte development and differentiation . Indeed , altered inside-out and outside-in integrin signaling in Rasa3−/− megakaryocytes probably results in the adherence and motility defects that we observed in this study . These defects may secondarily lead to an abnormal distribution of megakaryocytes between osteoblastic and vascular niches and to altered megakaryopoiesis . Constitutive activation of integrin signaling in Rasa3−/− megakaryocytes is associated with alteration in actin cytoskeleton organization , including a lack of stress fiber assembly , in talin recruitment to the plasma membrane and in cell adherence and spreading that occurred independently of integrin ligands . These alterations probably prevent terminal differentiation of Rasa3−/− megakaryocytes since megakaryocyte αIIbβ3 and β1 integrins are known to control proplatelet production and platelet release [30]–[32] . Moreover , stress fiber assembly is known to require optimal β1 integrin activation , a process also regulated by αIIbβ3 integrin [33] , [34] . In future work , it will be important to analyze Rasa3−/− platelets , since integrins play also important roles in these cells . Our preliminary studies indicate that unstimulated Rasa3+/− platelets have altered adhesion to BSA-coated plates and activation , as compared with Rasa3+/+ platelets , thus mimicking defect of Rasa3−/− megakaryocytes . Interestingly , constitutive αIIbβ3 integrin activation in human megakaryocytes mimics most of the Rasa3−/− megakaryocyte phenotypical traits . Indeed , in rare thrombocytopenic patients with activating mutations in ITGA2B or ITGB3 genes , megakaryocyte spreading on fibrinogen is abnormal , with 50% of spread cells showing a disordered actin distribution where focal adhesion points are more evident than stress fibers [21] . Sustained and substrate-independent activation of the outside-in αIIbβ3 signaling was detected in megakaryocytes of these patients , leading to severely impaired proplatelet formation and congenital thrombocytopenia [20]–[22] . It is noteworthy here that these patients do not develop the entire Rasa3−/− phenotype , like megakaryocytosis and bone marrow fibrosis , and its consequences . This discrepancy suggests that Rasa3 has additional function beside the control of integrin signaling , and/or that the enzyme has slightly different roles in man and mouse . In a recent report , Peters and collaborators have described a new spontaneous mutant mouse with a missense mutation in the Rasa3 protein [18] . The G125V Rasa3Scat mutation causes mislocalization of the protein to the cytosol and phenotypical traits that are clinically and biologically most often different from SCID-Rasa3−/− and Rasa3−/− phenotypes ( Table S5 ) . Indeed , Rasa3Scat/Scat mice have a cyclic phenotype of crisis-remission with a first embryonic to P9 wave of lethality – which affect ∼60% of the mutant mice – followed by a second wave of lethality at P30 ( affecting 94% of the first crisis survivors ) . This unexplained cyclic phenotype is fully transferable via hematopoietic stem cells injection into SCID or RAG−/− mice , ruling out the possibility that expression of the mutant Rasa3Scat protein outside the hematopoietic system is responsible for the different phenotype [18] . Another notable difference between Rasa3Scat/Scat and SCID-Rasa3−/− mice is the presence of a delayed erythropoiesis in the former mice . By contrast , in SCID-Rasa3−/− mice , many hallmarks of regenerative anemia are present . It is noteworthy that no bone marrow fibrosis nor extramedullar hematopoiesis have been reported in the Rasa3Scat/Scat model , and no mechanism was presented to explain the severe Rasa3Scat/Scat thrombocytopenia . Finally , no evidence for predisposition to oncogenesis was observed in Rasa3Scat/Scat mice , but the very small numbers of homozygous mice that survive the second crisis period ( ∼6% of Scat/Scat newborns ) may explain this difference and preclude more extensive analysis . The cause of the major differences between the Rasa3Scat/Scat and SCID-Rasa3−/− phenotypes is currently not known , but may be due to the different mutation present in the Rasa3 protein - affecting protein localization and enzymatic activity , respectively - and/or to the different genetic background of the two models . Indeed , relocalization of the Rasa3Scat/Scat protein from the membrane to the cytosol may eventually create a new function in this cell compartment and lead to phenotypic alterations that are not present in mice expressing a catalytically-inactive and truncated Rasa3 protein . About 20% of SCID-Rasa3−/− mice develop a preleukemia characterized by a massive infiltration of bone marrow and spleen with CD117+ Sca-1+ CD38+ cells , a phenotype very similar to acute myeloid leukemia in man . The exact mechanism of this preleukemia was not defined in this work , but active GTP-bound Ras level was similar in Rasa3+/+ and Rasa3−/− fetal liver CD117+ hematopoietic stem cells . However , our studies in the human K562 leukemic cell line which overexpresses Rasa3 suggest that Rasa3 is a probable negative regulator of proliferation in these cells ( Fig . S5 ) . Alternatively , it has been reported that β1 and β3 integrin signaling regulates the balance among hematopoietic stem cell self-renewal , differentiation and quiescence in the osteoblastic niche [35] , [36] . Furthermore , β1 and β3 integrins can regulate stem cell functions via direct or indirect participation in cellular signaling [37] , providing a potential mechanism to explain the predisposition to preleukemia in a minor percentage of SCID-Rasa3−/− mice . In conclusion , our results demonstrate that mice with a catalytic inactivation of Rasa3 protein in the hematopoietic system develop a lethal syndrome characterized by defects during megakaryocyte development and differentiation , and leading to a severe thrombocytopenia . This syndrome is associated with Rap1 and integrin signaling alterations and a predisposition to develop preleukemia .
All animal studies were authorized by the Animal Care Use and Review Committee of the Université de Liège and of the Université Libre de Bruxelles . Rasa3−/− mice with Rasa3 exons 11 and 12 replaced by a neomycin resistance cassette express a catalytically-inactive Rasa3 truncated protein [16] . These mice were analyzed on a hybrid 129/SvJ×C57BL/6J genetic background . C . B . -17 SCID mice were purchased from Charles River , Belgium . All mice were bred in a specific pathogen free facility at the GIGA–Research Centre . The Rasa3 genotype was determined by PCR as previously described [16] . For reconstitution , 4–6 week-old C . B . -17 SCID mice were irradiated ( 200 rad ) and a total homogenate of E12 . 5 fetal liver cells ( FLC ) obtained from Rasa3 embryos was intravenously injected . SCID-Rasa3−/− mice were killed and analyzed either when moribund ( ie presenting a severely reduced mobility and/or feeding incompatible with a more than 2 days survival ) or 14 months after irradiation/reconstitution . Individual liver was recovered from E12 . 5 embryo and single cell suspension was prepared by passage through a 23-gauge needle . Recovered cells were cultured in DMEM ( Gibco ) supplemented with 10% heat-inactivated FBS , 2 mM L-Glutamine , 50 U/mL Penicillin , 50 ng/mL streptomycin , 0 . 1 mM nonessential amino acids and 50 ng/ml of recombinant mouse TPO for megakaryocyte differentiation ( PreProtech ) . Bone marrow from SCID-Rasa3+/+ and SCID-Rasa3−/− femurs were flushed with PBS . The marrow was cut in 1 mm transverse sections and placed in an incubation chamber containing complete DMEM medium . Chamber was maintained at 37°C for 6 h . Megakaryocytes at the periphery of the explant were observed under a confocal microscope ( Nikon A1R , 20× objective ) . Each experiment was performed in duplicates . One transversal section was used to determine by flow cytometry the number of CD41+ cells present in the explant . Images were acquired sequentially at 10 min intervals and processed with NIS-software and ImageJ . Three mice from each genotype were analyzed . FLC from Rasa3+/+ and Rasa3−/− embryos were cultured in the presence of TPO as described above . On day 3 , recovered cells were enriched for mature megakaryocytes on a 1 . 5–3% bovine serum albumin ( BSA ) gradient under gravity for 45 min at room temperature . The percentage of mature megakaryoctes in the enriched population was always over 70% . Cells were resuspended in Tyrode's buffer containing 1 mM CaCl2 and 1 mM MgCl2 for 3 h . For inside-out integrin signaling , cells were incubated for 30 min at room temperature with FITC-fibrinogen ( 250 µg/ml ) and 100 ng/ml TPO , 1 mM MnCl2 or nothing , in the presence or absence of 10 mM EDTA . After a 10-fold dilution with PBS containing 1 µg/ml propidium iodide , fibrinogen binding was quantified by flow cytometry [22] . Specific fibrinogen binding was defined as binding that was inhibited by 10 mM EDTA . To compare independent experiments , specific fibrinogen binding was expressed as a percent of maximal binding obtained in the presence of 1 mM MnCl2 , an activator of integrins . For outside-in integrin signaling , coverslides were coated with murine fibrinogen ( 100 µg/ml ) , collagen-I ( 35 µg/ml ) or Poly-D-Lysine ( PDL , 15 µg/ml ) for 1 h at room temperature , blocked with denatured BSA ( 5 mg/ml ) for 30 min and washed with PBS before use . Cells ( 25×103 ) were incubated for 18 h on the indicated substrate and non adherent cells were removed . Adherent cells were fixed in 10% formalin , permeabilized with 0 . 2% Triton X-100 in PBS and stained as described below . Cells were then analyzed by confocal microscopy and ImageJ Software . For Rap1 inhibitor studies , purified mature megakaryocytes were cultured over PDL coated-plates as in outside-in experiments in the presence of 3 µM GGTI-298 ( Sigma ) or DMSO as control . Adherent cells were fixed in 10% formalin , permeabilized with 0 . 2% Triton X-100 in PBS and stained as described above . Cells were then analyzed by confocal microscopy and ImageJ Software . A single-cell suspension of femur bone marrow was prepared by flushing the bones with PBS followed by gentle disaggregation through Pasteur pipette . Cells were released from spleen by gentle disruption with a piston of syringe . Spleen cells were treated with ACK buffer to lyse erythrocytes and washed once with PBS . Cells were incubated with 2 . 4G2 to saturate Fcγ receptors II and IIIa before staining with primary and secondary antibodies in PBS containing 0 . 1% FBS and 0 . 1% NaF for 20 min , and washed with the same solution before flow cytometric analysis on a FC 500 ( Beckman Coulter ) . Cell counts were determined by adding fluorospheres ( Flow-Count Fluorospheres , Beckman Coulter ) to the cell suspension , as described by the manufacturer . The following anti-mouse biotinylated or fluorochrome-conjugated antibodies were obtained from BD Pharmingen: anti-CD3ε , anti-CD71 , anti-CD41 and anti-CD117 . Anti-B220 , anti-F4/80 , anti-IgM , anti-Mac1 , anti-Sca-1 , anti-CD34 , anti-CD38 , anti-Ter119 , anti-CD41 and anti-Gr1 , as well as streptavidine-cychrome 5 were obtained from eBioscience . Anti-FcεRIa ( Mar-1 ) was obtained from O . Leo's laboratory ( Université Libre de Bruxelles , Belgium ) . JON/A antibody was obtained from Emfred Analytics . Fetal liver cell were analyzed on a FACS CantoII ( Beckman Coulter ) . For hematopoietic stem and megakaryocyte progenitor cells staining , anti-mouse biotinylated or fluorochrome-conjugated antibodies specific for Ter-119 , Gr1 , Mac1 , CD4 , CD8 , CD5 , IL7Rα , B220 and c-Kit ( CD117 ) were used to define the c-Kit+ Lin− cell population [19] , [38] . Then , anti-Sca-1 , anti-CD34 and anti-Flk2/Flt3 were used to define the hematopoietic stem cells , whereas anti-Sca-1 , anti-FcRγII/III and anti-CD150 were used to define the megakaryocytes progenitor cells ( all antibodies were from eBioscience , except anti-Flk2 , from BD Pharmingen and anti-CD150 , from BioLegend ) . Streptavidin phycoerythrin-Texas Red was from Invitrogen . Debris , aggregates and propidium iodide-positive dead cells were first excluded . Cells were analyzed using an LSRII flow cytometer ( Becton Dickinson ) . Data were analyzed with FlowJosoftware ( Tree Star , Ashland , OR ) . Fetal liver cells were stained for CD41 as described above and fixed with 5% formalin for 15 min . Cells were permeabilized in PBS containing 0 . 25% Tx-100 for 5 min at 4°C . DNA was stained with DAPI for 20 min and DNA content in CD41+ cells was determined by flow cytometry . Spleen and liver were fixed in paraformaldehyde 4% and embedded in paraffin following standard procedures . Femurs were fixed in paraformaldehyde 3 . 7% , decalcified in 0 . 5M EDTA pH 8 for one week and then processed as spleen and liver . Serially cut 5-µm-thick sections were stained with hematoxylin/eosin or Sirius Red ( for femur ) according to standard protocols . Spleen was processed as described and sections were stained with an anti-B220 antibody [39] . Femur sections were stained with a rabbit polyclonal anti-von Willebrand Factor ( vWF ) antibody from Dako . For quantification of megakaryocytes in osteoblastic and vascular niches , the whole diaphysis of three consecutive femur sections was scanned with a conventional microscope ( 20× objective ) for vWF+ cells , as described [40] . Megakaryocytes in the osteoblastic niche were calculated as the number of megakaryocytes in contact with the endosteal border . Megakaryocytes in the vascular niche were calculated as the number of megakaryocyte per vessel border . Osteoblastic and vascular borders were calculated with ImageJ software . Results are means ± SEM of 3 mice per genotype . Platelet counts were determined with Unopette ( Becton Dickinson ) . Red cells , total white cells , lymphocytes , neutrophils , eosinophils , basophils , hemoglobin , hematocrit and red cell volume were quantified with a Cell Dyn 3500 analyzer ( Abott Diagnostic ) . Serum erythropoietin and thrombopoietin levels were determined with ELISA mouse EPO and mouse TPO Quantikine kits ( R&D Systems ) . Blood smears were stained with Giemsa's , methylene blue and Romanowsky's solutions . Immunofluorescence studies using conventional and confocal microscopes were performed on total FLC cultured in the presence of TPO , on purified mature megakaryocytes and on FL hematopoietic stem cells . Cells were fixed in 5% formalin for 15 min , washed , permeabilized with 0 . 2% Tx-100 in PBS containing 2% of FBS for 15 min and incubated 1 h at room temperature with APC-conjugated anti-CD41 ( MW Reg30 , eBioscience ) for megakaryocyte or CD117 ( BD Pharmingen ) for HSC . Active , GTP-bound Rap1 or Ras immunofluorescence was detected using GST-RalGDS-RBD or GST-Raf1-RBD , respectively , and a FITC-conjugated mAb against GST ( Santa Cruz ) as described [41] Negative controls included the omission of GST-RalGDS-RBD/GST-Raf1-RBD , the substitution of GST-RalGDS-RBD/GST-Raf1-RBD with GST and the substitution of the anti-GST antibody with an irrelevant FITC-conjugated mouse IgG . After several washes , phalloidin-TRICT ( Sigma ) and DAPI ( Sigma ) were added for 20 min in PBS . After 3 washes in PBS , samples were mounted in ProLong ( Invitrogen ) for observation under a confocal microscope ( NikonA1R ) and/or an epifluorescence microscope ( Nikon Eclipse 90i ) . For active Rap1 or Ras images , z-sections of 0 . 150 microns were acquired from megakaryocytes or HSC . Pseudocolor scale was used to depicture the intensity of active Rap1 or Ras staining along the cell membrane . ImageJ was used to quantify the intensity of active Rap1 or Ras staining on each cell . All images were acquired and analyzed in the same conditions . For immunofluorescence studies of adherent megakaryocyte , cells were fixed with 10% formalin for 15 min , washed , permeabilized with 0 . 2% Tx-100 in PBS containing 2% of FBS for 15 min and incubated 1 h at room temperature with the indicated primary and secondary antibodies . After several washes , phalloidin-TRICT ( Sigma ) and DAPI ( Sigma ) were added for 20 min in PBS . After 3 washes in PBS , samples were mounted in ProLong ( Invitrogen ) for observation under a confocal microscope ( NikonA1R ) . The following antibodies were used: APC-conjugated anti-CD41 ( MW Reg30 , eBioscience ) , anti-Rap1 ( Millipore ) , anti-Talin-FITC and anti-rabbit-alexa 488 . A collagen-based system ( MegaCult-C , StemCell Technologies , Inc . ) was used for the colony assay . Briefly , 1 . 25×105 freshly isolated fetal liver cells were resuspended in IMDM completed with recombinant mouse TPO ( 50 ng/ml ) , IL-3 ( 20 ng/ml ) and IL-6 ( 10 ng/ml ) , followed by addition of cold collagen . Suspension was dispensed into 2 wells of a four chamber slide ( Millipore ) for duplicates . Cultures were kept at 37°C in a 5% CO2 atmosphere for 3 days . The collagen matrix was then fixed in a methanol–acetone solution ( 1∶3 ) , at room temperature for 20 min for colony fixation . Slides were then allowed to air dry for 15 min and stained for Acetylcholinesterase . For scoring , acetylcholinesterase-positive colonies with 3 or more immature megakaryocytes of about 10 µm of diameter were scored as CFU-Mk . Mature megakaryocytes averaged approximately 30 µm in diameter . Eight- to twelve-week old male mice were bled under sodium pentobarbital anesthesia from the retro-orbital plexus . Blood was collected on acid citrate dextrose ( ACD: 93 mM Na3-citrate , 7 mM citric acid , 14 mM dextrose , pH 6 . 0 ) containing 1 U/ml apyrase ( Grade I , Sigma ) in a volume ratio of ACD to blood of 1∶6 . Blood was centrifuged for 5 s at 800× g followed by 5 min at 100× g to obtain platelet rich plasma ( PRP ) . PRP was diluted 3 fold in ACD containing 1 U/ml apyrase and centrifuged at 1000× g . The platelet pellet was resuspended at a concentration of 3×108/ml in Tyrode's buffer ( 137 mM NaCl , 12 mM NaHCO3 , 2 mM KCl , 0 , 34 mM Na2HPO4 , 1 mM MgCl2 , 5 , 5 mM glucose , 5 mM Hepes , 0 . 35% BSA ) . In order to test the adhesion of unstimulated platelets to BSA-coated surface , 3 . 5×106 platelets in 300 µl of tyrode's buffer were added to each well of a 8 chambers slide ( Millipore ) and incubated for 45 min in a CO2 incubator at 37°C . Adherent platelets were washed twice with PBS , fixed with 10% formalin , and stained with phalloidin-TRICT . Washed platelets were stimulated or not with ADP ( 25 µM ) or collagen-related peptide ( CRP ) ( 1 µg/ml ) , under non-stirring conditions . After 15 minutes of activation , saturating concentrations of FITC-conjugated CD62 anti-P-selectin and PE-conjugated JON/A antibodies were added to the platelets , and incubations were continued for additional 15 minutes in the dark . Samples were fixed before the analysis with a FACS Calibur flow cytometer ( BD Biosciences ) . Light transmission was recorded during platelet aggregation induced by ADP ( 50 µM ) in the presence of 2 mM CaCl2 on a Chrono-Log Lumi-Aggregometer ( Havertown , PA ) . The Rasa3-tet-ON-inducible K562 cell line was generated by GEnTarget Inc . Briefly , Rasa3 expression and TetR repressor lentiviruses were generated and cotransduced in K562 cell by the company . K562 mutant cell line ( K562-Rasa3 ) was cultured in IMDM supplemented with 10% heat-inactivated FBS , 2 mM L-Glutamine , 50 U/mL Penicillin , 50 ng/mL streptomycin , 0 . 1 mM nonessential amino acids , 10 µg/ml blasticidin and 1 µg/ml puromycin . Treatment of K562-Rasa3 cells with tetracycline ( 2 µg/ml ) induced Rasa3 expression from the lentiviral constructs after 48 h . For the proliferation assay , 4×105 cells per ml were cultured in the absence or presence of tetracycline for 12 days . At the indicated days , number of alive cells was counted with a hemocytometer . Death cells were excluded by trypan blue staining . Rasa3 expression was confirmed by western blot . Two independent experiments were performed in duplicates-triplicates . Results are expressed as means ± SEM . Statistical analyses were performed with Graphpad Prism 3 . 0 . The test used for each experiment is described in the corresponding legend . For each test , a difference of P<0 . 05 was considered significant .
|
Megakaryocytes are the bone marrow cellular precursors of circulating blood platelets and give rise to nascent platelets by forming branching filaments called proplatelets . Terminal differentiation of round megakaryocytes into branched proplatelet forming megakaryocytes is a complex cytoskeletal-driven process which is affected in rare human familial thrombocytopenias . Interactions of megakaryocytes with extracellular matrix proteins are essential in this process since constitutive megakaryocyte integrin activity caused by specific mutations in ITGA2B or ITGB3 genes encoding for extracellular matrix protein receptors may result in abnormal adherent megakaryocytes , defect in proplatelet formation and thrombocytopenia . Here , we show that Rasa3 , a GTPase activating protein of the GAP1 family , controls Rap1 activation and integrin signaling during megakaryocyte differentiation . We found that Rasa3 catalytic inactivation in mice altered megakaryocyte development , adherence , migration , actin cytoskeleton organization and differentiation into proplatelet . Thus , these mice presented a severe thrombocytopenia , bleeding and anemia .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"cell",
"differentiation",
"gene",
"function",
"animal",
"models",
"developmental",
"biology",
"integrins",
"model",
"organisms",
"gtpase",
"signaling",
"molecular",
"development",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"thrombocytopenia",
"cell",
"adhesion",
"extracellular",
"matrix",
"ras",
"signaling",
"mouse",
"models",
"extracellular",
"matrix",
"adhesions",
"hematology",
"signal",
"transduction",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"molecular",
"cell",
"biology",
"cell",
"signaling",
"hematopoiesis"
] |
2014
|
Rasa3 Controls Megakaryocyte Rap1 Activation, Integrin Signaling and Differentiation into Proplatelet
|
In Caenorhabditis elegans and Drosophila melanogaster , removing the germline precursor cells increases lifespan . In worms , and possibly also in flies , this lifespan extension requires the presence of somatic reproductive tissues . How the somatic gonad signals other tissues to increase lifespan is not known . The lifespan increase triggered by loss of the germ cells is known to require sterol hormone signaling , as reducing the activity of the nuclear hormone receptor DAF-12 , or genes required for synthesis of the DAF-12 ligand dafachronic acid , prevents germline loss from extending lifespan . In addition to sterol signaling , the FOXO transcription factor DAF-16 is required to extend lifespan in animals that lack germ cells . DAF-12/NHR is known to assist with the nuclear accumulation of DAF-16/FOXO in these animals , yet we find that loss of DAF-12/NHR has little or no effect on the expression of at least some DAF-16/FOXO target genes . In this study , we show that the DAF-12-sterol signaling pathway has a second function to activate a distinct set of genes and extend lifespan in response to the somatic reproductive tissues . When germline-deficient animals lacking somatic reproductive tissues are given dafachronic acid , their expression of DAF-12/NHR-dependent target genes is restored and their lifespan is increased . Together , our findings indicate that in C . elegans lacking germ cells , the somatic reproductive tissues promote longevity via steroid hormone signaling to DAF-12 .
Aging and reproduction are two central aspects of an animal's life history . Although evolutionary theorists have long hypothesized that an intrinsic cost to reproduction may shorten lifespan [1] , many studies suggest that the relationship between the reproductive system and lifespan is more complex . Interestingly , studies in worms , flies , and mice have demonstrated that unknown signals emitted by the reproductive tissues can actively modulate lifespan [2]–[5] . Reproductive tissues are known to be important signaling centers . In humans , the reproductive tissues secrete a variety of hormones such as estrogens and testosterone , which have profound effects on development and behavior . However , little is known about how signals from the reproductive tissues can affect aging . In C . elegans , the somatic reproductive tissues ( also called the “somatic gonad” ) can have a dramatic lifespan-extending effect . When the germ cells of the worm are removed , the resulting sterile adult animals live 60% longer than they would with an intact germline . Removal of the somatic reproductive tissues along with the germ cells suppresses this lifespan extension , suggesting that the somatic gonad dispatches lifespan-extending signals to other tissues when the germline is gone [2] . How does the reproductive system influence other tissues ? In order for animals lacking the germline to live long , they require a signaling pathway involving the nuclear hormone receptor DAF-12 , as well as genes that produce DAF-12 ligands ( called “dafachronic acids” ) , such as the cytochrome P450 gene daf-9 [2] , [6]–[8] . Thus , a steroid signaling pathway is embedded within this longevity system . This steroidal signaling pathway is a prime candidate for a pathway that allows the reproductive tissues to communicate with the rest of the animal . The analysis of this longevity pathway has been facilitated by the identification of molecular events that occur in the body when the germ cells are removed . A particularly important event caused by loss of the germ cells is the nuclear accumulation of the evolutionarily conserved lifespan-extending transcription factor DAF-16/FOXO in intestinal cells [9] . [The intestine is the major endodermal tissue in C . elegans , as it also performs functions characteristic of the adipose tissue , the liver , and the pancreas . ] DAF-16/FOXO is completely required for loss of the germ cells to increase lifespan , and expression of daf-16/FOXO exclusively in the intestine completely rescues the longevity of daf-16 ( − ) mutants lacking a germline [10] . Previously , we demonstrated that daf-12/NHR and daf-9/CYP450 are partially required for DAF-16/FOXO to accumulate in intestinal nuclei when the germ cells are removed [11] . Furthermore , treatment of ligand-defective daf-9/CYP450 mutants lacking germ cells with the DAF-12/NHR ligand Δ4-dafachronic acid rescues DAF-16/FOXO nuclear localization [12] . Together , these findings indicate that DAF-9/CYP450 and DAF-12/NHR play a role in the nuclear localization of DAF-16/FOXO . However , interestingly , daf-12/NHR is still required for lifespan extension in animals carrying a mutant DAF-16/FOXO protein that localizes constitutively to nuclei [11] . Thus , daf-12/NHR has another function , apart from regulation of DAF-16/FOXO nuclear localization , in the regulation of longevity by the reproductive system . What is the other function of DAF-12/NHR ? In this study we have asked whether daf-12/NHR might function in the signaling that takes place between the somatic reproductive tissues and the rest of the animal . Like DAF-12/NHR , the somatic gonad has a lifespan-extending function that does not involve DAF-16/FOXO nuclear localization . Germline-deficient animals that lack the somatic gonad do not live long even though DAF-16/FOXO accumulates in nuclei [13] . Here , we present data suggesting that this essential life-extending function of the somatic gonad is its ability to activate the DAF-12/dafachronic-acid signaling pathway .
Worms that lack germ cells [germ cell ( − ) ] have an extended lifespan that requires both the DAF-12 sterol-signaling pathway and the somatic gonad [2] . Therefore , we hypothesized that the somatic gonad might extend the lifespan of animals lacking germ cells by promoting the activity of DAF-12/NHR . In animals lacking germ cells , dafachronic acids are likely to stimulate DAF-12/NHR's lifespan-extending activity because the extended lifespan produced by germ-cell removal requires DAF-9/CYP450 , which catalyzes the synthesis of dafachronic acids . If the somatic gonad extends lifespan by promoting dafachronic-acid signaling , then it should be possible to bypass the requirement for the somatic gonad by providing germline-deficient animals with exogenous dafachronic acid . To test this hypothesis , we supplemented the media of animals lacking both the somatic gonad and the germline [germ cell ( − ) ; somatic gonad ( − ) ] with Δ4-dafachronic acid , one of several isoforms of dafachronic acid hormones shown to bind DAF-12/NHR [8] . To generate adults that lack both the somatic gonad and germ cells , we ablated the two somatic gonad precursor cells ( Z1 and Z4 ) in L1 larvae . The development of the germ cells requires the presence of a somatic gonad; therefore , killing the somatic gonad precursors eliminates the entire reproductive system . In three separate trials , we found that Δ4-dafachronic acid was able to increase the lifespan of germline-deficient animals that also lacked the somatic gonad ( Figure 1A , Table S1 ) . In contrast , we saw little to no extension of lifespan when daf-12 ( rh61rh411 ) null mutants that lack the somatic gonad and germ cells were treated with Δ4-dafachronic acid ( Figure 1B , Table S1 ) . Thus , as expected , Δ4-dafachronic acid exerts its effects through the DAF-12 nuclear hormone receptor . These findings are consistent with the hypothesis that the somatic gonad exerts its effect on lifespan by activating the DAF-12/dafachronic acid pathway . However , these data do not exclude the possibility that dafachronic acid extends lifespan through a parallel pathway that is not related to the reproductive system . Two experiments argue that this is not the case . First , if dafachronic acid extends lifespan via a pathway unrelated to the reproductive longevity system , then it would be expected to further extend the long lifespan of animals lacking germ cells . On the contrary , we found that Δ4-dafachronic acid did not further extend the lifespan of germline-deficient ( Z2/Z3-ablated ) animals that contained the somatic reproductive tissues . In two out of three experiments , we saw no effect of Δ4-dafachronic acid supplementation ( Figure 1C ) , as previously reported by Gerisch et al . [12] . In one of the three experiments , we observed a shortening of lifespan ( Table S1 ) . The fact that the effects of dafachronic acid and germline removal are not additive suggests that dafachronic acid is part of the reproductive longevity pathway . Moreover , it suggests that dafachronic acid is not a limiting factor for lifespan extension in germline-deficient animals . In addition , as predicted by the model that dafachronic acid can substitute for the somatic gonad in animals lacking a germline , the lifespan of germ cell ( − ) ; somatic gonad ( − ) animals treated with dafachronic acid was as long as that of germ cell ( − ) animals treated with dafachronic acid ( Table S1 ) . Second , if dafachronic acid extends lifespan via a pathway that is not related to the reproductive system , then it should extend the lifespan of normal , intact animals as well as that of animals that lack the reproductive system . However , Gerisch et al . reported previously that dafachronic acid does not increase the lifespan of intact animals . We repeated these experiments , measuring the lifespan of animals with an intact gonad maintained on Δ4-dafachronic acid plates , and also observed no change in lifespan ( Figure 1D , Table S1 ) . Thus , dafachronic acid only extends the lifespan of animals that lack both the germ cells and the somatic reproductive tissues . These results link the lifespan-extending effect of dafachronic acid to this reproductive signaling pathway in an interesting way: they suggest that loss of the germline is required for the somatic gonad to promote lifespan extension through the DAF-12/dafachronic acid pathway . Unexpectedly , in the dafachronic-acid supplementation experiments , loss of the entire reproductive system in the absence of dafachronic acid shortened lifespan . We do not know what caused this effect in these experiments , however we note that the Antebi lab also observed a small but statistically significant shortening of lifespan when the somatic gonad and germ cells were removed [6] , [7] . Nonetheless , dafachronic acid was able to overcome this lifespan shortening , as it caused animals lacking the entire reproductive system to live as long as long-lived germline-deficient animals grown in parallel under the same conditions . In summary , supplementation of exogenous Δ4-dafachronic acid increased the lifespan of animals lacking the entire reproductive system in a daf-12-dependent manner , but it did not increase the lifespan of animals lacking only the germline or the lifespan of animals with an intact gonad . Together , these findings are consistent with the idea that the somatic gonad transmits longevity signals to the rest of the animal through dafachronic acid signaling . Another way to investigate whether the dafachronic acid signaling pathway mediates the effect of the somatic gonad on lifespan was to ask whether overexpression of DAF-9/CYP450 , which is predicted to increase the amount of endogenously generated dafachronic acids , extends the lifespan of germline-deficient animals lacking the somatic gonad . We removed the somatic gonad and germ cells of animals that carry a transgenic array containing multiple copies of a functional daf-9::GFP gene fusion driven by the daf-9/CYP450 promoter . We found that in animals that overexpress DAF-9/CYP450 , removal of the somatic gonad and germ cells extended lifespan ( Figure 2A , Table S2 ) , in contrast to the case in wild-type animals , where removal of the somatic gonad as well as the germ cells does not increase lifespan [2] . The daf-9/CYP450 promoter drives daf-9/CYP450 expression in parts of the somatic gonad as well as in some non-reproductive tissues [6] , [14] , [15] . Thus , one might hypothesize that DAF-9/CYP450 acts in the somatic gonad to promote longevity of germ-cell deficient animals . However , as overexpression of DAF-9/CYP450 using its own promoter increased the lifespan of animals lacking the entire reproductive system , DAF-9/CYP450 can clearly function in tissues outside of the somatic gonad to promote longevity . Besides being expressed in parts of the somatic gonad , daf-9/CYP450 is expressed in the neuroendocrine-like XXX cells and the hypodermis [6] , [14] , [15] . Thus , we asked whether limiting DAF-9/CYP450 overexpression to just one of these tissues would be sufficient to restore the lifespan extension of germline-deficient animals lacking the somatic reproductive tissues . First , we used the XXX-cell-specific sdf-9 promoter to express daf-9/CYP450 in a hypomorphic daf-9 ( e1406 ) mutant . We found that this construct was able to extend the lifespan of germ cell ( − ) ; somatic gonad ( − ) animals ( Figure 2B , Table S2 ) . Likewise , expressing DAF-9/CYP450 using the hypodermal dpy-7 promoter , which is active during larval development , also extended the lifespan of animals lacking a reproductive system ( Figure 2C , Table S2 ) . Thus , DAF-9/CYP450 can function both in the hypodermis and XXX cells to extend lifespan of germ cell ( − ) ; somatic gonad ( − ) animals . We also examined whether DAF-9/CYP450 could extend lifespan when expressed in cells that do not normally express daf-9/CYP450 . We found that this was the case: expression of DAF-9/CYP450 in ciliated sensory neurons using the cilium-specific che-2 promoter also extended the lifespan of germ cell ( − ) ; somatic gonad ( − ) animals ( Figure 2D , Table S2 ) . This finding implies that both the substrates of DAF-9/CYP450 and its products , the dafachronic acids , can travel between the various tissues in germ cell ( − ) animals . Just as with addition of exogenous Δ4-dafachronic acid , increasing dafachronic acid levels by overexpression of DAF-9/CYP450 had a greater effect on the lifespan of germ cell ( − ) ; somatic gonad ( − ) animals than it had on animals with intact gonads: germ cell ( − ) ; somatic gonad ( − ) animals lived longer than did intact animals in several strains carrying multi-copy daf-9/CYP450 transgene arrays . These findings , too , imply that dafachronic acid extends lifespan specifically in germ cell ( − ) ; somatic gonad ( − ) animals . Reduction-of-function daf-9 alleles and the canonical daf-12 ( m20 ) allele both completely prevent loss of the germ cells from extending lifespan [2] , [6] . These findings led to the model that dafachronic acid extends the lifespan of germline-deficient animals simply by activating a lifespan extending activity of DAF-12/NHR . The m20 allele is predicted to eliminate the function of two isoforms of DAF-12/NHR while leaving a third isoform intact [16] , [17] . Therefore , it was interesting to ask how animals carrying a putative null DAF-12/NHR allele would respond to germline ablation . The double daf-12/NHR mutant rh61rh411 is predicted to inactivate all DAF-12/NHR isoforms [17] . Consistent with this interpretation , mutants carrying this allele appear phenotypically to lack the developmental functions of DAF-12/NHR [17] . We removed the germ cells of daf-12 ( rh61rh411 ) mutants and found , to our surprise , that they lived slightly longer than intact controls ( Figure S1 , Table S3 ) . A similar small lifespan extension was reported in similar experiments carried out by the Antebi lab [12] . Consistent with the hypothesis that the somatic gonad exerts its lifespan-extending function through DAF-12/NHR , we found that this small lifespan extension was somatic-gonad independent ( Figure S1 , Table S3 ) . Likewise , dafachronic acid had little or no effect on the lifespans of these animals ( Figure 1B , Table S1 ) . Because no lifespan increase was observed when the germline or entire gonad of daf-12 ( m20 ) animals was removed [2] , it appears that a second , lifespan shortening function of DAF-12/NHR is revealed by the rh61rh411 allele . Together , these findings necessitate a revised model for the role of DAF-12/NHR in animals that lack germ cells ( see Discussion ) . To assess DAF-12/NHR activity at the level of downstream gene expression , we examined expression of genes directly or indirectly regulated by DAF-12/NHR in the presence and absence of the somatic gonad . cdr-6 , which encodes a homolog of a C . elegans cadmium-responsive gene , was identified in microarray experiments searching for genes regulated by daf-12/NHR in animals that lack germ cells ( MM and CK , in revision ) . We found that expression of GFP under the control of the cdr-6 promoter ( Pcdr-6::GFP ) was up-regulated in the animal by germ cell ablation ( Figure 3 , Table S4 ) . This cdr-6 up-regulation required daf-12/NHR , as no increase in expression was seen when the germ cells were removed in a daf-12 ( rh61rh411 ) mutant ( Figure 3B ) . We verified the daf-12/NHR dependent increase in cdr-6 expression in germ cell ( − ) animals using quantitative RT-PCR ( qPCR ) ( Figure 4C , Table S9 ) . If signals from the somatic gonad cause DAF-12/NHR to up-regulate cdr-6 expression upon germline removal , then ( i ) the presence of the somatic gonad should be required for loss of the germline to increase Pcdr-6::GFP expression and ( ii ) dafachronic acid should be able to substitute for the presence of the somatic gonad in the regulation of this gene . We found that both predictions were met . When we removed the somatic gonad as well as the germ cells , Pcdr-6::GFP expression was no longer elevated; instead , it was similar to the level observed in animals with an intact gonad ( Figure 3 , Table S4 ) . Moreover , when these germ cell ( − ) ; somatic-gonad ( − ) animals were grown on plates containing Δ4-dafachronic acid , expression of Pcdr-6::GFP was restored to levels that were similar to those of Pcdr-6::GFP-transgenic animals lacking only the germline . As predicted , the increase in Pcdr-6::GFP expression caused by addition of dafachronic acid required a functional DAF-12/NHR protein . In a daf-12 ( rh61rh411 ) mutant , we observed no increase in cdr-6 expression upon Δ4-dafachronic acid treatment ( Figure 3B ) . Together , these findings support the interpretation that in germline-deficient animals , the somatic gonad activates a dafachronic-acid signaling pathway that turns on cdr-6 gene expression through the activity of DAF-12/NHR . Interestingly , Δ4-dafachronic acid increased the expression of Pcdr-6::GFP in animals with an intact gonad ( Figure 3B ) . This was noteworthy , as dafachronic acid does not increase the lifespan of animals with an intact gonad . This finding is consistent with the idea that loss of the germline has two effects . First , germline loss permits the somatic gonad to activate DAF-12/NHR via dafachronic acid signaling . Second , germline loss initiates additional events that are required for lifespan extension ( see Discussion ) . These findings were supported by our analysis of a second gene , dod-24 , which we found to be negatively regulated by daf-12/NHR in a different study ( MMG , SJL , and CK , in preparation ) . The expression of a transcriptional Pdod-24::GFP gene fusion was variable; however , we found that dod-24 expression depended on the presence of the somatic gonad in a daf-12/NHR dependent fashion ( Figure S2 , Table S5 ) , similar to results we obtained with cdr-6 . Since DAF-12/NHR appeared to be regulated by the somatic gonad , we wondered if DAF-12/NHR activity could be responsible for all the effects of the somatic gonad . Previously we showed that the somatic gonad is required for the increased expression of a subset of daf-16/FOXO-regulated genes in germ-cell deficient animals [13] . We wondered if daf-12/NHR was also required for the expression of these genes . We examined the expression of sod-3 , a direct target of DAF-16/FOXO [18]–[20] . When the germ cells are removed , sod-3 expression increases , and this increased expression requires daf-16/FOXO [13] . We found that this increased expression was largely independent of daf-12/NHR: in a daf-12 ( rh61rh411 ) mutant background , expression of a Psod-3::GFP fusion still increased upon germ cell removal . Furthermore , this increase in expression still required the somatic gonad ( Figure 4A , Table S6 ) . We observed a similar effect when we examined mRNA levels of sod-3 by quantitative RT-PCR ( qPCR ) in glp-1 ( e2141 ) mutants , which lack germ cells and are long-lived [5] , [21] . Whereas mutation of daf-16/FOXO resulted in a significant drop in sod-3 expression , mutation of daf-12/NHR did not have a statistically significant effect ( Figure 4C , Table S9 ) . In intact animals , we also observed a decrease of sod-3 expression by qPCR upon mutation of daf-16/FOXO , but not when daf-12/NHR was mutated ( Table S9 ) , consistent with sod-3's being a daf-16/FOXO regulated gene . We note that , whereas in several experiments sod-3::GFP up-regulation in germ-cell ablated animals was daf-12/NHR-independent ( Figure 4A ) , in one experiment , we did observe a slight decrease in the expression that was statistically significant ( Table S6 ) . Thus , DAF-12/NHR could potentially make a small , variable , contribution to sod-3 expression . As DAF-12/NHR had at best a modest effect on the expression of the DAF-16/FOXO-regulated sod-3 gene , we wondered about the converse situation—that is , whether DAF-16/FOXO might have a minor effect on the activity of DAF-12/NHR-regulated genes . We therefore examined the expression of genes whose expression requires daf-12/NHR in a daf-16/FOXO mutant background . First , we examined transgenic animals carrying the Pcdr-6::GFP construct in a daf-16 ( mu86 ) null mutant and observed that expression of Pcdr-6::GFP still increased relative to intact animals in response to loss of the germline ( Figure 4B , Table S7 ) . Furthermore , removal of the somatic gonad in germ cell ( − ) animals lowered the expression of Pcdr-6::GFP . Thus , the somatic gonad was still able to modulate expression of cdr-6 when DAF-16/FOXO was not functioning . Notably , the magnitude of up-regulation of Pcdr-6::GFP in the daf-16 ( mu86 ) mutant was lower than in a wild-type background . We observed similar expression patterns when we measured cdr-6 mRNA levels by qPCR in glp-1 ( e2141 ) mutants , which lack germ cells . Mutation of daf-12/NHR resulted in a significant drop in the level of cdr-6 mRNA , whereas mutation of daf-16/FOXO did not affect cdr-6 mRNA levels ( Figure 4C , Table S9 ) . Together , these findings suggest that daf-16/FOXO is at most only partially required for loss of the germline to increase the expression of cdr-6 . We also observed daf-16/FOXO-indepdendent modulation of dod-24 expression by the presence of the somatic gonad ( Figure S2 , Table S8 ) . In summary , mutation of daf-16/FOXO only slightly affects the transcription of the daf-12/NHR regulated genes cdr-6 and dod-24 , whereas mutation of daf-12/NHR only slightly affects the transcription of the direct daf-16/FOXO target , sod-3 . Thus , while it remains possible that DAF-16/FOXO and DAF-12/NHR co-regulate yet-unidentified target genes , these two transcription factors regulate at least some of their target genes largely , but not completely , independently of one another . As described above , dafachronic acid extends the lifespan of animals that lack the entire reproductive system . In contrast , dafachronic acid has no effect on the lifespan of intact animals . This finding suggests that the presence of the germline prevents dafachronic acid from extending lifespan . Loss of the germline activates the expression of multiple genes in a daf-16/FOXO-dependent fashion [13] , [22] , and , as described above , at least some of these genes are activated largely independently of daf-12/NHR . It seemed possible that these daf-16/FOXO-dependent genes are collectively required for loss of the germline to extend lifespan . If many of these genes cannot be activated by DAF-12/NHR , then the addition of dafachronic acid would not be expected to extend lifespan in the presence of the germline . We tested this idea by removing the germ cells and the somatic gonad of daf-16/FOXO mutants , and then treating these animals with dafachronic acid . If DAF-16/FOXO-dependent genes are required for lifespan extension , then these animals should not live long . We found this was the case . Whereas in a wild-type , daf-16 ( + ) , background , Δ4-dafachronic acid extended the lifespan of germ-cell ( − ) ; somatic gonad ( − ) animals , in a daf-16 ( mu86 ) background , there was no change in lifespan ( Figure 4D ) . Thus , daf-16/FOXO is still required for lifespan extension in animals with activated DAF-12/NHR . Together , these findings , along with our studies of germline-dependent gene expression , suggest that although there is some overlap , DAF-16/FOXO has an essential function in this lifespan extension pathway that is triggered mainly by germline loss , and DAF-12/NHR has another , distinct function that is activated by the somatic gonad when the germline is removed ( Figure 4E ) .
The primary finding of this study is that the somatic gonad extends the lifespan of germline-deficient animals by activating a DAF-12/NHR-dependent sterol-signaling pathway . The somatic gonad and genes that produce dafachronic acid are both required for germ cell removal to extend lifespan , and their ability to influence lifespan requires daf-12/NHR . Furthermore , the somatic gonad is important for the proper activity of DAF-12/NHR , as the presence of the somatic gonad is required for the correct expression of daf-12/NHR-regulated genes such as cdr-6 . Most compelling , increasing the level of the DAF-12/NHR ligand Δ4-dafachronic acid in animals that lack the somatic gonad is sufficient to rescue both lifespan extension and expression of daf-12/NHR-regulated genes , in a fashion that requires daf-12/NHR . One simple model to explain the longevity-promoting activity of the somatic gonad is that the somatic gonad stimulates dafachronic acid production when the germ cells are removed , which in turn affects DAF-12/NHR activity . Indeed , this model is not without precedent , as in humans the somatic reproductive tissues secrete steroid hormones such as androgens , which influence other tissues . In this model , germ cell ( − ) ; somatic gonad ( − ) animals fail to live long because they have insufficient dafachronic acid levels . We asked whether the somatic gonad might influence the level of daf-9/CYP450 gene expression , but this was not the case , as levels of DAF-9::GFP were not overtly different ( unpublished data ) . However , the somatic gonad could potentially affect the level of dafachronic acid by alternate mechanisms; for example , by increasing the level of a biosynthetic precursor of dafachronic acid . It is also possible that the somatic gonad regulates the activity of DAF-12/NHR without affecting the total level of dafachronic acid . For example , the somatic gonad could influence the proportion of dafachronic acid in the animal that is available to bind to DAF-12/NHR . Another possibility is that the somatic gonad influences the levels or activities of DAF-12/NHR inhibitors or co-activators , though in this scenario , it is necessary to postulate that increased levels of dafachronic acid can overcome the effects of these co-regulators . It would be interesting to directly measure levels of dafachronic acids in animals that lack the germ cells or the entire reproductive system . daf-9/CYP450 is expressed in the somatic gonad ( specifically , in the spermatheca [6] , [15] ) , so it was interesting to find that limiting DAF-9/CYP450 overexpression to one of several non-reproductive tissues—the XXX cells , the hypodermis , or even to sensory neurons that do not normally express daf-9/CYP450—could increase the longevity of germ-cell defective animals that lack the somatic gonad . This finding indicates that tissues other than the somatic gonad and the intestine can participate in this reproductive signaling pathway by producing dafachronic acid . Perhaps loss of the germ cells stimulates the synthesis or release of a precursor of dafachronic acid , which in turn diffuses among the tissues . Alternatively , as mentioned above , various tissues could synthesize dafachronic acid independently of any input from the reproductive system if the somatic gonad controls other factors in the dafachronic acid/DAF-12 signaling pathway . Finally , it is possible that when the germline is removed but the somatic gonad is present , gonadal DAF-9/CYP450 also contributes to the pool of dafachronic acid in the animal . Indeed , when daf-9/CYP450 was overexpressed under the control of its endogenous promoter , which drives some expression in the somatic gonad , germline ablation caused a larger increase in lifespan than did loss of the entire gonad ( Table S2 ) . This was not the case when daf-9/CYP450 was expressed only in non-gonadal tissues . This difference could be due to the loss of DAF-9/CYP450 in the somatic gonad . Besides modulating the lifespan of germ cell ( − ) animals , daf-9/CYP450 , dafachronic acid , and daf-12/NHR play a second role in facilitating the nuclear localization of DAF-16/FOXO [11] , [12] . Interestingly , the somatic gonad does not appear to modulate this second function of DAF-12/NHR , as removal of the somatic gonad does not affect the nuclear localization of DAF-16/FOXO in germ cell ( − ) animals [13] . It is possible that different levels of dafachronic acid are required for these two activities of DAF-12/NHR , and that residual levels of dafachronic acid are sufficient to promote DAF-16/FOXO localization when both the germ cells and somatic gonad are gone . In any case , it will be interesting to determine how the somatic gonad modulates one aspect of DAF-12/NHR function , its ability to promote longevity , without affecting the other aspect , DAF-16/FOXO nuclear localization . Giving dafachronic acids to animals with an intact gonad does not extend lifespan ( Figure 1D and [12] ) . However , dafachronic acid does stimulate DAF-12/NHR to regulate germline-dependent genes in intact animals , since it produces a daf-12/NHR-dependent up-regulation of cdr-6 and a daf-12/NHR-dependent down-regulation of dod-24 in intact animals . Since activation of DAF-12/NHR is not sufficient to extend lifespan , other lifespan-promoting factors turned on in germline-deficient animals must be necessary for an increased lifespan . Because dafachronic acid extends lifespan in the absence of the somatic gonad , the somatic gonad itself is unlikely to provide these other lifespan-promoting factors . Instead , DAF-16/FOXO is the most likely candidate for the factor activated by loss of the germ cells that is necessary , along with DAF-12/NHR and the somatic gonad , to increase lifespan . Consistent with this idea , genetic inactivation of daf-16/FOXO , like the presence of the germ cells , prevents dafachronic acid from extending lifespan . An unexpected finding from this study was that DAF-12/NHR has a more complex role in this longevity pathway than previously appreciated . We found that when the germ cells are removed in animals containing a daf-12/NHR null mutation , lifespan is extended slightly . This daf-12-independent lifespan-extension pathway ( referred to here as the “underlying pathway” ) does not require the somatic gonad and is not affected by dafachronic acid . When daf-12 ( + ) activity is present but the somatic gonad is absent or daf-9/CYP450 is mutated , then germ-cell loss does not extend lifespan . This suggests that unliganded DAF-12/NHR prevents germ-cell loss from activating the underlying pathway . In contrast , liganded DAF-12/NHR extends lifespan in response to germ-cell loss , as daf-12 ( + ) animals that have a daf-9 ( + ) genotype plus the somatic gonad live long when the germ cells are removed . In the future , it will be interesting to explore the nature of the underlying daf-12-independent pathway and to learn how daf-12/NHR ( + ) can affect its activity . Finally , we note that a dual ability of DAF-12/NHR to extend and shorten lifespan is not without precedent . In intact animals , DAF-12/NHR extends lifespan in daf-9/CYP450 reduction-of-function mutants when animals are cultured at 15°C [6] , [14] . In contrast , at warmer temperature ( 25°C ) , DAF-12/NHR shortens lifespan in response to decreased daf-9/CYP450 in the absence of thermosensory neurons [23] . Both sterol signaling and DAF-16/FOXO are required for the long lifespan of germline-deficient animals . The relationship between DAF-12/NHR and DAF-16/FOXO in animals that lack the germ cells is not well understood . Previous work has demonstrated that DAF-12/NHR is partially ( but not completely ) required for the nuclear accumulation of DAF-16/FOXO in animals that lack the germ cells [11] , [12] . Consistent with this result , in this study we showed that in daf-12/NHR mutants , the DAF-16/FOXO target sod-3 is still up-regulated ( though perhaps to a lesser extent ) . Likewise , Wang and Ruvkun showed that the lipase gene K04A8 . 5 is up-regulated by germline removal in a daf-16/FOXO-dependent but daf-12/NHR-independent fashion [24] . These data suggest that DAF-16/FOXO can promote the transcription of at least some of its target genes independently of DAF-12/NHR in animals that lack the germ cells . We have found that the converse also holds true . When daf-16/FOXO is mutated , DAF-12/NHR still retains the ability to affect transcription of genes such as cdr-6 and dod-24 . However , DAF-16/FOXO affects the magnitude of this regulation , suggesting that DAF-16/FOXO could have a partial effect on the activity of DAF-12/NHR . In summary , based on the several genes we examined , it appears that DAF-16/FOXO and DAF-12/NHR have distinct effects on the transcriptome of germ-cell deficient animals , although each has minor effects on the activity of the other . This interpretation is supported by a genome-wide microarray analysis of germline-defective daf-16/FOXO and daf-12/NHR mutants ( MM and CK , in revision ) . Although dafachronic-acid signaling and DAF-16/FOXO have distinct effects on gene transcription in animals that lack germ cells , both are required to extend lifespan . Furthermore , dafachronic acid does not override the requirement for DAF-16/FOXO to extend longevity , and rendering DAF-16/FOXO constitutively nuclear does not override the requirement for DAF-12/NHR . These two pieces of data make it unlikely that DAF-12/NHR and DAF-16/FOXO operate in a simple linear pathway , where the transcriptional effects of mutation of one gene would be completely mimicked by the mutation of the other . Instead , the simplest interpretation is that DAF-12/NHR and DAF-16/FOXO function in parallel to promote longevity in animals without germ cells . Therefore , we propose the following model ( Figure 4E ) : germ-cell removal has two important effects: ( i ) DAF-16/FOXO accumulates in the nucleus , and ( ii ) DAF-12/NHR is independently stimulated to extend lifespan . In these germline-deficient animals , activated DAF-12/NHR and DAF-16/FOXO act in parallel on different target genes ( for the most part ) to promote lifespan extension . The presence of the somatic gonad in germ-cell deficient animals promotes the activation of DAF-12/NHR by ensuring sufficient levels of available dafachronic acids , possibly through an increase in their levels . When the somatic gonad is removed , DAF-12/NHR no longer extends lifespan , and the animals no longer live long .
All strains used in this study were maintained under standard conditions [25] . The following strains were used: N2 CF2479 daf-12 ( rh61rh411 ) daf-9 ( e1406 ) ; mgEx662[daf-9p::daf-9 genomic::GFP] daf-9 ( e1406 ) ; mgEx670[sdf-9p::daf-9 cDNA::GFP; mec-7::GFP] daf-9 ( e1406 ) ; mgEx663[dpy-7p::daf-9 cDNA::GFP; mec-7::GFP] daf-9 ( e1406 ) ; mgEx666[che-2p::daf-9 cDNA::GFP; mec 7::GFP] daf-9 ( e1406 ) ; mgEx668[col-12p::daf-9 cDNA::GFP; mec 7::GFP] BC15369 dpy-5 ( e907 ) ; sEx15369[Pcdr-6::GFP + pCeh361] CF3595 sEx15369[Pcdr-6::GFP + pCeh361] obtained by outcrossing BC15369 3 times to the laboratory N2 CF3596 daf-12 ( rh61rh411 ) ; sEx15369[Pcdr-6::GFP + pCeh361] AU68 agIs6[Pdod-24::GFP] CF3556 agIs6[Pdod-24::GFP] obtained by outcrossing AU86 3 times to the laboratory N2 CF3600 daf-12 ( rh61rh411 ) ; agIs6[Pdod-24::GFP] CF3601 daf-16 ( mu86 ) ; agIs6[Pdod-24::GFP] CF1553 muIs84[Psod-3::GFP] CF3604 daf-12 ( rh61rh411 ) ; muIs84[Psod-3::GFP] CF3597 daf-16 ( mu86 ) ; sEx15369[Pcdr-6::GFP + pCeh361] CF1903 glp-1 ( e2141 ) CF1880 daf-16 ( mu86 ) ; glp-1 ( e2141 ) CF1658 glp-1 ( e2141 ) ; daf-12 ( rh61rh411 ) CF1037 daf-16 ( mu86 ) Some nematode strains used in this study were provided by the Caenorhabditis Genetics Center , which is funded by the NIH National Center for Research Resources ( NCRR ) . Construction of Psod-3::GFP was described previously in [10] . daf-9::GFP strains were provided by the Ruvkun Lab and were described previously in [15] . The Pdod-24::GFP strain was kindly provided by D . Kim . Strains containing Pcdr-6::GFP were obtained from the Genome British Columbia C . elegans Gene Expression Consortium [26] . Germ-cell ( Z2 , Z3 ) or somatic-gonad ( Z1 , Z4 ) precursor cells of newly hatched L1 larvae were killed by laser ablation as described previously [2] using a VSL-337 nitrogen pumped dye laser ( Laser Sciences , Inc . ) . At adulthood , absence of the gonad or germ cells was confirmed using a dissecting microscope . To obtain intact-gonad controls , un-ablated L1 larvae were anaesthetized and recovered from the same NaN3 agarose pads as experimental animals . Lifespan analysis was performed at 20°C as described previously [27] , [28] using OP50 bacteria . Lifespan analyses of animals grown on dafachronic acid were performed using 3 cm plates containing 5 mL of NG agarose media . Prior to use , 1 µl of 1 mM dafachronic acid in ethanol was diluted in 100 µl PBS and pipetted onto a plate containing a lawn of OP50 bacteria . As a control , 3 cm plates spotted with 1 µl of ethanol diluted in 100 µl PBS were used . Animals were placed on dafachronic acid or control plates as L1 larvae directly after laser ablation . Statistical analysis was performed using Stata/IC 10 . 0 software ( StataCorp LP ) . p values were determined using the log-rank ( Mantel-Cox ) method . On day 2 of adulthood , animals were anaesthetized on agarose pads containing 0 . 15 M NaN3 . Images were taken using a Retiga EXi Fast1394 CCD digital camera ( QImaging ) using the 10× objective on a Zeiss Axioplan 2 compound microscope ( Zeiss Corporation ) . Each image was taken with the intestine in focus , since expression of the various transgenes was primarily in the intestine . For each trial , exposure time was calibrated to minimize the number of saturated pixels for that set of animals . Openlab 4 . 0 . 2 software ( Improvision ) was used to quantify the total intensity of fluorescence per worm as measured by intensity of each pixel in the selected area of a frame ( i . e . the worm ) . Vulval expression of Psod-3::GFP , which was very bright , was excluded from quantification , since this structure is not present in animals lacking the gonad . Fluorescence of the entire animal was measured for all other GFP constructs . No expression of any of the constructs was visible in embryos prior to egg laying . Image processing for figures was performed using Adobe Photoshop 7 . 0 ( Adobe ) . Sterile glp-1 ( e2141ts ) and wild-type N2 animals were raised at 25°C from L1 to day 1 of adulthood , then shifted to 20°C . On day 2 of adulthood , animals were collected for RNA extraction . RNA extraction , purification , and reverse transcription were carried out as described in [29] . qPCR was performed using the 7300 Real Time PCR System ( Applied Biosystems ) and data were analyzed using the Ct method ( Applied Biosystems Prism 7700 Users Bulletin No . 2 , http://docs . appliedbiosystems . com/pebiodocs/04303859 . pdf ) . Data were generated from at least two biological repeats . mRNA levels of ama-1 , nhr-23 , and Y45F10D . 4 were used for normalization [29] , [30] . p values were determined using one-way ANOVA . Primer sequences are available upon request .
|
Reproductive tissues are known to generate important intercellular signals . For example , in mammals , the reproductive tissues produce steroid hormones such as estrogen and testosterone that have profound effects on development and physiology . Studies of the nematode C . elegans and other organisms have shown that the reproductive system can also affect the rate at which an animal ages . Removal of C . elegans' germ cells extends lifespan but this effect is not simply due to sterility , as removal of both the somatic reproductive tissues and the germ cells does not extend lifespan . Instead , loss of the germ cells extends lifespan by activating a pathway that requires input from the somatic gonad . In this study , we demonstrate that the somatic reproductive tissues promote longevity by controlling the activity of a steroid signaling pathway that regulates the DAF-12 nuclear hormone receptor .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/complex",
"traits"
] |
2010
|
The Somatic Reproductive Tissues of C. elegans Promote Longevity through Steroid Hormone Signaling
|
The spatiotemporal control of cell polarity is crucial for the development of multicellular organisms and for reliable polarity switches during cell cycle progression in unicellular systems . A tight control of cell polarity is especially important in haploid budding yeast , where the new polarity site ( bud site ) is established next to the cell division site after cell separation . How cells coordinate the temporal establishment of two adjacent polarity sites remains elusive . Here , we report that the bud neck associated protein Gps1 ( GTPase-mediated polarity switch 1 ) establishes a novel polarity cue that concomitantly sustains Rho1-dependent polarization and inhibits premature Cdc42 activation at the site of cytokinesis . Failure of Gps1 regulation leads to daughter cell death due to rebudding inside the old bud site . Our findings provide unexpected insights into the temporal control of cytokinesis and describe the importance of a Gps1-dependent mechanism for highly accurate polarity switching between two closely connected locations .
Cell polarization , i . e . , the asymmetric distribution of subcellular structures and components , is critical for a variety of biological processes in uni- and multicellular organisms [1]–[3] . Rho GTPases are the major evolutionarily conserved regulators of polarity in yeast and mammalian cells . Of these G-proteins , RhoA , Cdc42 , and Rac1 , in particular , play essential roles in establishing polarity in different contexts , including asymmetric cell division , wound healing , apical–basal polarity of epithelial cells , and front–rear polarity of migrating cells [3]–[5] . The activation of Rho GTPases is under tight spatiotemporal control in response to extra- or intracellular polarity cues ( e . g . , chemical gradients , cell–cell interactions , or other landmarks ) during the establishment and maintenance of cell polarity [4] . In this context , membrane-associated proteins often mediate activation or inhibition of Rho GTPases [1] , [2] , [4] , [6] . Another feature of polarity establishment and maintenance is that Rho GTPases can be either simultaneously activated at various locations to fulfill different functions ( e . g . , during cell migration ) or sequentially activated at the same location through cross-talk mechanisms ( e . g . , during single cell wound healing ) [5] , [6] . Both mechanisms require elaborate systems that enable both temporal and spatial separation of Rho GTPase activation . In budding yeast , the Rho GTPases Rho1 ( RhoA homolog ) and Cdc42 are essential for bud growth , mating , and cell separation . Although Rho1 and Cdc42 share common effectors ( formins , exocyst ) , they play distinct roles during the establishment and maintenance of cell polarity [3] . Unlike Cdc42 , Rho1 has a well-established role in cytokinesis and cell separation . During anaphase , the Rho1 guanine nucleotide exchange factor ( GEF ) Tus1 recruits Rho1 to the cell division site ( bud neck ) , thereby promoting the formation of the contractile actomyosin ring ( AMR ) , which drives membrane ingression and primary septum formation during cytokinesis [7] , [8] . Rho1 remains at the cell division site after the AMR has contracted and is important for the final step of cell abscission [9] . In budding yeast , abscission is a two-step process that involves the formation of the secondary septum ( cell wall deposition ) and septum cleavage mediated by a set of hydrolases . Both processes rely upon targeted vesicle transport , and both are mediated by the Rho1 effectors Bni1 ( formin ) [10] and Sec3 ( part of the exocyst complex ) [11] . Cell wall deposition also depends on the Rho1 effector Fks1 ( β-1 , 6-glucan-synthase ) [12] . Although Cdc42 co-localizes with Rho1 at the bud neck during cytokinesis and abscission , its contribution to the physical process of cell separation remains elusive . After completion of cytokinesis , Cdc42 establishes the new polarity site , from which the future daughter cell ( bud ) will emerge , thereby determining the new cell polarity axis [3] . Interestingly , the new bud never emerges from the site used for cytokinesis during the preceding cell cycle , despite the fact that Cdc42 is recruited to the bud neck at the same time as Rho1 . These observations raise important questions regarding the role that Cdc42 plays at the cell division site and whether mechanisms are in place to inhibit Cdc42-dependent bud emergence at the bud neck during cytokinesis . One possibility could be that Cdc42 is restrained in an inactive state at the cell division site . Alternatively , Cdc42 may activate effectors at the bud neck different from those employed at the new bud site . The scaffold protein Bem1 and the GEF Cdc24 are major regulators of Cdc42 localization and activation [3] , [13]–[16] . Given that both molecules co-localize with Cdc42 at the bud neck and bud site [3] , it is unlikely that they are responsible for the specificity of Cdc42 towards different effectors . Nevertheless , the shift of the Cdc42-associated polarity site suggests that regulatory mechanisms are in place to couple the establishment of a new cell polarity site with the successful completion of cytokinesis and cell separation . However , such mechanisms have not been reported so far . Here , we show that the uncharacterized gene product of YPL158C , which we named Gps1 ( GTPase-mediated polarity switch 1 ) , functions as a novel polarity cue at the cell division site and is required for both Cdc42 and Rho1 function at this site . Our data show that Gps1 regulates both Rho1- and Cdc42-dependent signaling in parallel . Gps1 marks the closed cell division site post–AMR contraction and is required to maintain the Rho1-dependent polarization that is necessary for secondary septum formation and cell separation . At the same time , Gps1 specifically inhibits Cdc42-dependent activation of the p21-activated kinase Cla4 at the bud neck . This inhibition of Cla4 is essential to prevent budding inside the established cell division site . Failure in Gps1 regulation leads to the death of the daughter cell . We therefore propose that Gps1 is a critical polarity cue that confers a high degree of accuracy in polarity switching between two closely connected locations .
The highly conserved Cdc42 GTPase has a well-known function in polarity establishment during bud growth , which is consistent with its localization at the incipient bud site . However , Cdc42 forms a ring-like structure at the bud neck before AMR contraction , is reorganized as a disc-like structure after AMR constriction , and then splits into two well-separated signals just before cell separation ( Figures 1 and S1A ) . This complex pattern of localization at the bud neck during cytokinesis implies that Cdc42 may either perform an important function during cytokinesis or be inhibited during AMR contraction . To understand the regulation and function of Cdc42 , we screened for bud neck proteins that interact with Cdc42 using the yeast two-hybrid system . We identified the protein Gps1 as a putative Cdc42 interactor ( Figure 1B ) . Gps1 is a protein of uncharacterized function that was previously reported to localize at the bud neck [17] . Co-immunoprecipitation using a pair-wise combination of functional tagged proteins confirmed that Gps1 and Cdc42 interacted with one another ( Figure 1C ) . Interestingly , we observed that Gps1 also associated with the Rho GTPase Rho1 in the yeast two-hybrid system ( Figure 1B ) . In addition , tryptic peptides corresponding to Rho1 were identified by mass spectrometry analysis of purified Gps1-TAP complexes ( Figure S1C ) . To map the Cdc42 and Rho1 interaction domain of Gps1 , we generated several Gps1 truncations ( Figure 1D ) . We found that the N-terminal 290 residues of Gps1 ( Gps1-1-290 ) were sufficient to interact with both Rho1 and Cdc42 in the yeast two-hybrid system ( Figure 1D; data not shown ) . These interactions were most likely direct , as bacterially purified 6His-Gps1-1-290 associated with purified GST-Cdc42 or GST-Rho1 but not GST in vitro ( Figure 1E ) . Interestingly , Gps1-1-290 bound to both GDP- and GTP-locked mutants of Cdc42 and Rho1 in the yeast two-hybrid system ( Figure S2A and S2B ) and in vitro ( Figure S2C ) , indicating that Gps1 interacts with Cdc42 and Rho1 independently of their nucleotide-bound forms . Given that the temporal recruitment of Rho1 to the bud neck was similar to that of Cdc42 ( Figures 1A , S1A , and S1B S1B ) , we postulated that both Rho1 and Cdc42 are functionally linked to Gps1 . Live cell imaging showed that Gps1 formed a ring-like structure at the bud neck that co-localized with Myo1 before the onset of AMR contraction ( Figure 1F and 1G ) . However , Gps1 did not contract with the AMR . Instead , the Gps1 signal spread over the area defined by the initial ring and now formed a disc-like structure after AMR constriction . This area was confined by the septin ring ( Figures 1H and S1D ) . The septin complex may therefore act as a diffusion barrier to maintain Gps1 at the bud neck region after AMR contraction , as previously suggested for other components [18] . Importantly , despite its localization with Cdc42 and Rho1 at the bud neck during cytokinesis , Gps1 differed from either partner protein , by not relocating to the new polarization site ( Figure S1D ) , suggesting that Gps1 does not play any role at the new bud site . Together , the data indicate that Gps1 is a novel Cdc42- and Rho1-interacting protein that defines a unique micro-domain at the bud neck region that might have a specific role in late steps of cytokinesis and cell separation ( Figure 1I ) . To investigate whether Gps1 plays a role in cytokinesis , we analyzed the phenotype of GPS1 knock-out ( gps1Δ ) cells . The absence of GPS1 gave rise to cells carrying more than the regular 2C amount of DNA , and a large proportion of mother cells retained a connection to two or more buds ( Figure 2A–2C ) . This phenotype was rarely seen in wild-type cells ( frequency of less than 0 . 2% , in comparison to 40% in gps1Δ cells ) and strongly indicates that cell separation was delayed in the absence of GPS1 . We also found that gps1Δ cells were unable to survive in the absence of the cytokinetic component HOF1 ( gps1Δ hof1Δ; Figure 2D , lanes 3 and 4 ) . Hof1 plays an important role in coordinating AMR contraction with septum formation and cytoskeleton polarization toward the cell division site during cytokinesis [19] , [20] . The synthetically lethal interaction between GPS1 and HOF1 therefore indicates that Gps1 may work in a parallel pathway to Hof1 in promoting AMR contraction and/or septum formation during cytokinesis . Through the inspection of AMR behavior and primary septum formation by live cell imaging and transmission electron microscopy ( TEM ) , we excluded the possibility that Gps1 plays a role in primary septum biogenesis and/or AMR constriction ( Figure S3 ) . Importantly , TEM analysis revealed that formation of secondary septum cell wall material was strongly reduced in gps1Δ cells ( Figure 2E and 2F ) . Despite this defect , gps1Δ cells eventually separated from each other , giving rise to newborn cells that had a 2- to 3-fold reduction in the thickness of the layer of secondary septal material that had been deposited at the bud/birth scar ( the site at which cytokinesis took place ) ( Figure 2E and 2F ) . We hypothesized that the thinning of the secondary septum at the bud/birth scar in gps1Δ cells would engage the cell wall integrity ( CWI ) pathway . The CWI pathway is a mitogen-activated protein ( MAP ) kinase cascade that is activated upon cell wall stress to ensure cell survival by up-regulating genes that play important roles in cell wall synthesis and repair [21] . This assumption was confirmed by the observation that both protein levels and phosphorylated forms of the CWI-MAP kinase Slt2 were up-regulated in gps1Δ cells , as described for cells lacking GAS1 ( a 1 , 3-beta-glucanosyltransferase involved in the synthesis of the cell wall ) [22] ( Figure 2G and 2H ) . In addition , the survival of gps1Δ cells was dependent on SLT2 ( Figure 2I ) . Together , these data established that Gps1 is required for secondary septum assembly at the bud/birth scar . The expression of GPS1 has been previously shown to be under the control of the transcription factor Swi5 [23] . Swi5 is activated at the transition from mitosis into the G1 phase of the cell cycle [24] . Accordingly , Gps1-6HA protein levels strongly increased as cells exited mitosis , as monitored by the decline in the level of the mitotic cyclin Clb2 and the accumulation of the mitotic cyclin inhibitor Sic1 , whose expression is also under Swi5 control [25] ( Figure 2J ) . In addition , Gps1-6HA was depleted from cells lacking SWI5 ( Figure 2K ) , confirming that Swi5 controls GPS1 expression in a cell-cycle-dependent manner . The deletion of SWI5 impeded secondary septum formation ( Figure 2E and 2F ) , engaged the CWI pathway ( Figure 2L ) , and led to the accumulation of cells with a DNA content that exceeded 2C ( Figure 2M ) , in a manner that was comparable to GPS1 deletion . We therefore postulated that GPS1 might be the most important , if not the sole , gene under Swi5 control that is required for cytokinesis . In agreement with our hypothesis , the thickness of the secondary septum , the lethality of swi5Δ slt2Δ cells , and the accumulation of multi-nucleated cells were all rescued by expressing GPS1 from a Swi5-independent promoter ( pTEF2-GPS1; Figure 2E , 2F , 2L , and 2M ) . Taken together , these findings point to an important role of Swi5 in cytokinesis via regulation of GPS1 expression that is manifested in a cell-cycle-dependent manner . The interaction between Gps1 and Rho1 suggested that Gps1 controls secondary septum formation via activation of Rho1 . Rho1 activates the cell-wall- and septum-synthesizing Fks1 and two proteins that target this transmembrane protein to the bud neck: the formin Bni1 and the exocyst landmark protein Sec3 . To determine whether Gps1 does indeed regulate Rho1 , we monitored the distribution of Rho1 . For this , we used a GFP-RHO1 fusion expressed from the Rho1 endogenous promoter in a strain carrying untagged Rho1 . Our GFP-Rho1 is partially functional , as it could promote the growth of Rho1-depleted cells once expressed at higher expression levels ( Figure S4 ) . As previously reported [9] , GFP-Rho1 was recruited to the bud neck shortly before the onset of AMR contraction and persisted there until the cells separated ( Figure 3A ) . This temporal bud neck association is consistent with the role of Rho1 in AMR assembly/contraction as well as in secondary septum formation [7] , [12] . In gps1Δ cells , Rho1 was recruited to the bud neck shortly before AMR contraction , as in wild-type cells ( Figure 3A and 3B; Video S1 ) . Moreover , Rho1 accumulated to similar levels at the bud necks of both wild-type and gps1Δ cells ( Figure 3A and 3B; Video S2 ) , which is consistent with our observation that AMR and primary septum assembly were not affected by GPS1 deletion ( Figures 2E and S3 ) . However , whereas Rho1 levels increased at the bud neck of wild-type cells after AMR contraction , they declined sharply in gps1Δ cells and only reappeared at the tip of the newly emerging bud ( Figure 3A [arrowhead] and 3B; Videos S1 and S2 ) . This reduced bud neck sequestration was also found for the Rho1 effectors Bni1 , Sec3 , and Fks1 ( Figure 3C ) . Consequently , actin and myosin polarization toward the bud neck was , in general , impaired after AMR ring contraction in gps1Δ mutants ( Figure S5A–S5F ) . These data suggest that gps1Δ cells cannot maintain Rho1 and downstream targets at the bud neck after the initial wave of polarization that controls Rho1 recruitment to promote AMR and primary septum functions . We postulated that the levels of Rho1 at the bud neck after AMR contraction in gps1Δ cells would be insufficient to promote efficient Fks1 targeting and activity , thereby explaining the marked reduction of secondary septum material in these cells . In line with this assumption , FKS1 overexpression partially rescued gps1Δ defects in secondary septum formation ( Figure S5G ) . One prediction arising from this observation is that increasing Rho1 activity or tethering Rho1 to the bud neck independently of Gps1 function should rescue the secondary septum defect of gps1Δ cells . We observed that low overexpression of RHO1 partially rescued the growth defect of gps1Δ cells ( Figure S6A; Table S1 provides an overview of genetic manipulations that rescued gps1Δ defects ) , whereas higher levels of RHO1 caused a toxic effect ( Figure S6B ) . In addition , we found that the rapid cycling mutant Rho1-C25A ( expressed from the Rho1 endogenous promoter ) , but not wild-type or GTP-locked Rho1 , partially rescued the growth and secondary septum formation defects of the gps1Δ cells ( Figure S6C–S6F ) [9] . Next , we targeted Rho1 artificially to the bud neck in gps1Δ cells using the GFP-binding protein ( GBP ) strategy [26] . GFP-RHO1 wild-type , GTP-locked , or rapid cycling mutants [9] , [27] were expressed in cells in which the septin Shs1 was fused to GBP . This constitutively tethered Rho1 at the bud neck both in SHS1-GBP and SHS1-GBP gps1Δ cells ( Figure 3D and 3E ) . Whereas GFP-Rho1 was mainly lost from the bud neck in gps1Δ cells after AMR contraction , GFP-Rho1 remained at the bud neck in GPS1 , GPS1 SHS1-GBP , and gps1Δ SHS1-GBP cells until the cells had completed cell separation ( Figure 3D , red arrowheads indicate the time point of cell separation ) . The growth defect , the reduction in secondary septum assembly , and the impaired Rho1 effector localization in gps1Δ cells were all partially rescued by bud neck targeting of GFP-Rho1 via Shs1-GBP ( Figures 3F , 3G , S6G , and S6H ) . In this context , rapid cycling Rho1 mutants rescued the growth defect of gps1Δ mutants even better than wild-type Rho1 , whereas GTP-locked mutants showed no or just a little rescue ability ( Figure S6D–S6F ) . A similar inability of GTP-locked Rho1 mutants to rescue cytokinesis defects was reported previously [9] . Importantly , the increased bud neck targeting of the other bud-neck-residing Rho-GTPases , Rho2 , Rho4 , and Cdc42 , failed to rescue the growth of gps1Δ cells ( Figure S7A–S7D ) . This indicated that Rho2 and Rho4 did not influence the bud neck localization of Rho1 effectors , which is in line with the observation that the deletion of neither RHO2 nor RHO4 impaired the recruitment of Sec3 to the bud neck ( Figure S7E ) . These findings highlight the specificity of Gps1 function in promoting Rho1 bud neck association . Collectively , our data indicate that Gps1 is required to maintain Rho1 at the bud neck after AMR contraction , thereby contributing to the later functions performed by Rho1 in secondary septum formation and cell separation . We next assessed how Gps1 promotes the retention of Rho1 at the bud neck . The recruitment of Rho1 to the bud neck in late anaphase has been shown to depend upon GEF proteins , which themselves are targeted to the bud neck in a cell-cycle-dependent manner . During anaphase , the GEF Tus1 localizes at the bud neck and recruits Rho1 to promote AMR contraction and primary septum formation [8] . As both processes occurred normally in gps1Δ cells , we excluded the possibility that Gps1 controls Rho1 by regulating Tus1 ( Figure S3 ) . After mitotic exit , the GEF Rom2 controls Rho1 localization [9] . Rom2 localizes at the bud neck after AMR contraction in a manner that is dependent upon phosphatidylinositol 4 , 5-bisphosphate [28] . However , neither the bud neck localization of Rom2 nor the accumulation of phosphatidylinositol 4 , 5-bisphosphate at the bud neck was impaired in gps1Δ cells ( Figures S8 and S9 ) . Furthermore , constitutive targeting of either Rom2 or other Rho GEFs to the bud neck failed to rescue the growth phenotype of gps1Δ cells ( Figure S9 ) . This reinforces the view that Gps1 promotes Rho1 bud neck targeting in a Rom2-independent manner . Given that Gps1 did not promote Rho1 GEF localization , we asked whether Gps1 could negatively influence Rho1 GTPase-activating proteins ( GAPs ) . Interestingly , deletion of LRG1 , but not of other Rho GAP genes , rescued the growth defect of gps1Δ cells ( Figures 3I and S10A ) . Since Lrg1 levels at the bud neck were not altered in gps1Δ cells ( Figure S10B and S10C ) , we concluded that , although Rho1 was transiently recruited to the bud neck in both gps1Δ and gps1Δ lrg1Δ cells , deletion of LRG1 led to the accumulation of higher amounts of active Rho1 that compensated for its function in secondary septum formation . In agreement with this hypothesis , the deletion of LRG1 could negate the need for GPS1 in secondary septum formation , most likely because of the increased targeting of Rho1 effectors to the bud neck ( Figure 3C , 3H , and 3I ) . However , since LRG1 deletion in gps1Δ cells did not completely restore Rho1 effectors at the bud neck to the same levels observed in wild-type cells ( Figure 3C ) , we consider it unlikely that Gps1 regulates Rho1 exclusively through inhibition of Lrg1 . Taken together , our data indicate that Gps1 regulates Rho1 bud neck localization by a mechanism that is independent of Rho1 GAP or GEF bud neck targeting . Analysis of gps1Δ cells by live cell imaging revealed a high percentage of daughter cells ( 30%–40% ) that lysed shortly after AMR contraction and , consequently , were of dark appearance and stained by the cell death marker propidium iodide ( Figure S11; Video S3 ) . Because more than 95% of gps1Δ cells had a defect in secondary septum formation , we reasoned that daughter cell death could not solely arise from alterations in secondary septum assembly . Given that Cdc42 interacted with Gps1 ( Figure 1B–1E ) , we pursued the idea that daughter cell death in gps1Δ cells could be a consequence of mis-regulation of Cdc42 . To investigate whether Gps1 controls Cdc42 , we compared Cdc42 localization in wild-type and gps1Δ cells ( Figure 4A; Videos S4 and S5 ) . In wild-type and gps1Δ cells , Cdc42 was recruited to the bud neck prior to AMR contraction ( Figure 4A , 2–4 min , white asterisk; Videos S4 and S5 ) . However , whereas Cdc42 moved to the new bud site in every wild-type cell ( Figure 4A , white arrows ) , Cdc42 returned to the bud neck in more than 30% of gps1Δ cells ( Figure 4A , red asterisk ) . Strikingly , this backward migration of Cdc42 correlated with the death of the daughter cell , as indicated by the shrinkage and darkening of the bud ( Figure 4B , red arrowhead; Video S5 ) . Thus , Gps1 inhibits Cdc42-dependent repolarization and budding at the site of former cell division . Consistently , TEM revealed that more than 30% of gps1Δ cells underwent budding inside the old bud neck , which was marked by the formation of two or more layers of cell wall material at the bud neck ( Figure 4B , additional collars ) . To test whether active Cdc42 remained associated with the bud neck in the absence of GPS1 , we made use of the established in vivo marker for active Cdc42 ( Gic2-PBD-RFP ) [29] . We found that Gic2-PBD-RFP accumulated at the bud neck after AMR contraction in gps1Δ but not in wild-type cells , where Gic2-PBD-RFP became concentrated next to the cell division site ( Figure 4C and 4D ) . Furthermore , we observed that the Cdc42-activating GEF , Cdc24 , was enriched at the bud neck in gps1Δ but not in wild-type cells after AMR contraction ( Figure S12 ) . These data therefore suggest that Gps1 constitutes a safeguard mechanism that inhibits Cdc42 function at the bud neck after cytokinesis . Interestingly , Cdc42 activation at the old cell division site and the “budding inside the old bud neck” phenotype has previously been reported for Rga1-deficient cells [29] . Rga1 acts as a GAP that inactivates Cdc42 . Consequently , RGA1 deletion causes the bud neck accumulation of active Cdc42 ( Figure 4C and 4D ) . We thus postulated that Gps1 may work either together with , or in parallel to , Rga1 to control Cdc42 . We took a genetic approach to discriminate between these possibilities . The deletion of RGA1 alone did not affect cell growth but compromised the viability of gps1Δ cells ( Figure 4E ) . Importantly , RGA1 was the only Rho GTPase GAP gene whose deletion was synthetic lethal with gps1Δ ( Figure S10A ) . In comparison to gps1Δ and rga1Δ single mutants , gps1Δ rga1Δ cells showed an enhanced “budding inside the old bud neck” phenotype , as observed by the reduced number of bud scars formed at mother cells ( Figure 4F ) . Thus , our results suggest that Gps1 may work in a parallel mechanism to the Cdc42 GAP Rga1 to regulate Cdc42 function after AMR contraction . An important function of Cdc42 is to organize the actin cytoskeleton to allow growth of the new bud . We hypothesized that Gps1 may create an inhibitory zone at the old bud neck to locally impede Cdc42 function after cytokinesis ( Figure 5A ) . Based on the growth pattern of yeast cells , we expected this regulation to be especially important in haploid cells , where the new polarity site is formed adjacent to the old division site ( axial budding pattern ) . In contrast , diploid cells establish a new polarity site distally ( bipolar budding pattern ) . The establishment of the axial polarity site depends on a set of proteins ( Bud3 , Bud4 , Axl1 , and Axl2 ) that localize at the septin rings at the previous division site [30] . To test whether the mode of bud patterning would affect the survival of gps1Δ cells , we deleted genes reported to activate axial budding in haploid cells ( Figure 5B ) . The deletion of these genes causes a shift from an axial to a bipolar or random budding pattern in haploid cells [30] . Interestingly , the deletion of the activators of axial budding rescued the growth defect of gps1Δ cells ( Figure 5C ) , without affecting secondary septum formation ( Figure 5D and 5E ) . Therefore , we concluded that diminishing the probability of having a new bud site close to the old one increases the viability of gps1Δ cells . Conversely , one could imagine that forcing diploid cells to bud axially would have the opposite effect , i . e . , would compromise the viability of diploid gps1Δ/gps1Δ cells , which were otherwise viable ( Figure 5F ) . As predicted , the viability of diploid cells lacking GPS1 was significantly reduced after changing the budding pattern from bipolar to axial or proximal ( Figure 5F ) . Collectively , our data indicate that Gps1 provides an inhibitory signal at the old bud neck that prevents repolarization and rebudding within this area after cytokinesis . This function is important for cell survival . Cdc42 activates several downstream effectors , including the kinases Cla4 and Ste20 . To understand the function of Cdc42 that is regulated by Gps1 at the old bud neck after cytokinesis , we assessed the behavior of established Cdc42 loss-of-function mutants in the gps1Δ background . If Gps1 inhibits Cdc42 activity , the down-regulation of Cdc42 function should rescue the growth phenotype of gps1Δ cells . In agreement with this hypothesis , a mutation in the effector binding switch I region of Cdc42 ( cdc42-T35A ) rescued the growth defect of gps1Δ cells ( Figure 6A ) . This rescue , however , was not consistently supported by locking Cdc42-T35A in the GTP- or GDP-bound forms ( Figure S13 ) . Therefore , it is unclear whether the nucleotide-bound state of Cdc42-T35A influences the rescue of gps1Δ cells . To gain a deeper insight onto Cdc42-T35A function in vivo , we monitored its subcellular localization during cytokinesis and polarity establishment in the presence of wild-type Cdc42 . Cdc42-T35A was recruited to the bud neck during cytokinesis like wild-type Cdc42 ( Figure 6B ) . However , in contrast to Cdc42 , Cdc42-T35A failed to move to the new polarity site after cytokinesis ( as judged by the recruitment of Myo1 to the new bud neck ) ( Figure 6B ) . Rather , Cdc42-T35A persisted at the old bud neck . Consequently , we suggest that Cdc42-T35A may compete with Cdc42 at the bud neck and probably compensates for the absence of the inhibitory function of Gps1 towards wild-type Cdc42 . Given that Cdc42-T35A has a weaker interaction with effectors , deletion of the gene that encodes the effector protein should also promote the growth of gps1Δ cells . Interestingly , the deletion of CLA4 , but no other Cdc42 effector , improved the growth fitness of gps1Δ cells ( Figures 6C and S14 ) , suggesting that activation of Cla4 by Cdc42 impairs cell growth in the absence of GPS1 . Consistently , ectopic overexpression of CLA4 reduced the viability of gps1Δ cells ( Figure 6D ) . Surprisingly , deletion of the Cdc42 effector kinase STE20 had the opposite impact of CLA4 deletion , as gps1Δ ste20Δ cells were unable to survive ( Figure 6C ) , and overexpression of STE20 rescued the growth lethality of gps1Δ cells ( Figure 6D ) . Importantly , the effect of CLA4 and STE20 on the growth of gps1Δ cells was dependent upon both the kinase activity and the CRIB-domain that mediates the binding of Cla4 and Ste20 to Cdc42 ( Figure 6E and 6F ) . We therefore concluded that Cdc42-dependent Cla4 activation must be down-regulated , while Cdc42-dependent Ste20 activation must be promoted , to provide viability of cells lacking GPS1 . To examine how Cla4 and Ste20 influence the growth of gps1Δ cells , we analyzed the phenotype of gps1Δ ste20Δ and gps1Δ cla4Δ cells in detail . The synthetic defect of ste20Δ gps1Δ double mutants was based on an increased “budding inside the old bud neck” phenotype ( Figure 6G ) . Interestingly , Ste20 is required to promote bipolar budding [31] , suggesting that increased budding in close proximity to the bud neck may account for the lethality of ste20Δ gps1Δ cells . Indeed , the lethality of ste20Δ gps1Δ double mutants was reverted upon deletion of BUD1 or BUD4 to switch the budding pattern from proximal to bipolar or random ( Figure 6H ) . This suggested that Ste20 activity may be especially important to place the new bud site next to the old cell division site . Analysis of cdc42-T35A gps1Δ and cla4Δ gps1Δ cells by TEM displayed a much less pronounced “budding inside the old bud neck” phenotype than was observed for gps1Δ cells ( Figure 6I ) , indicating that inhibition of Cla4 activity at the bud neck is required to avoid rebudding inside the old bud neck . Taken together our results indicate that Gps1 inhibits Cdc42-dependent polarity establishment at the old bud neck most likely by promoting Ste20 and inhibiting Cla4 activation . To determine whether Gps1 prevents premature activation of Cla4 at the old bud neck , we analyzed the phosphorylation status of Cdc24 , a well-known substrate of Cla4 ( Figure 7A ) [13] . In both wild-type and gps1Δ cells , Cla4-dependent hyperphosphorylated forms of Cdc24 were observed ( Figure 7A , asterisk ) , suggesting that hyperphosphorylation of Cdc24 per se was not impaired in the absence of GPS1 . To investigate the timing of Cdc24 phosphorylation , we transiently arrested wild-type and gps1Δ cells in metaphase and followed cell cycle progression and Cdc24 hyperphosphorylation after release from the metaphase block . Both wild-type and gps1Δ cells progressed through the cell cycle and completed anaphase with similar kinetics ( Figure 7A ) . However , hyperphosphorylation of Cdc24 and budding were both advanced in gps1Δ cells ( Figure 7A ) . This indicates that Cdc24 may prematurely activate Cdc42 in cells lacking GPS1 . In agreement with this hypothesis , we found premature accumulation of Cla4 at the cell division site in gps1Δ mutants ( Figures 7B , 7C , and S15 ) . In addition to Cdc24 activation , Cla4 supports the formation of a stable septin ring at the incipient budding site [32] . To do so , the septin ring has to disassemble at the bud neck after cytokinesis . If Cla4 is prematurely activated at the old bud neck in gps1Δ cells , the septin ring should be formed and/or maintained at the old bud neck after cytokinesis . Indeed , whereas the septin ring disappeared from the mother cell bud neck after AMR constriction and reappeared at the incipient bud in wild-type cells ( Figure 7D , GPS1 , red arrowheads ) , it persisted at the mother cell bud neck of all gps1Δ cells in which the daughter cell died after cytokinesis ( Figure 7D , gps1Δ , white arrowheads ) . Collectively , our data indicate that Gps1 is essential to keep Cla4 inactive at the old bud neck during cytokinesis .
We understand little about the spatiotemporal coordination of events that follow the contraction of AMR and lead to cell separation and establishment of a new cell polarity axis in budding yeast . The two major Rho GTPases , Cdc42 and Rho1 , are recruited to the cell division site prior to the onset of AMR contraction . Rho1 contributes to the formation of the AMR and , subsequently , to secondary septum biogenesis and cell separation . The functions executed by Cdc42 at the cell division site remain to be elucidated . Puzzlingly , Cdc42 plays an important role in the establishment of the axial polarity and bud growth that occurs adjacent to the site of the previous cell division . Although Cdc42 is recruited to the site of cell division prior to AMR contraction , bud growth does not begin at this site during cytokinesis . How this specific Cdc42 function is inhibited during cytokinesis is unknown . Here , we identified the protein Gps1 as a bud neck polarity cue that integrates polarity signaling by maintaining Rho1 and inhibiting Cdc42 at the old cell division site . We propose that the Gps1-mediated mechanism facilitates and safeguards a polarity switch between adjacent sites ( old bud neck and new bud neck ) , thereby coordinating cell division with the establishment of a new cell polarity axis . We identified Gps1 as a direct interactor of both Rho1 and Cdc42 in the yeast two-hybrid system and established by in vitro and in vivo pull-down experiments that Gps1 is in complex with Rho1 and Cdc42 . Live cell imaging of Gps1-GFP revealed an intricate pattern of bud neck recruitment . During AMR constriction and primary septum growth , a new membrane compartment is established by the fusion of membrane material provided by the secretory pathway . Gps1 covered exclusively this emerging disc-like compartment that forms after AMR contraction . Our data show that the Gps1 disc is confined at the bud neck by the splitted septin rings , which act as a diffusion barrier for membrane proteins [18] . Gps1 may then be maintained at the cell division site through a direct or indirect interaction with the membrane or transmembrane protein . Alternatively , Gps1 may associate with the septins via the septin interactors Nap1 , Nba1 , and Nis1 , which were found in the purified Gps1 complex ( Figure S1 ) . Our data are consistent with a role for Gps1 in the coordination of Rho1 and Cdc42 function after AMR contraction . During cytokinesis , actin repolarization towards the bud neck is required for both AMR assembly/contraction and targeted delivery of components involved in septum formation . Rho1 , but not Cdc42 , plays an important role in actin assembly and vesicle delivery through activation of the formin Bni1 and the exocyst landmark Sec3 at the bud neck . In late anaphase , Rho1 accumulates at the bud neck through its GEF , Tus1 . This initial pool of active , GTP-bound Rho1 is required for the completion of AMR assembly and contraction and primary septum formation [8] . Our data now indicate that Tus1 can sustain Rho1 at the bud neck for a very limited period , the time of AMR constriction . The maintenance of Rho1 localization at the bud neck post–AMR contraction was largely dependent upon Gps1 . Accordingly , the localization of Rho1 effectors , including Fks1 , Bni1 , and Sec3 , was impaired in cells lacking GPS1 , which reflected the diminished density of actin cables and the absence of secondary septum formation at the bud neck after AMR contraction in gps1Δ cells . Since the localization of other bud-neck-residing Rho1 GEFs such as Rom2 was not impaired , we suggest that Gps1 is directly responsible for the maintenance of Rho1 at the bud neck . Supporting this view , engineered tethering of Rho1 to the bud neck partially relieved the requirement of Gps1 for cytokinesis ( Figure 3E and 3F ) . In contrast to Rho1 , Gps1 inhibited the activation of Cdc42 at the old cell division site after cytokinesis . Consequently , in cells lacking GPS1 , frequent budding inside the old bud neck was observed by TEM analysis . A similar phenotype has been described for cells bearing hyperactive Cdc42 upon GAP depletion ( rga1Δ cells ) [29] . Rga1 co-localizes with the septin ring that surrounds the cell division site and the Gps1 disc . The synthetic lethal phenotype observed upon co-deletion of GPS1 and RGA1 suggests that Gps1 and Rga1 may act in different pathways to inhibit Cdc42 . We therefore suggest a model whereby Rga1 regulates the temporal activation of Cdc42 , whereas Gps1 dictates directionality by inhibiting Cdc42 at the old cell division site ( Figure 8 ) . Importantly , in contrast to rga1Δ cells , gps1Δ cells displayed a strong correlation between relocation of Cdc42 to the old bud neck and death of the daughter cell . Genetic manipulations that rescued secondary septum formation improved the growth viability of gps1Δ cells without reverting the phenotype of budding inside the old bud neck ( Table S1 ) . Conversely , genetic manipulations that decreased the probability of axial budding in haploid cells rescued the growth fitness of gps1Δ cells independently of secondary septum biogenesis ( Table S1 ) . This indicates that gps1Δ cells die as a consequence of two combined defects: formation of a new bud at the old bud neck , and an incorrectly formed secondary septum at that location . We therefore concluded that Gps1 is required for cell viability through the temporal coordination of both the Cdc42 and Rho1 pathways . Which function of Cdc42 does Gps1 inhibit ? Cdc42 and its GEF , Cdc24 , associate with the cell division site prior to the onset of AMR contraction . Although the function of Cdc42 at the bud neck at this early stage of cytokinesis is unclear , it is well established that GTP-bound Cdc42 activates several effectors required for polarized growth , including the kinase Cla4 . Several lines of evidence indicate that Gps1 inhibits premature Cla4 activation . First , deletion of CLA4 , but not other Cdc42 effectors , rescued the growth defect caused by deletion of GPS1 . Second , unlike the normal controls seen in wild-type cells , where Cla4 weakly associated with the bud neck , Cla4 accumulated to high levels at the bud neck and prematurely phosphorylated Cdc24 in gps1Δ cells ( Figure 7A–7C ) . Finally , the Cdc42-T35A protein , which fails to activate Cla4 [13] but efficiently localizes at the bud neck ( Figure 6B ) , rescued budding inside the old bud neck and cell viability of gps1Δ cells . Thus , we postulate that Gps1 specifically inhibits Cdc42-dependent Cla4 activation at the old bud neck , thereby supporting budding outside the area where cytokinesis takes place . Why should a cell prevent budding inside the old bud scars ? First , budding inside the old site of cell division makes cells more vulnerable to defects in secondary septum biogenesis , which may be inheritable ( as in case of GPS1 knock-out ) or induced by drugs/toxins that affect cell wall formation . Second , electron micrographs show that recurrent budding at the same position narrows the diameter of the bud neck ( Figure 4B ) . A narrower bud neck may hinder chromosome segregation or partitioning of organelles between mother and daughter cells . In either case , budding inside the old site of cell division increases the probability of cell death . Cell cycle progression is mainly regulated by differential gene expression , post-translational modifications , and protein degradation . In Saccharomyces cerevisiae , the hallmark of the mitotic exit is the activation of the phosphatase Cdc14 , which drives the M–G1 phase transition and is a prerequisite for cytokinesis . Cdc14 counteracts mitotic Cdk1 , thereby activating the transcription factors Ace2 and Swi5 [24] . Ace2 is well known for its cytokinesis-related function in activating the daughter-specific expression of the chitinase and glucanases that are responsible for the final cleavage of the daughter from the mother cell [33] . In contrast to Ace2 , a direct role for Swi5 in cytokinesis and cell separation has remained obscure to date . We showed that Swi5 is essential for Gps1 expression and consequently also for secondary septum formation ( Figure 2E ) . This dependency can be overcome by expressing GPS1 under the control of a Swi5-independent promoter . The finding that Swi5 regulates secondary septum formation , whereas Ace2 is essential for the following cell cleavage , raises the question whether there is a cross-talk between Swi5- and Ace2-dependent activation to ensure the correct order of function of the Swi5- and Ace2-dependent targets . In conclusion , the analysis of Gps1 function presented here has provided novel insights into the spatiotemporal coordination of polarity during the late cytokinetic and early morphogenetic events that are under control of Rho1 and Cdc42 signaling . We anticipate that similar mechanisms might coordinate the final stages of division and the reestablishment of cell polarity in other systems in which membrane-associated polarity cues act as a Rho GTPase positioning system and/or promote Rho GTPase activation and inactivation to ensure either a switch-like change of polarity or the maintenance of a stable polarity gradient that is characterized by a spatial separation of Rho GTPase activity .
Yeast strains and plasmids used in this study are listed in Tables S2 and S3 . CDC42 and RHO1 constructs were expressed from their endogenous promoter . Yeast growth and media were as described previously [34] . Gene deletions and epitope tagging were performed using PCR-based methods [35] , [36] . Yeast strains were grown in yeast peptone dextrose medium containing 0 . 1 mg/l adenine ( YPAD ) . Osmo-sensitive mutants were grown in medium containing 1 M sorbitol for osmo-stabilization . Strains carrying plasmids were grown in synthetic complete medium lacking the corresponding amino acids . For synchronization of cells in the G1 phase , 10 µg/ml synthetic alpha-factor ( Sigma-Aldrich ) was added to cultures in the early log phase ( 5×106 cells/ml ) and incubated for <2 . 5 h at 30°C or <4 h at 23°C until >95% of the cells formed mating projections . To arrest the cells with nocodazole , 15 µg/ml nocodazole ( Sigma-Aldrich ) was added to the culture medium and incubated 2–4 h until >90% of the cells arrested with large buds and one DNA-stained region ( DAPI staining ) . Loss of URA3-containing plasmids was assessed by using plates containing 1 mg/ml 5-fluoroorotic acid ( 5-FOA ) . The plasmid shuffle strategy was used to test the viability of single and double mutants ( genetic interaction ) . Briefly , mutant strains containing the corresponding wild-type gene on an URA3-based plasmid ( pRS316 ) were analyzed for growth on 5-FOA plates ( which select against URA3 ) . At least six individual transformants were analyzed for each mutant , and one representative mutant is shown in Figures 2D , 2I , 2L , 3F , 3I , 4E , 5C , 5F , 6A , 6C–6F , 6H , S4B , S6A , S6B , S6D , S7A , S9A , S10A , S13A , S13B , and S14 . For each mutant , 10-fold serial dilutions of wild type ( ESM356 or YPH499 ) and mutants with the indicated genotype were spotted onto synthetic complete medium and 5-FOA plates . Mutants were complemented by pRS316-HOF1 ( Figure 2D ) and by pRS316-GPS1 ( Figures 3F , 3I , 4E , 5C , 5F , 6A , 6C–6F , 6H , S4B , S6A , S6B , S6D , S7A , S9A , S10A , S13A , S13B , and S14 ) . Strains carrying genes on a LEU2-based plasmid ( Figures 6A , 6D–6F , S4B , S6A , S6B , and S13A ) were grown on medium lacking leucine to avoid plasmid loss . Plates were incubated for 1 or 2 d at 30°C . Yeast protein extracts and Western blotting were performed as described previously [35] . Antibodies were rabbit anti-GFP antibody , mouse anti-tubulin ( Tub1 ) , mouse anti-HA ( clone 12CA5 , Sigma ) , mouse anti-Myc ( clone 9E10 , Sigma ) , rabbit anti-Clb2 , and guinea pig anti-Sic1 [37] . Secondary antibodies were goat anti-mouse , goat anti-rabbit , and goat anti–guinea pig IgGs coupled to horseradish peroxidase ( Jackson ImmunoResearch Laboratories ) . Slt2 was probed with anti-Mpk1 ( y-244 , Santa Cruz Biotechnology ) , and phosphorylated Slt2 was probed with Phospho-p44/42 MAPK ( T202/Y204 , New England BioLabs ) . To determine the proportion of activated Slt2 , the measured values of phosphorylated and total Slt2 were corrected by subtracting the background signal . Corrected values were plotted as the ratio of phosphorylated Slt2 to total Slt2 ( Figure 2H ) . For fluorescence still image analysis , cells carrying GFP or Cherry fusion proteins were fixed in 4% formaldehyde for 10–30 min before inspection . Cells were fixed with 70% ethanol and re-suspended in PBS containing 1 µg/ml DAPI ( 4′ , 6-diamino-2-phenylindole , Sigma ) for DNA visualization ( DAPI staining ) . For actin staining , cells were fixed for 60 min in 4% formaldehyde solution in PBS . Actin and bud scars were stained with 2 µM rhodamine-phalloidine ( Invitrogen ) and 0 . 1 mg/ml calcofluor ( Fluorescence Brightener 28 , Aldrich-Sigma ) , respectively . Dead cells were stained with 5 µg/ml propidium iodide . Live cell imaging and quantification of fluorescence still images were performed as described previously [20] . For quantification and image preparation , 3–5 z-sections of the region of interest were averaged or maximum-projected , respectively . Specimens for electron microscopy were prepared as described previously [20] . Flow cytometry was performed as described previously [39] . Bud scars were stained with Fluorescence Brightener 28 ( calcofluor , Sigma-Aldrich ) . Cells were assigned to one of three categories: cells without any bud scar ( daughter cells ) , cells with one bud scar ( mother cells ) , and cells with two or more bud scars ( mother cells ) . PIP2 ( phosphatidylinositol 4 , 5-bisphosphate ) levels were determined using the established PIP2 reporter GFP-2xPH ( PLC∂ ) [40] . To quantify the bud-neck-associated levels of the PIP2 reporter GFP-2xPH ( PLC∂ ) , grey values were measured along the mother–daughter axis of still images , as depicted in Figure S8B . PIP2 levels were measured in cells with a large bud after AMR contraction ( Figure S8C and S8D ) or during cytokinesis ( Figure S8E–S8H ) . For normalization , the level of PIP2 associated with the mother cell membrane was set to 1 in each cell analyzed .
|
In budding yeast , cell polarization ( or the asymmetric distribution of subcellular components ) ensures the targeted transport of proteins and membrane material to the sites of cell growth or cell division in late mitosis . Two conserved members of the Rho-GTPase family , Rho1 and Cdc42 , are master regulators of cell polarity . While Rho1 has a well-established role in cytokinesis and cell separation , Cdc42 helps to establish the new polarity site from which the future daughter cell will grow after cytokinesis . Interestingly , despite the fact that Cdc42 is recruited to the site of cell division at the same time as Rho1 , the new daughter cell never emerges from the site previously used for cytokinesis during the preceding cell cycle , and it remains elusive how cells coordinate the distinct functions of Rho1 and Cdc42 during cytokinesis . Here , we show that the novel protein Gps1 marks the cell division site , where it maintains Rho1-dependent polarity until cell separation is completed . We also demonstrate that Gps1 prevents activation of Cdc42 at the site of cell division during cytokinesis . We propose that Gps1 provides a novel polarity cue that guides the establishment of a new polarity site , away from the old site of cell division , where the new daughter cell then emerges .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"cellular",
"structures",
"model",
"organisms",
"molecular",
"cell",
"biology",
"cell",
"division",
"cell",
"biology",
"cytokinesis",
"yeast",
"and",
"fungal",
"models",
"biology",
"saccharomyces",
"cerevisiae",
"microbiology",
"cytoskeleton"
] |
2013
|
A Safeguard Mechanism Regulates Rho GTPases to Coordinate Cytokinesis with the Establishment of Cell Polarity
|
Because of great challenges and workload in deleting genes on a large scale , the functions of most genes in pathogenic fungi are still unclear . In this study , we developed a high-throughput gene knockout system using a novel yeast-Escherichia-Agrobacterium shuttle vector , pKO1B , in the rice blast fungus Magnaporthe oryzae . Using this method , we deleted 104 fungal-specific Zn2Cys6 transcription factor ( TF ) genes in M . oryzae . We then analyzed the phenotypes of these mutants with regard to growth , asexual and infection-related development , pathogenesis , and 9 abiotic stresses . The resulting data provide new insights into how this rice pathogen of global significance regulates important traits in the infection cycle through Zn2Cys6TF genes . A large variation in biological functions of Zn2Cys6TF genes was observed under the conditions tested . Sixty-one of 104 Zn2Cys6 TF genes were found to be required for fungal development . In-depth analysis of TF genes revealed that TF genes involved in pathogenicity frequently tend to function in multiple development stages , and disclosed many highly conserved but unidentified functional TF genes of importance in the fungal kingdom . We further found that the virulence-required TF genes GPF1 and CNF2 have similar regulation mechanisms in the gene expression involved in pathogenicity . These experimental validations clearly demonstrated the value of a high-throughput gene knockout system in understanding the biological functions of genes on a genome scale in fungi , and provided a solid foundation for elucidating the gene expression network that regulates the development and pathogenicity of M . oryzae .
Magnaporthe oryzae is the best-studied phytopathogenic fungus , which was voted first in the top 10 list of fungal plant pathogens by an international community of molecular plant pathologists [1] . The importance of this filamentous ascomycete fungus is not only owing to the fact that the rice blast disease caused by the fungus is the most destructive disease of rice throughout the world , which typically leads to 20–30% losses and even complete loss in grain production during regional epidemics [1] , [2] , but also to its being a primary model in the study of host–fungal pathogen interactions [3] . The rice blast fungus has a complicated life cycle including hyphal growth , conidiogenesis , conidial germination , appressorium formation and plant infection , which provides substantial biological information in eukaryotic development and pathogenesis . The rice blast fungus is highly amenable to molecular genetic manipulation , and the functions of numerous genes are identified by gene knockout or ectopic insertion [4]–[6] . Transcription factors ( TFs ) are proteins that bind to specific DNA sequences , thereby controlling the flow of genetic information from DNA to mRNA . In the rice blast fungus , more than 522 putative TF proteins have been identified from 12 , 991 M . oryzae proteins ( www . ftfd . snu . ac . kr; www . broadinstitute . org ) . Therefore , nearly 4 . 02% of genes in the genome code for TFs , which makes this family the single largest family of M . oryzae proteins . Recent functional analyses of single or several TF genes revealed their critical biological roles in fungal development , pathogenesis and response to the environment , for instance , in hyphal growth ( MNH6 , MSTU1 , MoCRZ1 and MoSWI6 ) [7]–[10] , conidiogenesis ( COM1 , CON7 , COS1 , MNH6 , and MoHOX2/HTF1 ) [7] , [11]–[16] , conidial germination ( TRA1 ) [17] , appressorium formation ( MoLDB1 , MoSOM1 and MoCDTF1 ) [18] , [19] , plant infection ( COM1 , MNH6 , MIG1 , MST12/MoHOX8 , and MoSfl1 ) [7] , [11] , [16] , [20]–[22] , and response to oxidative stress ( MoATF1 and MoAP1 ) [23] , [24] or light ( MgWC-1 ) [25] . However , the biological functions of most TFs have not been revealed , mainly because it is difficult to delete genes on a large scale and because a high-throughput gene knockout method has not been previously established in M . oryzae . Several loss-of-function techniques , such as homologous recombination [4] , insertional mutagenesis [6] and RNA interference [26] , have been used to investigate gene functions in fungi . Although insertional mutagenesis and RNA interference have been effectively used to inactivate gene functions , they all have limitations that prevent them from becoming high-throughput gene knockout techniques , including 1 ) inability to cover the genome and the heavy workload to clone genes in the insertional mutagenesis method [6] , and 2 ) the silencing of potential unintended targets by the RNA interference method [26] . Gene knockout through DNA homologous recombination is a primary technique used to inactivate gene functions . More than 650 putative TF genes were disrupted in Fusarium graminearum using homologous recombination [27] . In that study , the double-joint PCR method [28] was used to generate gene knockout constructs , which were then transformed into fungal protoplasts . In Neurospora crassa , homologous recombination was developed into a high-throughput gene knockout technique [29] , in which gene knockout constructs generated by making use of a yeast recombinational cloning method were transformed into conidia by electroporation . Yeast recombinational cloning is a more suitable method than double-joint PCR for high-throughput gene knockout study . However , protoplast transformation is an elaborate , time-consuming and inefficient method , and electroporation of germinating conidia has not been built in the rice blast fungus and many other filamentous fungi . The frequency of homologous recombination after transformation is often quite low in filamentous fungi , typically <10% in M . oryzae [30] . In addition , mutant screening is a tedious and inefficient step in high-throughput gene knockout studies . The implementation of KU70/KU80 ( mus-51/mus-52 ) mutations as gene knockout background strains greatly increases the frequency of homologous recombination in N . crassa , Aspergilli and other filamentous fungi , and reduces the workload in screening [29] , [31]–[34] . However , the null mutants obtained from KU70/KU80 deletion background strains need to be complemented with native KU70/KU80 genes before analyzing their mutant phenotypes [29] . These shortcomings limit the application of the above methods in research in fungal functional genomics . Until now , there is still a surge in interest in functional genomics research through the systematic mutagenesis of identified genes sequenced in the genomes of a large number of fungi ( http://www . ncbi . nlm . nih . gov/genome/ ) . To learn the biological functions of TFs at the genome level , we constructed a high-throughput gene knockout method that enables the rapid knockout of large numbers of genes in the rice blast fungus . In this method , gene knockout vectors are built by a yeast recombinational cloning method using a high-throughput way , DNA transformations are performed by Agrobacterium tumefaciens-mediated transformation ( ATMT ) , and the null mutants are identified by the negative ( GFP ) /positive ( a resistant gene ) screening system through a novel yeast-Escherichia-Agrobacterium shuttle vector , pKO1B . With this system , we deleted 104 putative fungal-specific Zn2Cys6 TF genes in M . oryzae . The null mutants were then examined for their phenotypes in development , pathogenicity and responses to stress conditions . In particular , many Zn2Cys6 TF genes required for fungal growth , asexual development , conidial germination and appressorium formation , pathogenicity , and response to stress were identified . We further identified the genes regulated by GPF1 and CNF2 via RNA-sequencing ( RNA-seq ) and found that GPF1 and CNF2 have similar mechanisms in the regulation of gene expression related to fungal pathogenicity . Our findings will provide new insights into the transcriptional regulation of fungal development and pathogenicity , and a way to study the biological functions of genes in fungi at the genome level .
Three major challenges in high-throughput gene knockout of fungi are 1 ) how to build gene-deletion cassettes quickly , 2 ) how to transfer DNA into fungal cells easily , and 3 ) how to identify null mutants from the large number of transformants efficiently . To solve these problems , we developed a high-throughput gene knockout system using a newly designed yeast-Escherichia-Agrobacterium shuttle vector , pKO1B , ( Figure 1A ) . Through this plasmid , three techniques suitable for high-throughput manipulation ( yeast recombinational cloning to build gene-deletion cassettes , ATMT to transfer DNA into fungal cells and dual selection system to identify null mutants ) were combined into the high-throughput gene knockout system . pKO1B is a vector built on the framework of the binary vector pCAMBIA1300 ( www . cambia . org ) . It contains the URA3-2micro2_origin sequence from a yeast plasmid pYES2 ( Invitrogen , USA ) and eGFP gene under the control of a strong promoter of M . oryzae H3 histone gene [35] ( Figure 1A ) . pKO1B has a new characteristic: the ability to be replicated in three organisms , Saccharomyces cerevisiae , Escherichia coli and A . tumefaciens . Through the gapped vector pKO1B , three DNA fragments ( 5′ and 3′ flanking fragments of the targeted gene and a resistant gene fragment ) could be merged into a gene-deletion cassette in one step by yeast recombinational cloning ( Figure 1B ) . The knockout vectors obtained were transformed into A . tumefaciens without transferring the gene-deletion cassettes from a yeast plasmid to another binary Agrobacterium plasmid . The gene-deletion cassettes in pKO1B were transformed into fungal cells using the ATMT method . When the gene-deletion cassettes were ectopically integrated into fungal genomic DNA , GFP was activated in ectopic insertional transformants and was easily observed under a fluorescence microscope as a negative selective marker , but not in null mutants ( Figure 1C and 1D ) . To screen null mutants efficiently , the genomic DNAs of the transformants without green fluorescence were isolated using an improved CTAB method in a high-throughput way ( shown in Materials and Methods ) and then were detected for the targeted gene and β-tubulin gene by double PCR . If the targeted gene was deleted in a mutant , only one band for β-tubulin appeared as a positive control on an electrophoretic gel; otherwise , there were two bands with one for the targeted gene and the other for β-tubulin in ectopic insertion transformants ( Figure 1C and 1E ) . For those null mutants identified in the above negative screening PCR , we then continued to search for the unique recombinational DNA fragment that indicated a knockout event and only appeared in the null mutants by PCR . In this PCR , one primer was limited in the genomic DNA outside of the 5′ or 3′ flanking fragment of the targeted gene , and the other primer was limited in the resistant gene of the gene-deletion cassettes ( primers p1+p2 or p3+p4 in Figure 1C ) . There was one band 1 . 2–2 . 0 kb in length on the electrophoretic gel that appeared in null mutants . Otherwise , there was no band appearing in the ectopic transformants and wild-type strain ( Figure 1F ) . Gene knockout and ectopic insertion maybe happened coincidentally ( Figure 1C ) . To identify those mutants that were undetected in previous screening steps , copies of transformed gene deletion cassettes in null mutants were confirmed by qPCR after comparison with the wild-type strain using β-tubulin as a control . The null mutants had one copy of the gene deletion cassette , while the null+ectopic transformants had more than two copies and the wild-type strain none . Finally , the mutants containing a single copy of the gene deletion cassette were considered null mutants . TFs containing fungal-specific Zn2Cys6 zinc finger and fungal_TF_MHR Domain are the largest group of TFs in the rice blast fungus . The 163 putative fungal-specific Zn2Cys6 genes , previously annotated at the Fungal Transcription Factor Database ( http://ftfd . snu . ac . kr/index . php ) [36] or obtained by BLAST searches at NCBI under organism item ( rice blast fungus , taxid: 318829 ) with the Conserved Domain program [37] , were selected to generate gene deletion mutants ( Table S1 in Text S1 ) . As a result , 104 TF genes were identified to be deleted in M . oryzae ( Table S2 and Table S3 in Text S1 ) . The knockout rate of 104 genes was 15 . 03±12 . 02% ( from 0 . 68 to 62 . 07% ) . The causes for tens of TF genes not deleted in this experiment lie in the following . 1 ) Knockout vectors of twenty-one genes had not been constructed . 2 ) For those genes with a knockout rate less than 1% , more transformants need to be screened . 3 ) Because the SUR gene was selected as a resistance gene , the mutants must be screened on DCM medium . But the mutants of TF genes that are involved with amino acid metabolism cannot grow on DCM medium . 4 ) Some genes may be essential genes . We analyzed the phenotypes of the 104 Zn2Cys6 TF null mutants at multiple developmental stages that M . oryzae likely encounters during its infection cycle in rice . These phenotypes analyzed included: developmental characteristics ( mycelial growth , colony color and mycelial shape; conidiation; conidial germination; and appressorium formation ) and virulence to plants ( barley leaf explants , rice leaf explants and rice seedlings ) . A substantial fraction of the mutants ( 58 . 7% , 61/104 ) clearly displayed visible phenotypes ( Figure 2A ) . The results ( Figure 2B ) revealed that 27 of the 104 TF genes studied were involved in mycelial growth ( Figure S1 ) , 25 TF genes in conidial production , 12 TF genes in conidial germination and 10 TF genes in appressorium formation , and 5 TF genes were involved in pathogenicity in barley ( Figure S2 ) , while 7 TF genes were involved in pathogenicity in rice ( Figure S3 ) . Table 1 and Table S4 in Text S1 provide a complete list of the phenotypic analyses . Among these mutants displaying clearly visible mutant phenotype changes , 42 . 6% ( 26/61 ) of mutants exhibited multiple mutant phenotypes and 57 . 4% ( 35/61 ) exhibited a single mutant phenotype ( Figure 2A ) . The Zn2Cys6 TFs regulating asexual reproduction also often controlled vegetative growth ( conidial germination , colony growth , pigmentation and mycelial appearance ) of the fungus ( Figure 2C ) . Interestingly , 7 TF genes required for pathogenicity had multiple mutant phenotypes ( Figure 2C , Table 1 ) , and this phenomenon implied that the fungal pathogenic process is complex , where it is closely linked with many fungal development stages , and that the pathogenicity-related Zn2Cys6 TF genes are also involved in other developmental processes . We identified the homologs of the 104 M . oryzae Zn2Cys6 TF genes in N . crassa and F . graminearum by Blastp ( www . broadinstitute . org ) , and then compared the phenotypes of the knockout mutants with the homologs studied in N . crassa [29] or F . graminearum [27] ( Table S5 in Text S1 ) . Deletion of members of the Zn2Cys6 TF family in F . graminearum and N . crassa resulted in mutant phenotypes at 16% ( 46/296 ) and 42% ( 30/72 ) of the mutants [27] , [29] . However , more mutants ( 58 . 7% ) in M . oryzae displayed visible mutant phenotypes . But only two mutants showed defects in growth and asexual and/or sexual development in three species simultaneously ( Table S5 in Text S1 ) . This comparison of mutant phenotypes between the three fungi suggested that the large majority of Zn2Cys6 TF genes evolved to have a unique , not conserved function in regulating fungal development . The null mutants of 27 Zn2Cys6 TFs showed significant differences in fungal growth with the wild-type strain ( Figure 2B , Table 1 ) . Among these null mutants , three mutants ( 01C9-1/ΔMGG_07063 , 01F1-1/ΔMGG_17841 and 02G4-1/ΔMGG_07149 ) showed reduced colony growth on CM medium at 24 . 7 , 26 . 6 and 43 . 5% , respectively ( Table S4 in Text S1 ) . The TF genes deleted in these mutants were designated GCC1 ( growth , conidiation and cell wall regulatory factor 1 , MGG_07063 ) , GPF1 ( growth and pathogenicity regulatory factor 1 , MGG_17841 ) and GTA1 ( growth and tolerance to acidic stress regulatory factor 1 , MGG_07149 ) , respectively . To confirm the defects of the mutants in growth were caused by the knockout of the TF genes , we complemented the three mutants Δgcc1 , Δgpf1 and Δgta1 with their respective native copies from wild-type strain 70-15 ( Figure S4C ) . The phenotypic analyses showed the GCC1-rescued strain ( gcc1-c ) and GTA1-rescued strain ( gta1-c ) all recovered from their defects in growth , and GPF1-rescued strain ( gpf1-c ) also recovered most defects in growth when compared to the mutants and the wild-type strain ( Table 2 , Figure 3A ) . These results implied that Zn2Cys6 TFs GCC1 , GPF1 and GTA1 were required for the regulation of fungal growth in the rice blast fungus . Among the 104 Zn2Cys6 TFs , the null mutants of 25 genes displayed distinct changes in asexual development ( Figure 2B , Table 1 ) . Of these mutants , 21 mutants produced less conidia or even no conidia , but the other 4 mutants produced more conidia ( Table 1 , Table S4 in Text S1 ) . We selected 2 mutants without conidiation , 4 mutants with less conidiation and 2 mutants with more conidiation to be complemented by their native genes to show the relationship between mutant's phenotype and TF gene deleted ( Figure S4C ) . The genes deleted in these null mutants were CNF1 ( conidial production negative regulatory factor 1 , MGG_02962 in mutant 06B5-7 ) , CNF2 ( conidial production negative regulatory factor 2 , MGG_15023 in mutant 03C7-2 ) , PCF1 ( pathogenicity and conidiation regulatory factor 1 , MGG_17623 in mutant 03C11-4 ) , MoNIT4 ( a homolog of the nit-4 gene , which is a pathway-specific regulatory gene that mediates nitrate induction in N . crassa [38] , MGG_01518 in mutant 01A7-1 ) , CCA1 ( conidiation , conidial germination and appressorium formation required transcription factor 1 , MGG_05659 in mutant 06C2-14 ) , CONx1 ( conidiation required transcription factor x1 , MGG_12349 in mutant 06D5-21 ) , GCC1 ( MGG_07063 in mutant 01C9-1 ) and Mocod1 ( homolog of amyR from Aspergillus nidulans [3] , [39] , [40] , MGG_05343 in mutant 02F2-6 ) . The mutants Δcnf1 and Δcnf2 produced more conidia than did the wild-type; mutants Δpcf1 , Δcca1 and Δconx1 produced less conidia; and mutants Δgcc1 and ΔMocod1 did not produce any conidia . The complementation experiments displayed that the defects of mutants in conidiation were caused by the deletion of the TF genes , since the complemented strains of these mutants all recovered from their abnormal asexual developmental phenomena ( Table 2 , Figure 3B ) . CNF1 is a strong conidial production negative regulatory factor in rice blast fungus , and its deletion may result in greatly increased conidial production ( usually 20- to 40-fold of the wild-type strain ) . Moreover , the mutant Δcnf1 had significant changes in mycelial appearance , such as deep dark colony color from the gray-white of the wild-type strain , short spore-bearing aerial hyphae from long aerial hyphae of the wild-type strain , and spore-bearing aerial hyphae formed at the rim of mycelial colony earlier than in the wild-type strain ( Figure 4A ) . Although the conidiophore development of Δcnf1 seemed similar to that of the wild-type strain and CNF1-rescued mutant cnf1-c ( Figure 3B , Figure 4C ) , more aerial hyphae differentiated into conidiaophores in Δcnf1 ( Figure 4B ) . The increase in conidial production of Δcnf1 was possibly partly due to the spore-bearing hyphae differentiating earlier and more from the aerial hyphae than in the wild-type strain . Δcca1 produced few and odd , long and vacuolated conidia ( Table 2 , Figure 5 ) . Single or two tandem spores differentiated from the apex of conidiophores in the mutant Δcca1 , and the CCA1-rescued mutant cca1-c produced normal conidiophores and conidia similar as the wild-type strain ( Figure 4C , Figure 5 ) . These results suggested that CCA1 is required for the differentiation of conidiophores and the formation of conidia . When the null mutants of 101 Zn2Cys6 TFs were assayed for their conidial germination rate at 4 hpi ( hour post inoculation ) and their appressorium formation rate at 24 hpi , 12 mutants and 10 mutants displayed significant differences in conidial germination and appressorium formation , respectively ( Figure 2B , Table 1 ) . We assayed conidial germination and appressorium formation in the complemented strains of 6 null mutants ( Δfzc16 , Δgpf1 , Δgta1 , Δpcf1 , Δcnf1 and Δcca1 ) ( Figure S4C ) and found that these complemented strains all recovered from their defects in conidial germination or in appressorium formation ( Table 2 ) . These experiments implied that Zn2Cys6 TFs GPF1 , PCF1 , CNF1 , and CCA1 are required for conidial germination and appressorium formation and that FZC16 and GTA1 are required for conidial germination in the rice blast fungus . Among these 6 mutants , the mutant Δcca1 showed marked defects not only in conidial germination and appressorium formation , but also in conidiation ( Table 2 ) . The conidial germination rate of the mutant Δcca1 was 3 . 6% , and appressorium formation rate was 10 . 2% ( Table 2 ) . The structure and shape of conidium , germinated conidium and appressorium of the mutant Δcca1 displayed notable differences with the wild-type strain ( Figure 5 ) . Therefore , the mutant Δcca1 produced few multifarious and vacuolated conidia , which germinated and formed not fully melaninized appressoria at a low ratio . The pathogenicity of 104 null mutants was tested first on intact barley leaves by placing agar plugs containing mycelia on them . Mutant Δgpf1 ( MGG_17841 ) was found to lose pathogenicity to barley , and 4 mutants Δcnf1 ( MGG_02962 ) , Δcnf2 ( MGG_15023 ) , ΔMocod1 ( MGG_05343 ) and Δpcf1 ( MGG_17623 ) showed reduced pathogenicity ( Table S4 in Text S1 , Figure S2 ) . We then determined the pathogenicity of 101 mutants on rice seedlings ( CO-39 ) by spraying with conidial suspension ( 1×105 spores/ml ) . The results showed that it was decreased in 3 mutants Δcnf1 ( MGG_02962 ) , Δcnf2 ( MGG_15023 ) and Δgta1 ( MGG_07149 ) and that 2 mutants Δgpf1 ( MGG_17841 ) and Δcca1 ( MGG_05659 ) lost their virulence to rice ( Table S4 in Text S1 , Figure S3A ) . Impressively , Δcca1 caused lesions on barley leaves when inoculated with mycelial plugs , but lost virulence on rice seedlings when spayed as a conidial suspension . The other 3 mutants Δconx1 ( MGG_12349 ) , Δgcc1 ( MGG_07063 ) and ΔMocod1 ( MGG_05343 ) with little or no conidiation were tested on intact rice leaves with agar plugs containing mycelia . Δconx1 and ΔMocod1 were found to have lower virulence ( Table S4 in Text S1 , Figure S3B ) . After complemented with native genes ( Figure S4C ) , we retested the pathogenicity of the 7 mutants with their complemented strains on rice seedlings by spraying with conidial suspensions ( Δgpf1 , Δcnf1 , Δcnf2 , Δgta1 , Δcca1 and gpf1-c , cnf1-c , cnf2-c , gta1-c , and cca1-c ) ( Figure 6A ) or on rice leaf explants by inoculating with mycelial plugs ( ΔMocod1 , Δconx1 and Mocod1-c , conx1-c ) ( Figure 6B ) . The facts that the complemented strains recovered from their defects in virulence on rice implied that the loss or weakening of virulence in the mutants was caused by the deletion of the TF genes GPF1 , CNF1 , CNF2 , GTA1 , CCA1 , CONx1 and MoCOD1 , respectively . To assay the roles of these TF genes in plant penetration , onion cuticle and barley leaves were inoculated with a conidial suspension of the mutants and their complemented strains . Compared with the wild-type strain , the ability of Δgpf1 , Δgta1 , Δcnf2 , Δcnf1 and Δcca1 to penetrate the onion cuticle and barley leaf cuticle was weakened to some extent ( Figure 6C ) . Notably , the mutants Δgpf1 and Δcca1 were unable to penetrate barley cuticle , even after barley cuticle was abraded ( for Δgpf1 ) , while the wild-type strain penetrated into the epidermal cells of barley leaves and grew invasively after 24 hpi ( Figure 6C , Figure 6D ) . These results suggested that GPF1 is required by the rice blast fungus for penetration into the plant cuticle and possibly for invasive growth . Although the conidia of Δcca1 were unable to penetrate barley cuticle and could not cause lesions in rice , the mycelia of Δcca1 could cause lesions in barley and rice . The difference in virulence between conidia and mycelia may lie in the conidial defects of Δcca1 in germination , appressorium formation and host penetration ability . There are various stress conditions encountered by M . oryzae when it completes its infection cycle in rice . It is possible to predict the biological functions of these deleted Zn2Cys6 TF genes in cell activity regulation by assaying the resistance ability to stress conditions of the mutants . We screened and assayed in a rapid way the mycelial growth of the mutants under 9 abiotic stress conditions ( 0 . 5 mM H2O2 , 0 . 005% SDS , 0 . 3 M CaCl2 , 0 . 5 M NaCl , 1 M sorbitol , olive oil as a single organic carbon source , pH 5 . 0 and pH 9 . 0 , and minimal medium ) . Four mycelial inoculation blocks were placed in a 9-cm plate by needle point inoculation and the diameter of mycelial colonies were measured after 6 days culture . The results displayed that many mutants had significant visible phenotypes under those stress conditions when compared with the wild-type strain ( Table S4 in Text S1 ) : 14 mutants showed mutant phenotypes in MM medium ( nitrogen metabolism ) , 16 mutants in CM-C medium containing olive oil ( carbon metabolism ) , 22 mutants in CM medium with SDS ( cell wall stress ) , 17 mutants or 19 mutants in CM medium with high concentration of Ca2+ or Na+ ( ionic regulation ) , 18 mutants in CM medium with 1 M sorbitol ( hypertonic pressure ) , 30 mutants in CM medium with 0 . 5 mM H2O2 ( oxidative stress ) , and 30 mutants or 14 mutants in CM medium at pH 5 . 0 or pH 9 . 0 ( ambient pH regulation ) . After comparison of the relative growth rate of mutants on CM medium , we selected 5 mutants ( 01A7-1 , 01B11-1 , 01C9-1 , 01F1-1 and 02G4-1 ) that had significant changes in mycelial growth under stress conditions to be complemented with native genes to see if these phenotypes were caused by the deletion of TF genes ( Figure S4 ) . The TF genes in these 5 mutants were: MoNIT4 in mutant 01A7-1 ( MGG_01518 ) , GPF1 in mutant 01F1-1 ( MGG_17841 ) , GCC1 in mutant 01C9-1 ( MGG_07063 ) , TAS1 ( tolerance to acidic stress regulatory factor 1 ) in mutant 01B11-1 ( MGG_04108 ) and GTA1 in mutant 02G4-1 ( MGG_07149 ) . The mycelial growth of ΔMonit4 in minimal medium , Δgpf1 in CM-C medium containing olive oil , and Δtas1 at pH 5 . 0 was reduced ( Figure 7A ) . However , the mycelial growth of Δgcc1 in CM medium containing SDS and Δgta1 at pH 5 . 0 was increased . When rescued with native genes , these mutants all recovered their growth under stress conditions ( Figure 7A ) . Interestingly , the mutant ΔMonit4 grew normally as the wild-type strain on CM medium ( Table S4 in Text S1 , Figure S1 ) , but grew slowly as a very sparse and thin mycelium layer on MM medium ( Figure 7B ) . As previously reported , MoNIT4 is necessary for the regulation of nitrogen metabolism [38] , [41] . These data also suggested that TF gene GPF1 is necessary for gene expression regulation relative to fat metabolism , that GCC1 is necessary for gene expression regulation relative to cell wall synthesis , and that TAS1 and GTA1 are necessary for gene expression regulation in response to acidic stress . Eight Zn2Cys6 TF genes ( CCA1 , CNF1 , CNF2 , CONx1 , GPF1 , GTA1 , MoCOD1 and PCF1 ) are required for fungal pathogenicity to rice or barley . Δgpf1 and Δcnf2 are two mutants that have similar phenotypes on virulence with relatively limited differences in other fields , because ΔMocod1 and Δconx1 do not produce conidia , Δcca1 and Δgta1 have virulence when inoculated on rice leaves with mycelial plugs , Δpcf1 shows reduced virulence on barley inoculated with mycelial plugs , but normal when sprayed on rice , and Δcnf1 produces greatly increased conidia . To understand how these TF genes affect fungal pathogenicity , the two genes GPF1 and CNF2 , whose mutants lose ( Δgpf1 ) or have reduced ( Δcnf2 ) virulence to both barley and rice , were selected to analyze their functions in gene expression regulation . Since Δgpf1 could not penetrate rice cuticle , we could not obtain its mRNA during infection . The rice blast fungus experiences temporal nitrogen starvation when it infects the plant , and the gene expression pattern of the mycelia under starvation is similar to that at plant infection [42]–[45] , and the hyphae enduring starvation are similar to those inoculated as mycelial plugs in plant infection . We therefore analyzed the genome-wide gene expression of the mycelia after 4 h starvation of the wild-type strain , Δgpf1 and Δcnf2 by RNA-seq . The RNA-seq experiments were performed in biologic triplicate for each strain . The transcripts of 10 , 864 , 11 , 048 and 10 , 681 genes were identified in the starved mycelia of the wild-type strain , Δgpf1 and Δcnf2 , respectively . Compared with the wild-type strain , 2641 genes were differentially expressed significantly in Δgpf1 ( FDR<0 . 05 ) , with 1406 genes up-regulated and 1235 down-regulated ( Table S6 in Text S1 ) , and 3144 genes in Δcnf2 ( FDR<0 . 05 ) , with 1668 genes up-regulated and 1476 down-regulated ( Table S7 in Text S1 ) . The similarities and differences between genes regulated by GPF1 and CNF2 were analyzed in detail by comparing the DEGs in Δgpf1 or Δcnf2 . The comparison assays showed that the expression of 611 genes or 1114 genes were independently regulated by GPF1 or CNF2 , while 2030 genes were regulated by both GPF1 and CNF2 together ( Figure 8 ) . Among the 2030 genes commonly regulated by GPF1 and CNF2 , 916 genes were down-regulated and 1105 genes up-regulated simultaneously in both Δgpf1 and Δcnf2 , while only 9 genes were regulated in the opposite direction in both mutants ( Figure 8 , Table S8 in Text S1 ) . It is very surprising that the expression of so many DEGs were regulated in the same direction in Δgpf1 and Δcnf2 ( Figure 8 ) . The correlation coefficient between DEGs of Δgpf1 and Δcnf2 was determined by the linear trend model , and the result showed R2 = 0 . 88 with p<0 . 0001 . These similar patterns were not mainly caused by the starvation , as we also assayed the mutants of five TF genes containing non-Zn2Cys6 domains by RNA-seq with the same treatment at the same time , but the mutants displayed specific DEG patterns different from Δgpf1 and Δcnf2 . Therefore , these similar patterns in gene expression regulation in two mutants were possibly caused by the fact that TFs Gpf1 and Cnf2 had the same DNA-binding domain ( Zn2Cys6 domain ) , and Δgpf1 or Δcnf2 had similar phenotypes in pathogenicity . However , GPF1 and CNF2 still showed great differences in gene regulation since 1152 DEGs appeared in Δcnf2 when compared with Δgpf1 ( Table S9 in Text S1 ) . To see how GPF1 and CNF2 affect the pathogenicity to plants , we reviewed the functions of the DEGs which were studied in previous reports . Fifty DEGs in Δgpf1 and 68 DEGs in Δcnf2 ( total of 80 DEGs in both were studied by knockout . Of the 80 genes , the mutants of 57 genes ( 36 DEGs in Δgpf1 and 49 DEGs in Δcnf2 ) displayed defects in virulence ( Table 3 ) but the mutants of 23 genes ( 14 DEGs in Δgpf1 and 19 DEGs in Δcnf2 ) were dispensable for fungal pathogenicity ( Table S10 in Text S1 ) . Interestingly , of these reported pathogenicity-required genes , the number of down-regulated genes ( 45 DEGs ) was much higher than that of up-regulated genes ( 12 DEGs ) in the two mutants Δgpf1 and Δcnf2 ( Table 3 ) . These pathogenicity-required genes are mainly related to transcription ( such as CON7 , MoLDB1 , MoMIG1 , MoPAC2 , MSTU1 , MST12 and PTH12 ) [8] , [12] , [18] , [20] , [21] , [46] , [47] , G-protein signaling ( such as MAGB , MoRGS1 and MGB1 ) [48]–[51] , MAPK pathway ( such as MST7 , MoMCK1 and MoMPS1 ) [52]–[54] , autophagy ( such as MoATG1 , MoATG5 , MoATG9 and SNX41 ) [55]–[58] , regulation of reactive oxygen species ( such as MoHYR1 and GTR1 ) [59] , [60] , and amino acid , lipid and carbohydrate metabolism ( such as ILV2 , MoSNF1 , ICL1 and MoPLAA ) [61]–[64] . We also checked the relationships between Zn2Cys6 TFs and Gpf1 or Cnf2 in transcription regulation . Surprisingly , 60 or 66 Zn2Cys6 TF genes were listed as DEGS of Δgpf1 or Δcnf2 ( total of 79 genes in two mutants ) ( Table S11 in Text S1 ) . The Zn2Cys6 TF genes are also the important targeted genes regulated by two Zn2Cys6 TF genes GPF1 and CNF2 . These facts implied that the well-balanced and sophisticated expression regulation of TF genes and pathogenicity-required genes is necessary for the maintenance of pathogenicity in M . oryzae .
Since the genomes of a large number of fungi have been sequenced and are being sequenced , there is a surge of interest in functional genomics research through the systematic mutagenesis of identified genes . The construction of a genome-wide gene deletion mutant set of fungi based on the homologous recombinational gene knockout procedure is a valuable resource for the analysis of fungal development , pathogenicity , and many aspects of cell biology and biochemistry , such as those done in the budding yeast S . cerevisiae [65] , the fission yeast Schizosaccharomyces pombe [66] , and the saprobe filamentous fungus N . crassa [29] . However , except for N . crassa , F . graminearum and Aspergilli which are highly efficient in gene deletion experiments , no efficient system has been available to perform high-throughput gene knockout in filamentous fungi . In this study , we present an approach to knockout large numbers of genes that utilizes several methods suitable to high-throughput manipulation . Several time-consuming steps of the gene knockout procedure could be performed in a high-throughput way . In this procedure , the gene-deletion cassettes were built using a yeast homologous recombination method [29] with a yeast-Escherichia-Agrobacterium shuttle vector pKO1B , which is the first reported artificial plasmid that could be replicated in yeast , E . coli and Agrobacterium cells . The gene-deletion cassettes obtained in a binary vector pKO1B could be directly used to transform fungal cells through the ATMT method [67] . This advantage avoids the tedious and inefficient work to transfer the gene-deletion cassettes from a yeast plasmid to another binary Agrobacterium plasmid . The use of GFP fluorescence as a negative marker to eliminate most ectopic insertion transformants reduced the number of transformants one has to identify by PCR or Southern blot . The employment of the negative ( for the targeted gene ) /positive ( for the unique recombinational DNA ) identification PCR and qPCR to identify null mutants instead of Southern blot , which is not high throughput , makes the identification of null mutants a high-throughput procedure . The copies of the gene-deletion cassette in mutant genomic DNA quantified by qPCR has been widely performed in animals and plants [68] , [69] . The reliability and stability of PCR or qPCR were guaranteed by the extraction of high quality genomic DNA with CTAB [70] performed in a high-throughput way in this study . The result of the gene knockout events ( Figure S4A ) were also reconfirmed by Southern blot at the DNA level in 10 mutants ( Figure S4B ) and at the transcript level in 16 mutants ( Figure S4C ) . As a result , two researchers were able to complete one cycle of the gene knockout experiment for 96 genes in a month , including preparing knockout vectors , performing ATMT transformation , and identifying null mutants , which is a major improvement over the individual gene knockout protocol . More importantly , this procedure can be adapted for knocking out genes in other fungi without any modification or after the substitution of organism-specific promoter to drive the GFP reporter . Besides the GFP gene , a second drug resistance gene ( HPH/hygromycin B phosphotransferase gene , NEO/neomycin phosphotransferase II gene or BAR/glufosinate resistance gene , etc . ) or a herpes simplex virus thymidine kinase ( HSVtk ) gene , which converts 5-fluoro-2′-deoxyuridine to a toxic compound [71] , could be used as an alternative negative selection marker against ectopic transformants . As an alternative vector , pKO1B-HPH ( shown in Materials and Methods ) is another yeast-Escherichia-Agrobacterium shuttle which could be used in this high-throughput gene knockout system besides pKO1B . To increase knockout efficiency , M . oryzae KU80 null mutants , which are defective for non-homologous end joining ( NHEJ ) DNA repair [31] , could be alternative strains used in targeted gene replacement instead of the wild-type strain . In this study , we generated 104 fungal-specific Zn2Cys6 TF gene-deleted mutants in M . oryzae by a high-throughput gene knockout procedure and analyzed the phenotypes of individual TF mutants . The deletion of Zn2Cys6 TF genes resulted in phenotype changes in fungal development and pathogenicity in 58 . 7% mutants compared to the wild-type strain , while 26 mutants were defective in pleiotropic phenotype . Colony growth and asexual reproduction ( conidiation ) were the two phenotypic categories most observed in mutants , and most mutants defective in conidiation were often defective in vegetative growth ( conidial germination , colony growth , pigmentation and mycelial appearance ) . Seven Zn2Cys6 TF genes functional for pathogenicity in M . oryzae are also especially required for vegetative growth , conidiation or appressorium formation . Comparison of phenotypes between M . oryzae , F . graminearum and N . crassa Zn2Cys6 TF gene orthologs seemed to agree with the previously reported viewpoint that Zn2Cys6 TFs evolved divergently in how to regulate fungal growth and asexual development , rather than keeping the same function in different fungi [27] . These divergences in functions may be due to the differences in the life cycles between the three fungi . N . crassa is an obligate saprophyte which lives on dead organic material and cannot attack a living host , while M . oryzae and F . graminearum are facultative saprophytic plant pathogens that additionally need specialized structures ( such as an appressorium ) and functions to infect and obtain nutrients from living plants . Until now , 6 Zn2Cys6 TF genes ( MoCOD1 , MoCOD2 , MoNIT4 , PIG1 , TRA1 and XLR1 ) have been identified in the rice blast fungus [17] , [39] , [41] , [72] , [73] , and 4 of them were also in our Zn2Cys6 TF gene knockout mutant set . The expression of MoNIT4 and MoCOD1 were up-regulated during conidiation [41] , and their mutants ΔMonit4 and ΔMocod1 had reduced conidiation [39] , [41] , and ΔMocod1 also had lower pathogenicity [39] . In our study , the mutants ΔMonit4 and ΔMocod1 were also defective in conidiation or pathogenicity; furthermore , our mutant ΔMocod1 did not produce any conidia on CM medium . Similar to a previous report [74] , our ΔTra1 mutant also showed defects in conidial germination . PIG1 was identified to be involved in melanin biosynthesis , but not confirmed by knockout [73]; however , the Δpig1 mutant did not show notable phenotypes in fungal development in our study . These small discrepancies in mutant phenotypes with previously reported data may be due to the differences in wild-type strains and experimental conditions . TFs Mnh6 , Moatf1 , Mocrz1 and Mstu1 were reported to regulate hyphal growth in the rice blast fungus , and the colony growth of the mutants Δmnh6 , ΔMoatf1 , ΔMocrz1 and ΔMstu1 was about 70–90% of the wild-type strain [7]–[9] , [24] . We found 27 Zn2Cys6 TF genes involved in fungal growth , and 3 genes GCC1 , GPF1 and GTA1 were required for normal colony growth . Interestingly , the colony growth of Δgta1 was only about 56% of that of wild-type strain , and GTA1 is a gene which affects colony growth the most among known TF genes in the rice blast fungus . GTA1 also functions in conidiation and pathogenicity to plants . The change in asexual reproduction in a mutant is a phenotype that mostly happens when TF genes or other genes are deleted in the rice blast fungus . The TF gene-deleted mutants ( Δcom1 , Δcon7 , Δcos1 , Δmnh6 , ΔMohox2 ) showed reduced conidiation [7] , [11]–[13] , [15] , [16] . In our study , 25 Zn2Cys6 TF genes were identified as being involved in fungal conidiation . Of them , CCA1 , GCC1 , MoCOD1 and CONx1 are necessary for the differentiation of conidiophores . Unexpectedly , the deletion of four Zn2Cys6 TF genes ( CNF1 , CNF2 , CNF3 and CNF4 ) led to increased conidial production . In particular , Cnf1 is the strongest negative regulatory factor of conidial production identified in the rice blast fungus until now . The deletion of CNF1 led to earlier and more differentiation of spore-bearing hyphae than in the wild-type strain . Impressively , the mutants Δcnf1 and Δcnf2 produced more conidia , but their conidia had reduced virulence to barley and rice . Conidia are the main way to spread the rice blast disease . It is clear that a balance exists between the ability to produce conidia and the pathogenicity to plant in the phytopathogenic fungus . We identified 2641 genes regulated by GPF1 and 3144 genes by CNF2 through RNA-seq . Interestingly , 2021 genes were regulated in the same direction in Δgpf1 and Δcnf2 . This fact suggested GPF1 and CNF2 have similar mechanisms in the regulation of fungal pathogenicity . Nearly sixty DEGs in Δgpf1 and Δcnf2 were confirmed to be required for fungal pathogenicity ( Table 3 ) . In particular , the expression of two pathogenicity-required genes , an APSES TF gene MSTU1 [8] and a hydrophobin gene MPG1 [75] , were greatly down-regulated in Δgpf1 ( 7 . 80- and 12 . 49- fold , respectively ) and Δcnf2 ( 11 . 85- and 15 . 83-fold , respectively ) . About half of Zn2Cys6 TF genes and many other DNA-binding domain TF genes were regulated by GPF1 and CNF2 . These data primarily revealed the gene expression network in the regulation of pathogenicity controlled by GPF1 and CNF2 . It is necessary to keep the well-balanced expression of TF genes and pathogenicity-required genes for maintenance of pathogenicity in M . oryzae . The high-throughput analysis of mutant phenotypes showed that some TF gene mutants shared highly overlapping phenotypes , such as mycelial growth in Δgcc1 , Δgpf1 and Δgta1 , and pathogenicity in Δgpf1 , Δgta1 , Δcnf1 and Δcnf2 . On the other hand , several mutant phenotypes were shared by one TF gene mutant . Also , it is fascinating to know how to regulate similar phenotypes by different TF genes . In general , every TF gene regulates the expression of many downstream genes and the expression of every gene is regulated by many TF genes , and therefore each mutant phenotype is controlled by the changes in expression of several genes . When two TF genes show similar mutant phenotype , there are at least 3 types of phenotype regulation mechanisms between them . First , TF gene A regulates another TF gene B , and TF gene B continues to regulate a group of downstream genes . Second , TF genes A and B commonly regulate a group of downstream genes . Third , TF genes A and B regulate two different groups of downstream genes independently , but they lead to similar phenotypes . However , two or three of these gene regulation types often occur simultaneously . For example , GPF1 and other TF genes were down-regulated in Δcnf2 , while CNF2 and other TF genes were also down-regulated in Δgpf1 ( Table S11 in Text S1 ) . A common group of genes ( Figure 8 ) were regulated both by CNF1 and GPF1 which led to similar pathogenicity phenotypes in mutants , while two other different groups of genes ( Figure 8 ) were regulated by CNF1 and GPF1 independently which led to different phenotypes in conidiation in mutants . In conclusion , this study represents a major advance in a high-throughput gene knockout system suitable for filamentous fungi , provides the functional characterization of 104 Zn2Cys6 TF genes in the rice blast fungus , and reveals gene expression patterns of two virulence-required TF genes GPF1 and CNF2 . These studies will help us build more fungal gene-deletion mutant libraries in an economical way and uncover the transcriptional network in fungi and fungal pathogenic mechanisms .
M . oryzae strain 70-15 and its mutants , S . cerevisiae strain FY834 , and E . coli strain DH5α and A . tumefaciens strain AGL1 were used in this study . A 1184-bp promoter fragment of M . oryzae H3 histone gene amplified from pKD5 [35] with primers H3sF and H3SR ( Table S1 in Text S1 ) and 720-bp eGFP CDS fragment amplified from pEGFP ( Clontech , USA ) with primers KoGFPF and KoGFPR ( Table S1 in Text S1 ) were inserted into the XhoI/EcoR1 sites and the EcoR1/SacI sites of the binary vector pCAMBIA1300 ( Cambia , USA ) to produce the vector pKO1 . pKO1B was then produced by inserting a 2 . 9-kb URA3-2micro2_origin fragment amplified from pYES2 ( Invitrogen , USA ) with primers uraf and urar ( Table S1 in Text S1 ) into the SacII site of pKO1 by a yeast recombinational cloning method using S . c . EasyComp Transformation Kit ( Invitrogen , USA ) . A list of 2-kb regions on both sides of each ORF ( 12775 ORF totally ) was retrieved from the rice blast fungus database MG8 ( www . broadinstitute . org ) and saved in Excel type files by a program written by Mr . Tan Cheng . 23-nt primers ( Table S1 in Text S1 ) of each gene-specific flank ( 1000 bp–1500 bp in length and 1200 bp in optimum length ) were designed by the BatchPrimer3 program [76] . For each gene , primers were designed and synthesized with the following common 33-nt 5′ regions: 5f:GCTGTACAAGTAAGAGCTCGGTACCCGGGGATC… ( Homologous to pKO1B ) 5r:CCGGGAGATGTGGGGCACTGTGGCGTTGGCACA… ( Homologous to SUR gene ) 3f:TTGATTATTGCACGGGAATTGCATGCTCTCACA… ( Homologous to SUR gene ) 3r:TTAAGTTGGGTAACGCCAGGGTTTTCCCAGTCA… ( Homologous to pKO1B ) The flank fragments were produced from genomic DNA of M . oryzae strain 70-15 using Primer Star or EX Taq ( TaKaRa , China ) in 96-well PCR plates . The SUR cassette fragment was generated by PCR with primers surf and surr ( Table S1 in Text S1 ) from pBS-SUR [35] . All PCR products were verified by agarose gel electrophoresis analysis and then used in subsequent steps without further purification . Among the 163 TF genes , both flanking fragments were successfully amplified for 142 genes . However , one or sometimes both flanking fragments failed to be amplified for 21 genes even after trying several reaction conditions . The yeast transformation procedure was conducted following a small-scale yeast transformation protocol in the pYES2 user manual ( Invitrogen , USA ) in 96-well deep well plates as described by Colot et al [29] . The cocktail mixture was made by adding 1 . 8 ml competent yeast cells , 100 µl linearized pKO1B by XbaI and HindIII ( 100 ng/µl ) , 210 µl PCR production of SUR cassette , 60 µl denatured salmon sperm DNA ( Sangon , China ) and 2 . 6 ml DMSO to 20 . 7 ml freshly prepared 1×LiAc/40%PEG-3350/1×TE solution ( 100 mM lithium acetate , pH 7 . 5; 50% PEG-3350; 10 mM Tris-HCl , pH 7 . 5; 1 mM EDTA ) . Next , 200 µl of the mixture , followed by 4 µl of 5′ and 3′ flank PCR fragments of the targeted gene , were pipetted into each well of a 96-deep-well plate with a multichannel pipette . The plates were sealed and processed further following the user manual . After 3 days , the yeast cells cultured in SC-Ura liquid medium were collected and disrupted in Fastprep-24 homogenizer ( MP , USA ) at 4 . 0 m/s for 2 min . The plasmids were extracted with TIANprep yeast plasmid DNA kit ( Tiangen Biotech , China ) following the user manual . Next , the yeast plasmids were transformed into the competent cells of E . coli strain DH5α prepared following the Inoue method for “ultra-competent” cells [77] . Four bacterial colonies on each LB plate were placed in the wells of 96-deep-well plates , with each well containing 1 ml of LB liquid medium with 50 µg/ml kanamycin . The plates were shaken at 220 rpm overnight at 37°C . The correctness of homologous recombinational cloning for gene-deleted cassettes was confirmed by bacterial double PCR . Common primers were designed to amplify the 5′ and 3′ flanking fragment of each target gene-deleted cassette . Forward primer KO1Bf1 for the 5′ flanking fragment and reverse primer KO1Br2 for the 3′ flanking fragment were located inside pKO1B , and reverse primer SURr1 for the 5′ flanking fragment and forward primer SURf2 for the 3′ flanking fragment were located in SUR cassette ( Table S1 in Text S1 ) . The resulting PCR products in knockout plasmids constructed correctly would be 1 . 0–1 . 5 kb for 5′ flanking fragment and 1 . 5–2 . 0 kb for 3′ flanking fragment . The gene deletion cassettes were further confirmed by sequencing with the primers SURf2 and SURr1 . PCR and DNA sequencing results showed a success rate of 100% for the recombinant plasmids ( Table S2 in Text S1 ) . We found that it was enough to confirm the recombinant plasmids by PCR and not necessary to confirm by sequencing plasmids . The correctly built knockout plasmids were extracted from the bacterial cultures using the AxyPrep-96 plasmid purification kit ( Axygen , China ) following the protocol in the user manual . The knockout plasmids were transformed into the competent cells of A . tumefaciens strain AGL1 using the freeze/thaw shock transformation method following the procedure described elsewhere [67] . The plasmids in Agrobacterium cell were confirmed by culture PCR with primer set SurP1 and SurP2 ( Table S1 in Text S1 ) for a 368-bp SUR gene fragment . M . oryzae strain 70-15 was grown on CM medium for 12 days at 25°C under constant fluorescent light , and the conidia were harvested and transformed with the knockout plasmids mediated by A . tumefaciens according to a previously reported procedure [67] , whereas performed in groups , usually 24 or 48 genes in a batch each time . The nitrocellulose membrane strips containing the conidia co-cultivated with A . tumefaciens were placed on AIM medium in the dark at 22°C for 2 days and then transferred onto the selection defined complex medium ( DCM; 0 . 17% yeast nitrogen base without amino acids , 0 . 2% asparagine , 0 . 1% ammonium nitrate and 1% glucose , pH 6 . 0 with Na2HPO4 ) plates containing 100 µg/ml sulfonylurea , 50 µg/ml kanamycin , 400 µg/ml cefotaxime and 60 µg/ml streptomycin . Sulfonylurea-resistant transformants grown on selection medium were individually transferred onto new selection DCM plates using sterile toothpicks . A total of 8741 primary transformants corresponding to 133 Zn2Cys6 TF genes were selected on the basis of sulfonylurea resistance ( Table S2 in Text S1 ) . No transformant was obtained for the other nine genes . After culture for 2 days , a little mycelium of each transformant was picked out and placed onto a glass slide . We usually placed the mycelia of six transformants on a slide in proper order . The green fluorescence emitted by transformants was then observed one by one under a fluorescence microscope . Ectopic transformants emitted green fluorescence and null mutants did not when excited under the fluorescence microscope ( Figure 1D ) . The transformants without green fluorescence were picked out and inoculated on a new selective plate and grown for 3 days . In total , 3191 transformants without green fluorescence were screened from the primary transformants . Extraction of genomic DNA was performed following the CTAB protocol of Rogers and Bendich [70] with modifications . A small piece ( >9 mm2 ) of mycelium along with some medium in the selective plate was transferred to a 2-ml round-bottom tube , and 400 µl ddH2O were added along with 0 . 1 g porcelain beads . The mycelial cells were disrupted in a Fastprep-24 homogenizer at 4 . 0 m/s for 2 min; 400 µl 4×CATB buffer ( 4% CTAB , 100 mM Trisma base , 20 mM EDTA , 1 . 4 M NaCl ) were then added to each tube and the tubes incubated at 65°C for 30 min . Next , genomic DNA was extracted following the normal CTAB protocol [70] . In total , genomic DNA of the 3191 transformants was extracted with the CTAB method in a high-throughput way . The gene-deleted mutants from the transformants without green fluorescence were identified by negative screening double PCR . Double PCR was performed using primers CKF and CKR internal to the targeted gene and primers Tbl-gF and Tbl-gR for the β-tubulin gene ( Table S1 in Text S1 ) . PCR was performed in 25-µl reaction mixtures in 96-well PCR plates: 0 . 15 µl primers Tbl-gF and Tbl-gR ( 20 µM ) , 0 . 5 µl primers CKF and CKR ( 20 µM ) , 2 . 5 µl 10×PCR buffer , 0 . 4 µl dNTP mix ( 25 µM ) , 0 . 3 µl Taq ( 5 U/µl ) , 19 . 5 µl ddH2O and 1 µl genomic DNA . The PCR program was: 94°C 3 min followed by 35 cycles of 94°C for 30 s , 57°C for 30 s and 72°C for 30 s , and a final extension at 72°C for 10 min . The PCR products were detected by 1 . 0% agarose gel electrophoresis . If the targeted gene was deleted in a transformant , there was only one band for β-tubulin with 554 bp in length , appearing as a positive control; otherwise , in ectopic transformants and the wild-type strain there were two bands with one of 300–400 bp ( the targeted gene ) and another of 554 bp ( β-tubulin ) ( Figure 1E ) . After checking 2280 genomic DNA samples , we identified 477 transformants with deletions in a total of 104 TF genes ( Table S2 in Text S1 , Figure S4A ) . No null mutants were found for the other 29 genes after screening all DNA samples . For the transformants identified as null mutants in the negative screening double PCR , a second PCR was performed to verify the gene deletion event . One primer p1 or p4 ( Table S1 in Text S1 ) was limited in the genomic DNA outside the 5′ or 3′ flanking fragment in gene-deletion cassettes , and another primer p2 or p3 ( Table S1 in Text S1 ) was limited in the SUR gene in gene-deletion cassettes . If the targeted gene was deleted , there was a band with 1 . 2–2 . 0 kb in length appearing in the gel . Otherwise , there was no band for the ectopic transformant and wild-type strain 70-15 ( Figure 1F ) . A total of 477 transformants of 104 genes identified by the preceding negative screening PCR were also identified as null mutants in this positive screening PCR ( Table S2 in Text S1 , Figure S4A ) . The copies of transformed gene-deletion cassettes in null mutants were identified by qPCR . The concentration of DNA samples of null mutants was standardized to 25 ng/µl by DNA fluorometry . The primers for the genomic DNA of SUR gene are qSurF and qSurR ( Table S1 in Text S1 ) . The genomic DNA fragment coding for β-tubulin gene ( one copy in the genome ) was selected as a control . The primers of β-tubulin gene are qtblF and qtblR ( Table S1 in Text S1 ) . PCR mixture of 25 µl was prepared: 12 . 5 µl 2×RT-PCR buffer ( SYBR Green , premix Ex Taq , TaKaRa ) , 0 . 5 µl forward primer and 0 . 5 µl reverse primer ( 20 µM ) , 2 µl genomic DNA , and 9 . 5 µl ddH2O . Real-time PCR was performed in a Mastercycler ( Eppendorf , USA ) with the following program: 95°C for 2 min , 40 cycles ( 95°C for 10 s , 60°C for 20 s ) , and ending with a melting curve step . Each sample was repeated three times . The copies of transformed gene-deletion cassettes in null mutants were calculated by comparing with the data of the β-tubulin gene and wild-type strain . The exogenous gene was inserted into the mutant's genomic DNA in an integral multiple number . If the copy number of the selective marker gene was 1 . 0±0 . 2 times the β-tubulin gene in genomic DNA , the mutant was regarded as containing a single insertion of the selective marker gene . As a result , 477 null mutants of 104 genes were identified as single insertion null mutants ( Table S2 and Table S3 in Text S1 ) . The gene deletion events in mutants of ten randomly selected TF genes were reconfirmed by Southern blot ( Figure S4B ) , which was performed according to a previously reported procedure [7] . The mutants were complemented with native gene copies of the wild-type strain 70-15 . First , the pKO1B-HPH was built by the replacement of Ph3-GFP cassette with a HPH gene from pCB1003 [78] in pKO1B . The copies of the complementation genes were then cloned from the genomic DNA of the wild-type strain with the primers listed in Table S1 in Text S1 and were inserted into the XbaI/HindIII sites of pKO1B-HPH by the yeast recombinational cloning method . The constructed complementation plasmids were transformed into the mutants using the ATMT method , and the transformants were screened on selective medium containing 200 µg/ml hygromycin B . The gene-rescued transformants were identified by RT-PCR at the mRNA level ( Figure S4C ) with primers specific for the targeted genes ( Table S1 in Text S1 ) . The phenotypic screening analyses were performed by testing 20–30 knockout mutants and one control ( the wild-type strain ) once in a batch according to previously reported protocols [7] , [79] , [80] . Mutant phenotypes were assayed in triplicate with five replicates for each strain . The values of colony growth , conidiation , conidial germination and appressorium formation of mutants in the different experimental groups were compared after normalization with the wild-type strain 70-15 in the same group . The phenotypes of the mutants were also screened under different stress conditions ( Table S4 in Text S1 ) with 20–30 knockout mutants once in a batch and the wild-type strain in an experiment . Four pinpoint-like mycelia of 9-day-old strains were inoculated in the solid medium of different stress conditions in a 9-cm plate with a space interval between each other and then incubated at 25°C . Colony images were captured and the diameters of the mycelial colonies were recorded at 6 dpi . The growth rates under stress conditions were compared between strains after normalization with the wild-type strain . The phenotypes of the mutants of 5 TF genes and their complementation strains under stress conditions were reassessed . Mycelial blocks of 5 mm from 9-day-old strains were inoculated in the center of solid medium of different stress conditions in 6-cm plates followed by culture at 25°C under constant fluorescent light , along with the strains grown on CM medium as controls . Colony images were captured and the diameters of the mycelial colonies were recorded at 8 dpi . The experiments were performed in triplicate with five replicates for each strain . Growth inhibition rate of each mutant was calculated as the growth inhibition rate = ( colony diameter on CM medium−colony diameter under stress condition ) ÷colony diameter on CM medium×100 . The virulence tests on barley ( Hordeum vulgare ) and rice ( Oryza sativa cv CO39 ) were performed following previously reported protocols [7] . For assays with leaf explants of barley or rice , 5-mm mycelium blocks of mutants ( along with the wild-type strain 70-15 and mock ) were inoculated on the leaf , followed by incubation in a wet box at 25°C for 4 days . For spraying assay on rice seedlings , 4 ml conidial suspension ( 1×105 conidia/ml ) containing 0 . 2% ( w/v ) gelatin were sprayed onto 15–20 rice seedlings between the third and fourth leaf stages using an artist's airbrush . Inoculated plants were placed in a wet box at 25 C for 2 days and then allowed to grow in controlled environment chambers with a photoperiod of 12 h using fluorescent lights for 5 days . Plant penetration by the wild-type strain or mutant appressoria was assayed on leaf explants of barley or onion cuticles according to previously reported protocols [7] . A droplet of conidia ( 5×104 conidia/ml ) was inoculated on barley leaf cuticles or onion cuticles and incubated at 25°C for 24 or 48 h . The barley leaves were then treated in methanol ( overnight , room temperature ) to remove chlorophyll , and fixed in alcoholic lactophenol ( 1 h , 95% alcohol/lactophenol = 2∶1 ) . Appressorium penetration on barley leaf cuticle or onion cuticles was assessed under a microscope . The wild-type strain , Δgpf1 and Δcnf2 were grown in liquid CM medium at 25°C with shaking at 180 rpm for 2 days . The cultures were then collected and incubated in H2O at 25°C for 4 h . The treatments were repeated in triplicate for each strain . Total RNA was extracted from ground mycelia in liquid nitrogen with the RNeasy Plant Mini Kit ( QIAGEN ) , and mRNA was isolated using AMPure XP beads ( Beckman ) . RNA-seq libraries were constructed using NEBNext RNA sample preparation kit ( NEB ) in accordance with the standard low-throughput protocol . Samples were sequenced in a 1×100 nt way on an Illumina Hiseq2500 instrument using the TruSeq PE Cluster Kit v3 - cBot - HS ( Illumina ) and TruSeq SBS Kit v3-HS ( Illumina ) . The clean reads were generated by removing adaptor sequences , tags with>10% “N” , and low quality tags by FastQC ( http://www . bioinformatics . babraham . ac . uk ) , and were mapped to the M . oryzae genome database ( MG8 ) ( www . broadinstitute . org ) using Tophat software [81] . The data from triple biological replicates were then analyzed using Cufflinks software and resulted in quantified genome-wide transcript levels of genes ( expressed in fragments per kilobase of exon model per million mapped fragments – FPKM ) [82] . The significant differences in FPKM between different samples were assayed by the Cuffdiff component of the Cufflinks package [82] . The functions of 104 Zn2Cys6 TF genes were studied by knockout and phenotypic analysis in this study ( Table 1 ) . RNA-sequencing data were deposited in NCBI's Gene Expression Omnibus ( GEO accession number GSE57146 ) .
|
Magnaporthe oryzae is not only the fungus causing the rice blast disease , which leads to 20–30% losses in rice production , but also a primary model pathosystem for understanding host-pathogen interactions . However , there is no high-throughput gene knockout system constructed , and little is known about most of the genes in this fungus . We developed a high-throughput gene knockout system , and using this system , we obtained null mutants of 104 fungal-specific Zn2Cys6 transcription factor ( TF ) genes by screening 8741 primary transformants in M . oryzae . We analyzed the functions of these TF genes in development , pathogenesis , and stress responses under 9 conditions . We found that 61 Zn2Cys6 TF genes play indispensable and diversified roles in fungal development and pathogenicity . CNF1 is the first reported TF gene that strongly and negatively regulates asexual development in the rice blast fungus , and CCA1 , CNF1 , CNF2 , CONx1 , GPF1 , GTA1 , MoCOD1 and PCF1 are required for pathogenicity . We further found via RNA-seq that GPF1 and CNF2 have similar mechanisms in gene expression regulation related to pathogenicity . The resulting data provide new insights into how Zn2Cys6 TF genes regulate important traits during the infection cycle of this rice blast pathogen .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"mycology",
"medical",
"microbiology",
"fungal",
"genetics",
"gene",
"expression",
"genetics",
"microbial",
"pathogens",
"biology",
"and",
"life",
"sciences",
"gene",
"disruption",
"microbiology",
"microbial",
"genetics",
"fungal",
"pathogens",
"gene",
"function"
] |
2014
|
Systematic Analysis of Zn2Cys6 Transcription Factors Required for Development and Pathogenicity by High-Throughput Gene Knockout in the Rice Blast Fungus
|
Cdk1 kinase phosphorylates budding yeast Srs2 , a member of UvrD protein family , displays both DNA translocation and DNA unwinding activities in vitro . Srs2 prevents homologous recombination by dismantling Rad51 filaments and is also required for double-strand break ( DSB ) repair . Here we examine the biological significance of Cdk1-dependent phosphorylation of Srs2 , using mutants that constitutively express the phosphorylated or unphosphorylated protein isoforms . We found that Cdk1 targets Srs2 to repair DSB and , in particular , to complete synthesis-dependent strand annealing , likely controlling the disassembly of a D-loop intermediate . Cdk1-dependent phosphorylation controls turnover of Srs2 at the invading strand; and , in absence of this modification , the turnover of Rad51 is not affected . Further analysis of the recombination phenotypes of the srs2 phospho-mutants showed that Srs2 phosphorylation is not required for the removal of toxic Rad51 nucleofilaments , although it is essential for cell survival , when DNA breaks are channeled into homologous recombinational repair . Cdk1-targeted Srs2 displays a PCNA–independent role and appears to have an attenuated ability to inhibit recombination . Finally , the recombination defects of unphosphorylatable Srs2 are primarily due to unscheduled accumulation of the Srs2 protein in a sumoylated form . Thus , the Srs2 anti-recombination function in removing toxic Rad51 filaments is genetically separable from its role in promoting recombinational repair , which depends exclusively on Cdk1-dependent phosphorylation . We suggest that Cdk1 kinase counteracts unscheduled sumoylation of Srs2 and targets Srs2 to dismantle specific DNA structures , such as the D-loops , in a helicase-dependent manner during homologous recombinational repair .
Homologous recombination ( HR ) is a fundamental DNA repair pathway and its deregulation is responsible for a variety of genomic rearrangements , including chromosome loss , DNA translocations and inversions , which are typical of the genetic alterations seen in tumor cells ( reviewed in [1] ) . The mechanisms and proteins involved in HR have been well conserved throughout evolution and much of our knowledge on HR comes from studies conducted in the yeast Saccharomyces cerevisiae ( reviewed in [2]–[5] ) . HR targets multiple DNA lesions , including single-stranded DNA ( sDNA ) breaks and DSBs , promoting their repair using a region of homology as a template . Diverse pathways can seal sDNA breaks , but the role of HR in DSB repair is essential . Different HR sub-pathways compete for DSB repair and some are less accurate than others [6] , [7] . The position of DNA sequences involved in recombination and the extent of their homology influence the kinetics of DSB repair . Irreparable DNA breaks [8] , or even those repaired slowly [9] , appear to be sequestered to the nuclear periphery , through a mechanism resembling that used to tether telomeres at the nuclear membrane [10] . When a region of homology is found on both sides of a DSB , the preferred pathway of repair is gene conversion ( GC ) . Among the initial steps in GC is the formation of Rad51 presynaptic nucleofilaments assisted by accessory factors . While Rad51 nucleation can occur directly at sDNA breaks , the ends of DSBs must be first processed to produce sDNA tails in order to recruit Rad51 . Multiple factors with nuclease and/or helicase activities , including the Mre11/Rad50/Xrs2 complex , Sae2 , Exo1 , Dna2 and Sgs1 cooperate in 5′ to 3′ DSB resection ( reviewed in [11] ) . Assembled Rad51 nucleofilaments invade and displace a duplex donor homologue DNA template leading to the formation of a D-loop structure . The D-loop is the site of DNA synthesis , which is promoted by extension of the 3′ invading strand . According to the canonical DSB repair model [12] , the capture of the second end of the DSB generates a double Holliday junction ( dHJ ) whose resolution , by cutting or branch migration , influences the formation of crossover products associated with GC , that is , the extent of DNA exchanges associated with DSB repair . If the second DSB end is not captured , it can anneal to the invading strand evicted from the D-loop soon after DNA synthesis . In this process , called synthesis-dependent strand annealing ( SDSA ) , GC is limited to DNA synthesized from donor strand and crossovers are prevented [4] . Another HR pathway , known as single strand annealing ( SSA ) , is used when DSB repair occurs between direct repeats [4] . In this case resected homologous sequences anneal without DNA synthesis and DSB repair is associated with deletion of the sequence between the repeats . Notably , during SSA , a D-loop is not formed and Rad51 is not required . The formation of presynaptic Rad51 nucleofilaments is fundamental for HR commitment during GC . However , Rad51 could nucleate improperly on DNA or even be engaged into damaged filaments when other recombination factors are inactivated: in both cases HR is not proficient , rather it becomes toxic for other DNA transactions . Many in vivo studies suggest that Srs2 , a member of UvrD family of DNA helicases conserved from bacteria to human , is involved in the removal of toxic Rad51 filaments from sDNA [13]–[17] . Further , the Srs2 protein disrupts presynaptic Rad51 filaments through its DNA translocase activity in vitro [18] , [19] . This Srs2 anti-recombination activity requires a physical interaction with sumoylated PCNA , as it was evidenced in the absence of the post-replication repair ( PRR ) pathway , a context in which Srs2 prevents deadly the recombinational repair [20] , [21] . Srs2 also exhibits 3′ to 5′ DNA helicase activity on duplex DNA [22] . Recent in vitro studies in yeast and plants suggest that Srs2 unwinds DNA structures mimicking a D-loop [23] , [24] . Genetic evidence , indeed , suggests that Srs2 favors the SDSA pathway , since the loss of Srs2 results in an increase in crossover products [25]–[27] . Moreover , Srs2 is essential for DSB repair through either SSA or ectopic GC [25] , [28] , [29]; in SSA repair , Srs2 is required to mediate recovery from checkpoint-mediated arrest [29] . Since Srs2 affects HR in several ways , Srs2 functions in recombination are probably regulated . Previous studies demonstrated that Srs2 is a target of the cell cycle-dependent kinase ( Cdk1 ) in vivo [30] and in vitro [31] . Cdk1 has been implicated in the DNA damage response and in DSB repair [32]; by monitoring repair of one HO-induced break , it was shown that Cdk1 is required both at the level of resection and at a step after Rad51-dependent strand invasion [33] , [34] . It is known that Cdk1 triggers the resection of DSB ends by phosphorylating Sae2 [35] , but other direct targets in DSB repair are unknown . We found that srs2 mutants that are unable to undergo Cdk1-dependent phosphorylation can still remove toxic Rad51 nucleofilaments , but these srs2 mutants fail to promote homologous recombinational repair . Analysis on repair of a single HO-induced break through ectopic GC shows that the proper turnover of Srs2 , at D-loop intermediates , is dependent on its modification by phosphorylation and this phosphorylation is essential for completion of the SDSA reaction that results in non-crossover products . Moreover , the phosphorylation-dependent role of Srs2 does not require an interaction with PCNA and does not affect the turnover of Rad51 at invading filaments . In the absence of Srs2 phosphorylation , the protein is sumoylated and this is the main cause of the recombinational repair defects seen in the nonphosphorylatable srs2 mutant . Thus , coordination of the sumoyaltion and phosphorylation modifications on Srs2 is essential during homologous recombinational repair .
Saccharomyces cerevisiae Srs2 contains characteristic amino acid motifs important for ATP-binding and DNA-binding that are highly conserved among members of UvrD family [36] . All these motifs are located in the N-terminal domain of the Srs2 protein ( grey box in Figure 1A ) and are sufficient for the helicase activity [22] , but not for translocase-dependent removal of Rad51 nucleofilaments , as tested in vitro [37] , [38] . The C-terminal tail of Srs2 protein plays an important regulatory function , since it mediates protein-protein interactions , including interaction with Rad51 and PCNA [15] , [21] , [37]–[40] . Moreover , a cluster of five consensus sites for Cdk1 kinase is present in the C-terminal region of Srs2 , while two additional sites are located in the helicase domain ( Figure 1A; [39] ) . The last 138 amino acids ( aa ) of the Srs2 C-terminal tail are required for the interaction with PCNA [21] and also contain three consensus sites for sumoylation ( Figure 1A ) . We previously showed that changing the seven serine/threonine Cdk1 consensus sites to the unphosphorylatable residues alanine/valine abolished DNA damage-induced phosphorylation of Srs2 , which can be monitored as an electrophoretic mobility shift on SDS-PAGE ( Figure 1B; [39] ) . We then produced a new srs2 allele in which the same serine/threonine residues were changed to the negatively charged aspartic acid/glutamic acid residues , with the aim of producing a mutated version of Srs2 that mimics the constitutively phosphorylated protein isoform . As shown in Figure 1B , the levels of wt Srs2 and the two mutated Srs2 isoforms are similar , both in normal conditions and in response to DNA damage by methyl methanesulfonate ( MMS ) -treatment ( data not shown ) . Henceforth , we will refer to the unphosphorylated and phosphorylated srs2 mutants , respectively , as srs2-7AV and srs2-7DE . To investigate whether Cdk1-dependent phosphorylation of Srs2 is important for its roles in HR , we first evaluated cell survival of the two srs2 phospho-mutants following UV-light and zeocin treatments . Wild type ( SRS2 ) and srs2Δ strains were used as controls . Previous studies have shown that the UV-sensitivity of srs2Δ strains is suppressed by mutations in RAD51 , indicating that cell lethality is due to accumulation of toxic Rad51 nucleofilaments at gaps whose repair can occur in the absence of HR [16] . We found that srs2Δ and rad51Δ mutants are also sensitive to zeocin , a radiomimetic chemical that induces DSBs ( Figure 2A and data not shown ) . Thus , zeocin-treatment induces DNA lesions whose repair is strictly HR-dependent and prevented in the absence of Srs2 . As shown in Figure 2A , we found that both srs2-7AV and srs2-7DE mutants , as SRS2 strains , survive UV-light doses that kill srs2Δ mutants . Conversely , the srs2-7AV mutant , but not the srs2-7DE mutant , is sensitive to zeocin and , indeed , is even more sensitive than the srs2Δ strain . Previous reports showed that srs2Δ mutations are synthetically lethal with either sgs1Δ or rad27Δ mutations [13] , [14] , [41] . While the synthetic lethality of srs2Δ sgs1Δ double mutants is suppressed by rad51Δ [13] , single rad27Δ mutants are themselves lethal in combination with rad51Δ [42] . Thus , the types of spontaneous DNA damage accumulating in sgs1Δ and rad27Δ mutants mirror those induced by UV and zeocin treatments: only in rad27Δ mutants and under zeocin treatment , HR is essential for DNA repair . We crossed the srs2-7AV and srs2-7DE phospho-mutants and srs2Δ as control with sgs1Δ or rad27Δ mutants . Heterozygous diploid mutants were sporulated and tetrad analysis was performed . As shown in Figure 2B , neither srs2Δ sgs1Δ nor srs2Δ rad27Δ double mutants form viable spores; the srs2-7AV mutation , but not the srs2-7DE mutation , is synthetically lethal with the rad27Δ mutation , while both srs2-7AV sgs1Δ mutants and srs2-7DE sgs1Δ mutant spores form colonies . Hence , the phenotypes of srs2-7AV mutants suggest that Srs2 phosphorylation is dispensable for the reversal of toxic Rad51-dependent recombination intermediates induced at sDNA by UV or by the absence of Sgs1 , but phosphorylation is required to promote recombinational repair in zeocin and in the absence of Rad27 . Previous data suggested that the Srs2 protein sensitizes postreplication repair ( PRR ) mutants , because it prevents HR [43] , [44] . Accordingly , as show in Figure 2C , the sensitivity of rad5Δ mutants to MMS is alleviated by deleting SRS2 . srs2 mutants encoding a protein that displays attenuated translocase activity also suppress the DNA damage sensitivity of PRR mutants , although they are not sensitive to DNA damaging agents by themselves [40] . Hence , we analyzed the srs2 phospho-mutants in a PRR mutant context , in which the importance of having an intact DNA translocation activity should be revealed . We constructed srs2 phospho-mutants in rad5Δ or rad18Δ backgrounds and then tested viability on medium containing MMS . We found that srs2-7AV mutation hypersensitizes rad5Δ and rad18Δ mutants to DNA damage , but , conversely , the srs2-7DE mutation partially suppresses the lethality of rad5Δ or rad18Δ mutation ( Figure 2C and data not shown ) . Notably , srs2-7AV and srs2-7DE mutants are not sensitive to MMS , even at a higher MMS dose than those employed in Figure 2C ( data not shown ) . Thus , we conclude that , even in a PRR context , unphosphorylatable Srs2 can remove Rad51 at DNA gaps . On the other hand , the phosphorylated Srs2 protein isoform appears to be less proficient in the anti-recombinational role . The observation that srs2-7AV mutants are sensitive to treatment with zeocin suggests that phosphorylation of Srs2 is important in DSB repair . To directly examine this , we tested the behavior of srs2 phospho-mutants in response to a single DSB created by a galactose-inducible HO endonuclease . Previous studies have shown that srs2Δ mutants can not survive a single HO-induced DSB when repair of this break occurs either by ectopic GC or by SSA [25] , [28] , [29] . While the GC pathway strictly depends on RAD51 , SSA can occur in the absence of Rad51 . There are important differences in the requirement for Srs2 in the two pathways: Srs2 is not required to complete DSB repair during SSA , but it is required for recovery from the DNA damage-induced cell cycle arrest [29] . RAD51 deletion rescues the checkpoint recovery defect in srs2Δ mutants [29]; thus , one hypothesis is that Rad51 accumulates on DNA contributing to the lethal checkpoint-induced arrest , since it can not be removed in absence of Srs2 [29] , [45] . Conversely , during ectopic GC , srs2Δ mutants are unable to complete DSB repair , with a specific reduction in non-crossover products formation [25] . Since the region of DNA homology involved is limited in ectopic DSB repair , the formation of crossovers might be prevented because the formation of the dHJ intermediate is reduced [46] . Thus , the failure to carry out SDSA results in loss of non-crossover products and there is a marked reduction in DSB repair efficiency [25] . To analyze the requirement of Srs2 phosphorylation in the DSB repair response , we assayed cell survival of srs2 phospho-mutants in a SSA system in which DSB repair occurs between repeated sequences , one of which is located 25kb from the DSB and results in a chromosomal deletion [29] or in an ectopic GC system in which DSB repair occurs between chromosomes V and III [25] . In agreement with previous findings , the rate of cell survival of srs2Δ mutants is 2% in both the SSA and GC systems ( Figure 3A and 3B ) . This high cell lethality in srs2Δ mutants correlates with inability to dephosphorylate the checkpoint kinase Rad53 , which is activated in response to DSB induction ( Figure 3A and 3B ) . Cell survival of srs2-7AV mutants is 25% in the GC system where they also fail to fully dephosphorylate Rad53 24 hours after DSB induction ( Figure 3A and 3B ) . Survival of the srs2-7AV mutant is normal in the SSA system and survival of the srs2-7DE mutant is normal in both systems . Thus , Srs2 phosphorylation is necessary for cell survival when DSB repair proceeds through the Rad51-dependent GC pathway , but is dispensable in the SSA pathway , which does not require Rad51 . As mentioned above , although SSA is Rad51-independent pathway , in absence of Srs2 , Rad51 might improperly accumulate on DNA and interfere with checkpoint recovery [29] , [45] . Since srs2-7AV survive DSB repair via SSA , this further strengthens the conclusion that Srs2 phosphorylation is not required for reversal of toxic Rad51-dependent intermediates . We used Southern blotting with a probe that recognizes the MAT locus , to physically observe DSB repair in the GC system ( Figure 3B ) . As mentioned above , in this system a DSB is induced at a MAT locus inserted into Chromosome V and is repaired using a unique uncleavable MAT-inc cassette on chromosome III ( Figure 3B ) . Notably , crossovers that are associated with the GC event can be evaluated by restriction analysis , since crossovers give rise to chromosomal bands that differ in size from the parental chromosomes and the non-crossover GC products ( Figure 3B ) . As shown in Figure 3C , DNA of SRS2 and srs2 mutants were analyzed by Southern blotting . DSB repair efficiency is about 30% in srs2Δ strain , in agreement with previous findings [25] and in srs2-7AV it is reduced to 70% compared to SRS2 or srs2-7DE ( Figure 3C ) . Moreover , the percentage of crossovers associated with GC increases three-fold in srs2Δ and two-fold in srs2-7AV compared to SRS2 or srs2-7DE ( Figure 3C ) . Similar to the srs2Δ mutants , the increase in crossovers is associated with a reduction in non-crossover repair efficiency in the srs2-7AV mutant ( Figure 3C ) ; thus , DSB repair defects in the absence of Srs2 phosphorylation likely indicate a specific failure to carry out repair via the SDSA pathway that results in non-crossover products . Our analysis indicates that Srs2 phosphorylation is required for Rad51-dependent DSB repair . Although we found that Srs2 phosphorylation is not essential for the removal of toxic Rad51 nucleofilaments at DNA gaps or during DSB repair by SSA , it might be specifically required to remove Rad51-dependent recombination intermediates initiated at D-loop intermediate . To investigate this possibility , we analyzed Rad51 binding to DSBs by ChIP and Q-PCR in SRS2 and srs2 phospho-mutants . We used DNA primers that amplified the region of homology located on donor chromosome III . Using this strategy , proteins localizing either at broken or recipient chromosomes will be immunoprecipitated at the DSB when the invading strand is in duplex DNA , which most likely represents the D-loop . As shown in Figure 4A , Rad51 protein is undetectable at the donor MAT locus before HO induction , while it is loaded at the DSB with similar kinetics in SRS2 and all srs2 mutated strains . Thus , we conclude that Rad51-mediated strand invasion occurs with similar kinetics in SRS2 and srs2 mutants . We also conclude that Rad51 is removed from the DSB with similar kinetics in all contexts and strains analyzed . Thus , DSB repair defects in srs2Δ or srs2-7AV mutants are unrelated to an abnormal persistence of Rad51 after strand invasion . Previous findings have indicated that Srs2 is loaded at DSBs [47] . We asked whether Srs2 phosphorylation could influence its ability to be recruited to DSBs in our GC system . Using the same ChIP strategy employed above , we found that Srs2 is sited at the invading strand with a three-fold enrichment ( Figure 4B ) . The Srs2 and Srs2-7DE proteins are loaded and dislodged from DNA with kinetics resembling that of Rad51 , but the Srs2-7AV protein accumulates only at later times and abnormally persists on DNA for at least 24 hours after DSB induction; notably , Rad51 has been displaced from DNA , when Srs2-7AV protein accumulates ( Figure 4 ) . In summary , the data in Figure 4 suggest that Srs2 is loaded at the D-loop during GC and its proper recruitment is governed by Cdk1-dependent phosphorylation . However , the DSB repair defects in srs2-7AV or srs2Δ mutants appear not be related to inefficient metabolism of Rad51 nucleofilaments at donor DNA sequences . In the course of our studies on Srs2 phosphorylation , we noticed that in response to massive DNA damage , such as treatment with 0 . 3% MMS , Srs2 accumulates as additional modified isoforms , which can be visualized as a ladder on SDS-PAGE analysis ( Figure 5A ) . These Srs2 protein isoforms are recognized by SUMO-specific antibodies ( Figure 5A ) . Preliminary characterization of Srs2 sumoylation indicates that none of the well-characterized SUMO ligases , including Siz1 and Siz2 , are involved in this modification ( Figure S1A ) . Three putative sumoylation sites have been mapped to the C-terminus tail of Srs2 ( Figure 1A ) . Our data indicated that DNA damage-induced sumoylation of Srs2 was abolished in srs2-3KR mutants , in which the three lysine residues in the motifs identified as modified by SUMO were mutated to arginine ( Figure 5A ) . Notably , the Srs2-3KR protein can be fully phosphorylated ( Figure S1B ) . Intriguingly , while sumoylation of native Srs2 is induced at 0 . 3% MMS , the unphosphorylatable Srs2-7AV protein can be detected as SUMO-modified isoforms at ten-fold lower MMS doses ( Figure 5A ) . Thus , while sumoylation and phosphorylation can occur independently , Srs2 accumulates in a sumoylated form in the absence of phosphorylation . The biological relevance of Srs2 sumoylation is still obscure , as extensive studies of the phenotypes of the srs2-3KR mutant were inconclusive ( D . Callahan and H . Klein , unpublished results ) . However , the finding that unphosphorylatable Srs2 is hyper-sumoylated prompted us to ask if the srs2-7AV mutant defects in recombinational repair might be related to Srs2 sumoylation . To test this , we mutagenized the sumoylation consensus sites in the srs2-7AV mutant to create the srs2-7AV3KR allele , which is simultaneously impaired for phosphorylation and sumoylation . We then tested the behavior of the srs2-7AV3KR mutant in the DSB repair GC system in which srs2-7AV mutant was highly sensitive ( see Figure 3 ) . We found that srs2-7AV3KR mutant survived DNA damage ( Figure 5B ) ; DSB repair is accomplished efficiently and a normal level of crossovers is seen in srs2-7AV3KR ( Figure 5B ) . In addition , the srs2-7AV3KR mutant correctly reversed the checkpoint response after DSB induction and repair , as seen by Rad53 kinase dephosphorylation ( data not shown ) . Furthermore , the srs2-3KR mutant , which is only impaired in sumoylation , can accomplish DSB repair ( Figure 5B ) . To see if ablation of Srs2 sumoylation rescues the phosphorylation defects in recombinational repair in other contexts , we crossed the srs2-7AV3KR mutant with the rad27Δ mutant to generate rad27Δ srs2-7AV3KR double mutants . While the rad27Δ srs2-7AV double mutants never form viable spores ( see Figure 2A ) , we found that rad27Δ srs2-7AV3KR double mutants developed into visible colonies ( 17/25 of total cases analyzed ) , although the double mutant grew very slowly ( Figure 5C ) . This partial suppression highlights the importance of Srs2 protein modifications when it is likely that more than one lesion is formed . Taken together , the data in Figure 5 indicate that Srs2 is sumoylated in vivo . Sumoylation of Srs2 is not required for DSB repair , but the recombinational repair defects in unphosphorylatable srs2-7AV mutants are largely related to the unscheduled sumoylation of the protein . The sumoylation consensus sites are located in the last 138 residues of the C-terminus tail of Srs2 ( Figure 1A ) , which also mediates the interaction with PCNA [21] . Hence , we asked if this tail is important for the Cdk1-dependent role of Srs2 . As shown in Figure 6 , we found that the srs2−ΔC138 mutant is viable after induction of a HO-mediated DSB and also when combined with a rad27Δ . Conversely , unphosphorylatable srs2-7AV mutants lacking the PCNA-interaction domain ( srs2-7AVΔC138 ) are lethal in both contexts . These data suggest that Cdk1 targets Srs2 to promote recombinational repair independently of the interaction with PCNA and sumoylation . Moreover , elimination of sumoylation sites , but not deletion of the Srs2 tail containing the same sites , suppresses the recombination defects in the srs2-7AV mutant .
Recombination can be both prevented and stimulated in srs2 mutants , suggesting a dual role for Srs2 in HR . The finding that Srs2 is a DNA translocase that antagonizes the formation of unscheduled Rad51 filaments explains certain srs2 phenotypes in HR that are suppressed by ablating RAD51; these include the synthetic lethality with sgs1 mutants or high sensitivity to UV-light [13] , [16] . Nevertheless , srs2 mutants are defective in Rad51-dependent DSB repair [25] , [28] or lethal when combined with rad27Δ mutants [14] , [41] . These are contexts in which HR is essential to restore DNA lesions and the activity of Srs2 is required to promote homologous recombinational repair . In this study we analyzed the recombination phenotypes of two srs2 mutants that mimicked either the constitutive unphosphorylated ( srs2-7AV ) or Cdk1-dependent phosphorylated ( srs2-7DE ) protein isoforms . We found that srs2-7AV unphosphorylatable mutants display only a subset of srs2Δ phenotypes and , in particular , they do not display those phenotypes that are suppressed by RAD51 deletion . In fact , srs2-7AV mutants are not UV-sensitive or synthetically lethal with sgs1Δ , but are non-viable when combined with rad27 mutants or treated with the DSB-inducing drug zeocin . Thus , functions of Srs2 in preventing unscheduled recombination or in allowing efficient recombinational repair are genetically separable . The phosphorylation of Srs2 is dispensable for the removal of toxic Rad51 nucleofilaments assembled at gaps , while it is essential to promote recombinational repair . In accordance with the finding that Srs2 phosphorylation is essential to promote recombination , we found that it is also required for Rad51-mediated DSB repair . In particular , we have been able to show that Srs2 phosphorylation is necessary to complete SDSA in DSB repair . ChIP data on Rad51 are consistent with the idea that strand invasion is not affected and that Rad51 protein does not persist on the D-loops in srs2Δ or srs2-7AV mutants , although we cannot rule out that presynaptic filament assembly may somehow be affected in the absence of Srs2 or its phosphorylation . ChIP analysis conducted on Srs2 suggests that the protein is found at DSBs upon strand invasion , thus it is likely loaded at D-loops . Taken together , these data are consistent with a role of phosphorylated Srs2 in SDSA pathway , but another helicase/translocase may be implicated in removing Rad51 at the D-loops . We favour the idea that Cdk1 targets Srs2 to dismantle the D-loop intermediate in SDSA ( Figure 7 ) perhaps after DNA synthesis has extended the invading strand . Srs2 helicase activity might be stimulated by binding to the D-loop structure and/or by interaction with other recombination factors . ChIP data conducted on unphosphorylatable Srs2-7AV at the invading strand suggest that the mutated protein accumulates at later times and is not rapidly dislodged from DNA as the wild-type protein . The fact that unphosphorylatable Srs2 appears glued at the D-loops is evocative of a protein working very inefficiently and whose turnover is largely prevented . It is likely that the unscheduled accumulation of the protein on the DNA might contribute to impaired cell viability and , consistent with this idea , the lethal phenotype of srs2-7AV mutant in response to DSBs is dominant ( Figure S2 ) . Our data indicate that the proportion of srs2-7AV cells that do not survive DSB repair via GC is higher than the one , which fails to repair DNA lesion ( Figure 3 ) . This suggests that a fraction of srs2-7AV cells might die because of checkpoint-mediated arrest , as in srs2Δ mutants [25] . However , Srs2 phosphorylation is not required for recovery during DSB repair by SSA , that is , when Srs2 is probably engaged to remove toxic Rad51-depedent DNA structures , rather than working at the D-loop intermediate [45] . Thus , the checkpoint recovery defect in srs2-7AV mutants might have different causes during DSB repair by GC or SSA; as described below , perhaps some aspects of recovery defect in srs2 mutants in GC could be explained considering that Cdk1-dependent phosphorylation was no longer required for Srs2 recombination activity , if sumoylation is also prevented . We found that Srs2 sumoylation can be detected in vivo in response to heavy DNA damage . Protein modification is prevented ablating three lysine residues located in the extreme C-terminus tail of Srs2 . Sumoylation and Cdk1-dependent phosphorylation modifications of Srs2 are independent events , but when phosphorylation fails , sumoylated Srs2 accumulates . There is a functional relationship between these two DNA damage induced modifications , since ablation of sumoylation residues largely rescues the recombinational repair phenotypes of srs2-7AV mutants . What may be the mechanism for the toxicity of sumoylation in the absence of phosphorylation ? Sumoylation of Srs2 alone appears unnecessary for many of its recombination functions ( D . Callahan and H . Klein , unpublished results ) ; here we show that it is not essential in DSB repair ( see Figure 5B and Figure 6 ) . While the biological significance of Srs2 sumoylation waits to be elucidated , we speculate that it might be important for degradation of Srs2 protein . Srs2 can interact physically with Slx5 [39] , that in complex with Slx8 , has been implicated in degradation of sumoylated proteins bound to irreparable DNA breaks at the nuclear periphery [8] , [9] . Our data suggest that Cdk1-dependent phosphorylation of Srs2 counteracts its sumoylation , which takes over only in response to massive DNA damage . Thus , in a possible scenario , unphosphorylated and sumoylated Srs2 is trapped at DSB and becomes channeled via the Slx5/Slx8 pathway to the nuclear periphery ( Figure 7 ) . Since this emergency nuclear periphery pathway intervenes to degrade proteins in response to irreparable DSBs , it might normally act on phosphorylated and sumoylated Srs2 and , therefore , Srs2-7AV cannot be eliminated . Conversely , after successful DSB repair , phosphorylated Srs2 could be recycled by other routes , and independent of sumoylation . Intriguingly , the unscheduled Srs2-dependent sequestration of DSBs to the periphery might explain the checkpoint recovery defects in srs2-7AV and perhaps also that of srs2Δ , if we imagine that another unregulated DNA helicase takes over in the absence of Srs2 . Our studies did not show any obvious alterations in Srs2 protein levels in srs2 phospho-mutants and/or SUMO-mutants ( M . Saponaro and G . Liberi , unpublished results ) , but local protein degradation events at damaged DNA could be relevant . Elimination of sumoylation compensates for the absence of phosphorylation of Srs2 in DSB repair , but paradoxically this rescue requires the last 138 residues of Srs2 that are not normally necessary for DSB repair . Hence , this suppression might require interaction with other factors . Preventing sumoylation in the unphosphorylatable Srs2 rescues recombination defects that ensue after a single DSB , but the importance of these Srs2 modifications become evident when many breaks occur , as in the rad27Δ mutants . We found that Srs2 phosphorylation is essential for recombinational repair of spontaneous damage occurring during S-phase in rad27Δ mutants . Similar to the response to DSBs , sumoylation of Srs2 is a main cause of death in srs2-7AV phospho-mutants . It is more difficult to predict the kind of damage which requires phosphorylated Srs2 in rad27Δ mutants . Rad27 is required for Okazaki DNA fragment processing [48] and in its absence , Srs2 might dismantle DNA and/or RNA structures that block HR . In any case , based on our conclusion that Srs2 phosphorylation is not essential for the processing of toxic Rad51 filaments , we think it more probable that the helicase activity , rather than translocase activity , is crucial for the survival in rad27Δ mutants . This proposed role of phosphorylated Srs2 in replication might seem at odds with the role suggested for Srs2 in preventing recombinational repair during S-phase through recruitment by sumoylated PCNA [20] , [21] . However , in PRR mutants , Srs2 is proposed to be recruited by PCNA to disrupt Rad51 filaments at DNA gaps , while in the absence of Rad27 , we are considering that Srs2 acts in a PCNA-independent and phosphorylation-dependent role as a helicase , rather than as a translocase . Importantly , srs2-7DE mutants slightly suppress the MMS sensitivity of PRR mutants , suggesting that the phosphorylated Srs2 is less efficient as a DNA translocase than the non-phosphorylated isoform . This is unmasked in PRR mutants , where it is likely that many sDNA breaks occur . Srs2 phosphorylation might modulate its interaction with PCNA , a hypothesis that will be interesting to test in the future . Our data indicate that Srs2 is a new target of Cdk1 kinase in DSB repair , acting at the level of strand invasion , rather than during DNA end resection . Srs2 phosphorylation is required to limit the extent of DNA exchanges during DSB repair with a function that is genetically separable from its role in processing toxic Rad51 filaments . We suggest that Cdk1-mediated phosphorylation might control , throughout the interaction with PCNA and/or other factors , the ability of Srs2 to function as a translocase or a helicase that inhibits or allows HR depending on the context . Furthermore , our data unravel a novel aspect of Cdk1-dependent regulation in counteracting untimely sumoylation events , which might become toxic for recombination if not properly scheduled .
Genotypes of the strains used in this study are listed in Table S1 . Deletion strains were obtained by the one-step PCR method and multiple mutant strains were derived from meiotic segregants of appropriate crosses . The srs2-7DE phospho-mutant was constructed by a site-directed mutagenesis strategy already described to construct the srs2-7AV mutant [39] . Mutations in SRS2 were introduced at the seven consensus sites for the Cdk1 kinase ( T604D , S698E , S879E , S893E , S938E , S950E and S965E ) . Construction of srs2-3KR strain , containing mutations K1081R , K1089R and K1142R at SUMO-consensus sites , will be described in detail elsewhere . srs2-7AV3KR mutant was constructed as follow: a NAT selection cassette was integrated downstream of the srs2-3KR mutated gene . DNA primers were designed to amplify a DNA region containing both the 3KR mutations and the NAT cassette . This DNA region was then used to replace , by transformation , the C-terminus of the srs2-7AV . A similar PCR-mediated strategy was used to delete the C-terminus-PCNA interaction domain in both wild type and srs2-7AV mutants . SRS2 and srs2 mutants were also cloned into the low copy-number Ycplac22 vector by gap-repair procedure or using PCR-based strategies described above and were used in all HO-based experiments . As tested by Western blotting , protein levels are similar when Srs2 or its mutated versions were expressed from SRS2 chromosomal locus or from the Ycplac22 centromeric plasmid . Log-phase cells were spread on YPD plates , irradiated with UV light ( 254 nm ) and incubated in the dark; cell survival was compared to that of untreated controls . Log-phase cultures were incubated with different doses of zeocin ( Invitrogen ) for 1 hour and cell survival was calculated by comparing the plating efficiency with untreated cells . The UV and zeocin curves are the average of three independent experiments . Spot assays were performed by evaluating the growth of serially diluted cultures on synthetic complete medium containing adenine at a final concentration of 0 . 7 g/liter , with or without MMS ( SIGMA ) . Relative frequencies of survival of cultures plated on glucose and galactose using the SSA and GC DSB HO-inducible systems were calculated as previously described [25] , [29] . Product formation and analysis of crossover formation were assessed by Southern blotting analyses as described in [25] . The results shown are the average of three to five independent experiments . Asynchronous cultures were grown overnight at 28°C in YEP media containing 2% raffinose . When cultures reached mid log-phase , 15 µg/ml nocodazole was added to synchronize the cells in G2/M . Expression of the HO endonuclease was induced by addition of 2% galactose for the indicated times . ChIP analysis was carried out as previously described [49] . Samples were incubated with 1% Formaldehyde for 20 min with the anti-Rad51 ChIP and 45 minutes with the anti-Srs2 ChIP . Immunoprecipitation was carried out with clarified extracts using anti-Rad51 ( kindly gift of Patrick Sung , Yale University , New Heaven ) or anti-Srs2 ( Santa Cruz Biotechnology ) antibodies overnight at 4°C . Levels of immunoprecipitated DNAs were measured by quantitative real-time PCR using the SYBR Green technique ( SYBR Green PCR Master Mix , Applied Biosystems ) and run in an Applied Biosystems 7500 Fast Real-Time PCR System . Sequences of the DNA primers are listed in the Table S2 . Dissociation stage curves were checked to test primer specificity . The results were analyzed with the 2-DCT method as previously described [50] . For the ChIP anti-Rad51 experiments the relative enrichment was determined by the fold increase of ChIPed DNA relative to that prior to DSB induction; for the ChIP anti-Srs2 the absolute 2-DCT variation after the subtraction of the 2-DCT of the ChIP carried out in parallel in a srs2Δ mutant is shown . The total amount of DNA is normalized respect to an unrelated locus near ARS305 . Proteins were extracted using a TCA protocol [30] . Western blotting analysis was performed as previously described [30] with anti-Rad53 [51] and anti-Srs2 ( Santa Cruz Biotechnology ) antibodies . Immuno-precipitation of Srs2 was carried out with an anti-Srs2 antibody or without antibody as a control on proteins extracted by the TCA protocol using a 50 ml culture of cells at a density of 107 cells/ml . After a series of washings with JS buffer ( 50 mM Hepes pH 7 , 5 , 150 mM NaCl , 1 , 5 mM MgCl2 , 1% glycerol , 5 mM EGTA , 1% Triton X-100 ) , proteins were resuspended in a suitable volume of Laemmli buffer and separated on a 10% acrylamide SDS-PAGE gel . The blots were probed with a rabbit anti-SUMO antibody ( kindly gift of Xiaolan Zhao , Memorial Sloan-Kettering Cancer Center , New York ) and a HRP-labeled secondary antibody ( Amersham ) . Blots were then stripped with the commercial solution Restore Western Blot Stripping Buffer ( Thermo Scientific ) and probed with the anti-Srs2 antibody ( Santa Cruz Biotechnology ) .
|
Broken DNA molecules can be repaired by copying a homologous DNA sequence located elsewhere in the genome . This process , called homologous recombination , needs to be carefully regulated , because unwanted DNA exchanges can lead to genome rearrangements and cell death . Cdk1 kinase is required for cell cycle progression and phosphorylates DNA repair factors , such as Srs2 , a protein that can both translocate on single-stranded DNA and open the two strands of DNA double helix . DNA translocation activity of Srs2 is crucial to prevent unwanted recombination , while DNA unwinding activity might be important to promote recombination . In this study , we used two srs2 mutants that constitutively express the unphosphorylated or Cdk1-dependent phosphorylated Srs2 protein isoforms . We found that Srs2 performs genetically distinct functions in preventing or promoting homologous recombination . Cdk1 targets Srs2 to promote accurate repair of double-stranded DNA breaks , but is not essential for the removal of toxic recombination intermediates assembled at single-stranded DNA breaks . Further , Cdk1 counteracts sumoylation of Srs2 , which is responsible for recombination defects due to the lack of Srs2 phosphorylation . In summary , Cdk1-dependent Srs2 phosphorylation prevents its unscheduled sumoylation and targets the helicase to promote accurate homologous recombinational repair .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"molecular",
"biology/recombination",
"genetics",
"and",
"genomics/gene",
"function",
"molecular",
"biology/dna",
"repair"
] |
2010
|
Cdk1 Targets Srs2 to Complete Synthesis-Dependent Strand Annealing and to Promote Recombinational Repair
|
Ca2+ contributes to a myriad of important cellular processes in all organisms , including the apicomplexans , Plasmodium and Toxoplasma . Due to its varied and essential roles , free Ca2+ is tightly regulated by complex mechanisms . These mechanisms are therefore of interest as putative drug targets . One pathway in Ca2+ homeostatic control in apicomplexans uses a Ca2+/H+ exchanger ( a member of the cation exchanger family , CAX ) . The P . falciparum CAX ( PfCAX ) has recently been characterised in asexual blood stage parasites . To determine the physiological importance of apicomplexan CAXs , tagging and knock-out strategies were undertaken in the genetically tractable T . gondii and P . berghei parasites . In addition , a yeast heterologous expression system was used to study the function of apicomplexan CAXs . Tagging of T . gondii and P . berghei CAXs ( TgCAX and PbCAX ) under control of their endogenous promoters could not demonstrate measureable expression of either CAX in tachyzoites and asexual blood stages , respectively . These results were consistent with the ability of parasites to tolerate knock-outs of the genes for TgCAX and PbCAX at these developmental stages . In contrast , PbCAX expression was detectable during sexual stages of development in female gametocytes/gametes , zygotes and ookinetes , where it was dispersed in membranous networks within the cytosol ( with minimal mitochondrial localisation ) . Furthermore , genetically disrupted parasites failed to develop further from “round” form zygotes , suggesting that PbCAX is essential for ookinete development and differentiation . This impeded phenotype could be rescued by removal of extracellular Ca2+ . Therefore , PbCAX provides a mechanism for free living parasites to multiply within the ionic microenvironment of the mosquito midgut . Ca2+ homeostasis mediated by PbCAX is critical and suggests plasmodial CAXs may be targeted in approaches designed to block parasite transmission .
Free Ca2+ is essential for signalling in all cell types and plays a central role in many processes during the complex life cycles of apicomplexan parasites ( e . g . Plasmodium and Toxoplasma ) , including secretion of adhesins , motility , cellular invasion and egress , and intracellular development [1]–[3] . Proteins that interact with Ca2+ are therefore important to identify as novel drug targets and can provide fundamental insights into the biology of parasites when they are functionally characterised [4] . Our current understanding of Ca2+ homeostatic control in apicomplexan parasites is limited . To control cytosolic free Ca2+ concentrations , eukaryotic cells use a range of Ca2+ binding proteins , Ca2+ diffusive “leak” pathways and active Ca2+ transporters . For example , several apicomplexan P-type Ca2+ ATPases have been characterised , such as the sarco ( endo ) plasmic reticulum Ca2+ ATPases ( SERCAs ) of P . falciparum , PfATP6 , and T . gondii , TgSERCA [5] , [6] . Plasmodium P-type ATPases are already being investigated as new or existing drug targets . More recently , an initial characterisation of the P . falciparum Ca2+/H+ exchanger ( PfCAX , also termed the Ca2+/H+ antiporter , PfCHA ) has been undertaken [7] . PfCAX and other apicomplexan orthologues belong to the Ca2+/cation antiporter ( CaCA ) superfamily and members have been identified across the biological Kingdoms including some lower vertebrates , although not in more complex metazoa including mammals [8] , [9] . CAX genes are classified into 3 subfamilies . Type II CAXs are found in fungi , Dictyostelium , and lower vertebrates and Type III CAXs are found in bacteria , while Type I CAXs include bacterial , fungal , plant and protozoan CAXs . Type I CAXs are divided into 8 subgroups ( A to H ) and protozoa are classified into the Type I-C phylogenetic group . The first member of this group to be characterised functionally was a Ca2+/H+ exchanger ( CrCAX1 ) from the unicellular green alga Chlamydomonas reinhardtii [10] . Functional characterisation of CAX proteins ( mainly from plants and fungi ) has shown that these H+ coupled exchangers all transport Ca2+ , with some being highly specific for Ca2+ , whilst others mediate the transport of a broad range of additional divalent cations or , in some cases , transport additional monovalent cations [11]–[13] . Their primary role in plants and fungi is to enable tolerance to high extracellular Ca2+ concentrations , by internal sequestration of Ca2+ into acidic organelles when cytosolic levels rise [14] . PfCAX demonstrates Ca2+/H+ exchange activity , an ability to exchange a limited range of other divalent cations , and a high transport capacity but low affinity ( Km value of ∼2 mM ) for Ca2+ , when expressed in Xenopus oocytes [7] . In vivo studies characterising PfCAX are consistent with an atypical localisation to the inner mitochondrial membrane , and an atypical function , where the protein provides a pathway for removal of Ca2+ from this organelle back into the parasite cytosol . The aim of this study was to determine the physiological importance of apicomplexan CAXs . In addition to developing a yeast heterologous expression system for the functional characterisation of apicomplexan CAXs , the genetically amenable P . berghei and T . gondii parasites were used with tagging and knock-out strategies to define expression and essentiality of their respective CAXs . The data demonstrate that under the control of their respective endogenous promoters only the expression of the tagged P . berghei CAX , PbCAX , could be established , and this , only in sexual stages of development ( predominantly female specific ) . Furthermore , genetic disruption of pbcax , while having little or no effect on blood stage growth , was lethal during ookinete development . However , mutant parasites could be rescued by the removal of extracellular Ca2+ .
The P . falciparum , P . berghei and T . gondii CAX genes , pfcax ( PFF0170w ) , pbcax ( PBANKA_010230 ) and TgCAX ( TGME49_007910 ) , have 1326 , 1323 and 1506 base pair open reading frames , respectively , with only the latter having ( 12 ) introns . They are located on chromosomes 6 , 1 and 1b in their respective genomes and encode polypeptides of 441 , 440 and 501 amino acids , with estimated sizes of 48 , 49 and 53 kDa , respectively ( Figure 1 and S1 ) . All the apicomplexan cax genes identified are single copy genes with no close paralogues . PfCAX has greater than 80% , approximately 50% and 39% amino acid sequence identity compared with other Plasmodium spp . , Coccidia ( Toxoplasma , Cryptosporidium and Eimeria ) and C . reinhardtii CAX sequences , respectively . The phylogenetic relationship between the apicomplexan putative CAX transporters is shown in Figure S2 , in which CrCAX1 has been added as the first functionally characterised Type 1-C CAX [10] . Interestingly , BLAST searches , using the PfCAX amino acid sequence , did not reveal cax genes in the genomes of the Piroplasmida , Babesia bovis and bigemina or Theileria annulata , even though they are closely related to Plasmodium parasites . As with CrCAX1 and higher plant CAXs , the apicomplexan CAXs identified here are predicted to have 11 membrane spanning regions ( with the TMHMM tool at www . cbs . dtu . dk/services/TMHMM/; Figure 1 and S1 ) , although tagging experiments suggest that PfCAX spans the membrane 10 times only [7] . Apicomplexan CAXs also have long N-terminal sequences ( ranging from 71 amino acids for Cryptosporidium ssp . CAXs to 126 amino acids for TgCAX ) . All contain conserved residues ( Figure 1 and S1 ) that are essential for Ca2+ transport in Arabidopsis thaliana CAX1 and Oryza sativa CAX1a [15] , [16] but not the non-consensus residues identified in the c-1 and c-2 repeat regions of CrCAX1 that may be involved in Na+ transport [10] . Also annotated on Figure 1 and S1 are the predicted mitochondrial targeting sequences and phospho-acceptor sites reported previously [7] , [17] and on GeneDB ( www . genedb . org ) in the case of P . berghei ookinetes . To investigate whether apicomplexan CAXs can provide a tolerance mechanism against elevated Ca2+ concentrations and confirm that apicomplexan CAXs do function as Ca2+/H+ exchangers , a yeast heterologous expression approach was used . This was chosen over the Xenopus oocyte system used previously to study PfCAX [7] , as our understanding of CAXs from many diverse organisms has been advanced more thoroughly through studies in yeast and for which there is i ) a positive control , CrCAX1 [10] , ii ) optimised ways to detect function ( see Materials and Methods ) , and iii ) a context for interpretation of results [14] . CAXs were expressed in the Saccharomyces cerevisiae mutant K665 , which lacks two vacuolar Ca2+ transport pathways; a Ca2+-ATPase ( Pmc1 ) and a Ca2+/H+ exchanger ( Vcx1 ) . This yeast mutant is unable to sequester Ca2+ into the vacuole and is therefore hypersensitive to external Ca2+ , which causes toxicity due to elevation of cytosolic Ca2+ concentrations [11] . A codon optimized pfcax cDNA ( Figure S3 ) was expressed in K665 yeast and the ability of PfCAX to provide expressing yeast tolerance to high Ca2+ stress was assessed . Ca2+ tolerance by PfCAX was compared alongside the previously characterised Ca2+/H+ exchanger from Chlamydomonas [10] . The 2009 study demonstrated that the Chlamydomonas exchanger was more efficient at transporting Ca2+ when expressed in yeast if the N-terminal tail is truncated , removing a putative regulatory domain . An N-terminal truncated variant of pfcax ( spfcax ) , in which translation was initiated from AUG at nucleotide position 181 ( encoding Met-61 ) , was therefore generated for comparison with PfCAX and sCrCAX1 . Expression of pfcax and spfcax in yeast was detectable by RT-PCR ( Figure 2A ) . Comparison of PfCAX , sPfCAX , sCrCAX1 and empty vector expressed in yeast grown on YPD ( yeast-peptone-dextrose ) media supplemented with 50 mM CaCl2 ( Figure 2B ) found that sPfCAX could suppress the Ca2+ hypersensitivity of the yeast mutant as efficiently as sCrCAX1 , while full-length PfCAX-expressing yeast growth was slightly reduced ( as determined by the non-uniform growth of yeast expressing PfCAX at a starting cell density of 0 . 04 absorbance units at A600 nm ) . The Ca2+ tolerance of yeast expressing PfCAX and sPfCAX was demonstrated further by cell growth in liquid YPD media supplemented with various concentrations of CaCl2 ( Figure 2C ) . sPfCAX provided K665 yeast with tolerance to high concentrations of CaCl2 . sPfCAX-expressing yeast growth was significantly greater than PfCAX-expressing yeast at each CaCl2 concentration ( p<0 . 02 unpaired , two-tailed Student's t-test; n = 4 ) , while the cell growth of PfCAX-expressing yeast was significantly greater than that of the empty vector control yeast at all CaCl2 concentrations , including at 150 mM CaCl2 ( p<0 . 02 , unpaired , two-tailed Student's t-test; n = 4 ) . A truncated version of TgCAX ( sTgCAX ) , in which translation was initiated by introducing an AUG prior to nucleotide position 295 ( encoding Ala-99 ) , was also generated and expressed in K665 yeast . The transformation was assessed to have worked by PCR ( Figure S4A ) and , as with PfCAX and sPfCAX , sTgCAX suppressed the Ca2+ hypersensitivity of the yeast ( Figure S4B ) . These data demonstrate that apicomplexan CAXs and truncated variants are functional in yeast , and in this heterologous system , can function in providing Ca2+ tolerance . To confirm that growth of PfCAX-expressing yeast on Ca2+-containing media was due to enhanced vacuolar Ca2+/H+ exchange activity , ΔpH-dependent 45Ca2+ uptake in the presence of the Ca2+-ATPase inhibitor vanadate was examined in vacuolar membrane vesicles isolated from K665 yeast expressing each of the cax plasmids . ΔpH across the vacuolar membrane vesicles was established by activation of endogenous H+-ATPase activity by the addition of Mg2+-ATP . Significant Ca2+/H+ exchange activity , which could be inhibited by the protonophore carbonyl cyanide 3-chlorophenyl hydrazone , CCCP , was measured for each CAX variant ( Figure 3 ) . However , sPfCAX activity was reduced compared with that of sCrCAX1 ( by 42% at the 12 min time point; 0 . 662±0 . 032 versus 1 . 142±0 . 076 nmol mg protein-1 , after subtraction of basal accumulation; p = 0 . 0004 , unpaired , two-tailed Student's t-test; n = 5 ) . Ca2+/H+ exchange activity mediated by full-length PfCAX was significantly reduced compared with that of sPfCAX ( 36% reduced Ca2+ uptake at 12 min time point; 0 . 425±0 . 063 versus 0 . 662±0 . 032 nmol mg protein−1 , after subtraction of basal accumulation; p = 0 . 01 , unpaired , two-tailed Student's t-test; n = 5 ) but was detectable over basal ( empty vector ) levels ( p = 0 . 0007 , unpaired , two-tailed Student's t-test; n≥4 ) . Previously , transient expression of a green fluorescence protein ( GFP ) -tagged version of pfcax , using a strong but non-authentic/non-sequential promoter , was used to localise PfCAX to the parasite mitochondrion [7] . To enable expression and localisation of a Plasmodium CAX to be determined over the parasite's full life cycle , the rodent parasite , P . berghei , was used . A GFP-tagged P . berghei parasite line was generated by endogenous C-terminus fusion of the gfp sequence onto pbcax , leaving control of gene expression under the endogenous promoter . This was achieved with a single crossing over homologous integration approach ( Figure S5A ) . PCR on genomic DNA obtained from a tagged clone was positive for integration of the tagging construct ( using primers INT N43tag and ol492; Table S1 , Figure S5B and Protocol S1 ) and pulse field gel electrophoresis demonstrated that the integration occurred , as expected , on chromosome 1 ( Figure S5C and Protocol S1 ) . Expression of PbCAX-GFP in activated gametocytes was confirmed by Western blotting in the particulate fraction ( contain membranes ) , using anti-GFP polyclonal antibodies ( Figure S5D and Protocol S1 ) . A protein band of ∼77 kDa was identified in samples of PbCAX-GFP expressing parasites , which corresponds to the predicted mass of the PbCAX-GFP fusion protein . No smaller bands could be identified , suggesting that the tagged protein remains intact in situ . As a positive control for blotting and antibody staining , a P . berghei line that constitutively expresses GFP in the cytosol throughout the life cycle was used [18] . An appropriate protein band of approximately 29 kDa ( the predicted size of GFP ) was observed in the supernatant fraction derived from these parasites ( note that a similarly sized band was also present in the particulate fraction , which may represent contamination from the supernatant ) . Using epi-fluorescence microscopy in an initial screen of live parasite stages , only very low level , diffuse fluorescence signal was observed , predominantly , in asexual blood stages of pbcax-gfp transgenic parasites , with stronger parasite-associated GFP signal observed only on rare occasions ( data not shown ) . Stronger GFP signal was observed in female but not male gametocytes and the same was true for gametes ( Figure S6A ) . Thus , the asexual parasites with stronger GFP signal may have been immature female gametocyte stages . Good GFP signal was also observed in zygotes , ookinetes and oocysts . To improve resolution of live parasite fluorescence images , deconvolution microscopy was used . In activated gametocytes , GFP signal was confined to membranous regions surrounding the parasite nucleus and a mass to the side ( possibly the endoplasmic reticulum ) . Little signal colocalised with MitoTracker , used as a marker for parasite mitochondrion ( Figure 4 ) . Non-converting female gametes/zygotes and ookinetes 24 h after activation had more dispersed GFP signal and only very partial colocalisation with the mitochondrion could be observed ( Figure 4 ) . Additional images of the latter stages co-stained with a parasite surface marker rather than MitoTracker can be seen in Figure S6B . To determine the physiological importance of Ca2+/H+ exchange activity during the erythrocytic stages of P . berghei , a direct pbcax knock-out strain was generated by double crossing over homologous recombination ( Figure S7A ) . After transfection of P . berghei ANKA and a second line that constitutively expresses cytosolic GFP [18] with a PbCAX knock-out vector , parasites containing the knock-out construct were selected by pyrimethamine drug pressure . PCR on genomic DNA obtained from two independent clones derived from each of the parasite lines ( cl9 and cl5 gfp ) were both positive for integration of the knock-out construct ( using primers INT N43 and ol248 ) and disruption ( using primers N43 KO1 and N43 KO2 ) of pbcax ( Table S1 , Figure S7B and Protocol S1 ) . Further confirmation that the gene deletion was successful came from Southern hybridisation of digested genomic DNA ( Figure S7C and Protocol S1 ) and pulse field gel electrophoresis , using a probe that recognises the resistance marker ( Figure S7D and Protocol S1 ) . The latter demonstrates that the integration occurred on the correct chromosome ( i . e . chromosome 1 ) . The ability to derive these clones and their apparently normal in vivo asexual blood-stage growth and gametocyte production ( not tested quantitatively ) suggest that PbCAX function is not essential for blood stages of the P . berghei life cycle . Next , the in vitro sexual development of mutant parasite clones was studied [19] . Firstly , the ability of male gametocytes to undergo exflagellation was tested by adding gametocyte containing blood to “activation” ookinete medium ( see Methods and Materials ) and counting exflagellation centres . Figure 5A presents data , using cl9 Δpbcax parasites , in which no statistical difference was determined between the number of exflagellation centres produced by wild-type parasites compared with the mutant line ( 8 . 4±0 . 6 versus 9 . 5±0 . 3; mean ± SEM; n = 3; p = 0 . 2 , unpaired , two-tailed Student's t-test ) . Apparently normal exflagellation was also observed for cl5 gfp Δpbcax parasites ( not tested quantitatively ) . Secondly , ookinete conversion was determined by culturing parasites for 24 h in ookinete medium and measuring the number of the motile ookinete forms in relation to unconverted “round” forms ( female gametes/zygotes ) at the end of this period , using fluorescently-labelled P28 antibody to aid identification ( note that P28 is a surface expressed antigen found on activated female gametes , zygotes and ookinetes ) . Figure 5B presents the data from three experiments using cl9 Δpbcax parasites in which no ookinete conversion was observed , while wild-type control parasites converted to ookinetes on average by 74±3% ( mean ± SEM; n = 3 ) . No ookinete conversion was also observed for cl5 gfp Δpbcax parasites ( n = 4 ) , while in two matched experiments , using wild-type control parasites that constitutively express GFP , estimated conversion rates were 87 and 71% . Note that very occasional ookinete-like forms were observed in these in vitro cultures . Figure 6 shows examples of the typical phenotype that was observed for the mutant parasites . At 8 h post-activation , wild-type parasites were predominantly “retort” forms ( a round parasite that contains the nucleus , with an apical protrusion ) , whereas Δpbcax parasites remained round . At 24 h post-activation , wild-type parasites were predominantly fully converted into ookinetes ( elongated forms with the nucleus in the centre ) . At this time , Δpbcax parasites were still round in form , fewer in number and were often smaller in size and had degenerated membranes ( as judged from discontinuous P28 staining ) . Furthermore , they often had diffuse nuclei , suggestive of possible necrosis or late stage apoptosis . In a single semi-quantitative experiment with three repeats , the number of “round” form PbCAX knock-out parasites per field of view was counted in ookinete development cultures at 2 , 6 and 24 h after activation ( Figure S8 ) . An approximate 70% reduction in the number of parasites present after 24 h was observed , suggesting that the parasites are degrading rather than arrested or slow growing . Cross-fertilization experiments were performed in an effort to establish if PbCAX activity is an important male and/or female trait ( Figure 5C ) . This was performed by co-culturing Δpbcax cl9 gametocytes with gametocytes of either the female-defective nek-2 or nek-4 null mutants [20] , [21] or a male-defective map-2 null mutant [22] and determining ookinete conversion . While none of the mutants alone underwent ookinete conversion , crossing Δpbcax with Δnek2 or Δnek4 resulted in recovered ookinete conversion of 14±1 and 13±2% , respectively ( mean ± SEM; n = 3 ) . This is due to the fertilisation of female Δpbcax cl9 parasites by functional Δnek2 or Δnek4 male gametes . However , crossing Δpbcax cl9 with Δmap2 resulted in recovered ookinete conversion of 42±2% ( mean ± SEM; n = 3 ) . This is due to the fertilisation of functional Δmap2 female gametes by male Δpbcax parasites . These data suggest that PbCAX activity , while not specific , is predominantly important to female gametes . Parasite development was assessed in vivo by feeding Anopheles stephensi mosquitoes on mice infected with Δpbcax cl9 or cl5 gfp parasite mutants and measuring oocyst formation in the mosquito midgut 14 days later . Figure 5D shows three independent experiments , the first using Δpbcax cl9 and the second two using Δpbcax cl5 gfp parasites ( with 20 mosquito midguts analysed per assay ) . It demonstrates that while mosquito infection with wild-type parasites produced mean numbers of oocysts per mosquito midgut of 110 , 144 and 57 , respectively , and infection prevalences ( number of mosquitoes with oocysts ) of 80 , 90 and 65% , infection with parasites lacking PbCAX produced none . At day 21 post-infection , the infected mosquitoes were used to re-infect mice but no mutant asexual blood-stage parasites could be identified after 15 days , while wild-type asexual blood-stage parasites were observed by day 5 on average . Given that transmission stages of parasite development are extracellular ( the in vitro culture medium used here contained 0 . 42 mM Ca2+ ) and that CAX activity often provides a mechanism to enable organisms to survive in the presence of extracellular Ca2+ [14] , we tested the ability of the Ca2+ chelator EGTA ( ethylene glycol tetraacetic acid ) to restore ookinete conversion of both Δpbcax cl9 and cl5 gfp parasites ( Figure 7 ) . EGTA ( 10 mM ) had no significant effect on the normal ookinete conversion efficiency of wild-type parasites , which constitutively expressed GFP ( control parasites used for comparison with Δpbcax cl5 gfp; open triangles in Figure 7 ) , measured at 24 h post-activation when added immediately prior to gametogenesis ( t = 0 h ) , 30 min post-activation ( at which point exflagellation has occurred ) , 2 h post-activation ( at which point zygote formation will be complete , predominantly ) or 3 h post-activation ( p>0 . 05 , ANOVA with Dunnett's post test; n = 3 ) . Apparently similar results were found for the development of non-GFP expressing wild-type parasites ( control parasites used for comparison with Δpbcax cl9; open circles in Figure 7 ) grown in the presence of EGTA , although this could not be tested statistically as only 2 control experiments were performed . The effect of EGTA on Δpbcax cl9 and cl5 gfp parasites was to restore ookinete conversion , when added 3 h post-activation or prior to this point , although the ability to restore ookinete conversion reduced the longer after activation EGTA was added . In the case of all Δpbcax cl9 experiments and one of the three using Δpbcax cl5 gfp parasites , addition of EGTA at 0 and 0 . 5 h led seemingly to complete restoration of ookinete conversion . Images of restored Δpbcax cl9 ookinetes in the presence of EGTA added at t = 0 h can be seen in the lower panels of Figure 6 . Addition of EGTA at 3 h post-activation still enabled ookinete conversion of Δpbcax cl9 and cl5 gfp parasites but at significantly lower levels than the GFP-expressing wild-type control ( p<0 . 05 , ANOVA with Dunnett's post test; n = 3 ) . In case the rescuing effect of EGTA was osmotic rather than due to its Ca2+ chelation properties , additional controls were performed in the presence of 20 mM NaCl added to the culture medium immediately prior to gametogenesis . This had no effect on the ability of wild-type parasites to develop into ookinetes ( p = 0 . 6 , unpaired , two-tailed Student's t-test; n = 3 ) or the inability of mutant parasite to develop ( data not shown ) . In a highly complementary approach to allow comparative tagging and knock-out studies with P . berghei , additional experiments were performed using the genetically amenable apicomplexan , T . gondii , at the tachyzoite stage . When TgCAX was expressed transiently under the control of the tubulin promoter , as a second copy detectable by a C-terminal Ty-tag ( Figure S9A ) , the protein was found predominantly in a large vesicular-like compartment located in the apical end of the intracellular parasite , as well as in much smaller vesicle-like structures dispersed throughout the parasite cytosol ( Figure 8A ) . This compartment is reminiscent of the plant-like vacuole ( PLV ) described recently in extracellular tachyzoites [23] . However , there was no co-localization with antibodies directed against a marker for the PLV and acidocalcisomes , the vacuolar proton pyrophosphatase , VP1 ( Figure S9B ) . When TgCAX-Ty was stably expressed in a pool of intracellular parasites , only the dispersed signal was observed ( Figure 8A ) , similar but seemingly not identical to that reported for VP1 [23] . Antibodies raised against an N-terminal region of the protein colocalised with the transiently expressed TgCAX-Ty ( Figure 8B ) . However , in intracellular wild-type parasites no signal was detected by immunofluorescence , using the N-terminal directed antibodies . Although proteomic data from tachyzoite stage preparations include TgCAX ( e . g . [17] ) , it may not be expressed at a detectable level by IFA at this stage . To test this hypothesis , the endogenous TgCAX was tagged with a 3Ty-tag ( Figure S9C ) . This was introduced at the C-terminus of TgCAX by a knock-in strategy , using the ΔKU80 strain of T . gondii [24] , [25] . In this strain , the non-homologous end-joining DNA repair pathway is defective and therefore integration essentially occurs only by homologous recombination at the targeted locus . The integration was confirmed by RT-PCR ( Figure S9D ) but no signal was detected by immunofluorescence ( data not shown ) , supporting the hypothesis that the protein is expressed at a very low level during this stage of the parasite's life cycle . Furthermore , transiently expressed Ty-tagged TgCAX shows no colocalization with HSP70 or with co-transfected SOD2-GFP ( Figure 8C ) that label specifically the mitochondrion of T . gondii [26] , [27] . To determine the physiological importance of Ca2+/H+ exchange activity during the tachyzoite stage of T . gondii , a similar approach to that used for the study P . berghei was taken ( Figure S9E and S9F ) . Plaque assay experiments were performed and revealed no obvious defect in the lytic cycle of the ΔTgCAX strain compared with wild-type parasites after 7 days ( Figure 9A ) . To confirm this observation , intracellular growth and egress were assessed specifically . In the case of intracellular growth , the distributions of parasites per vacuole were essentially identical for both the ΔTgCAX and wild-type strains ( Figure 9B ) and parasite egress induced by the Ca2+ ionophore A23187 was also similar between strains ( Figure 9C ) . These data suggest that TgCAX is dispensable during the tachyzoite stage .
Nearly all cax genes characterised to date from bacteria , fungi and plants encode H+-coupled Ca2+ transporters [11] , [12] , [28] . While some CAX transporters are highly specific for Ca2+ , such as yeast Vcx1 , others have a broader ion specificity and can mediate the transport of different divalent cations [13] , [15] or can transport both divalent ( e . g . Ca2+ ) and monovalent ( e . g . Na+ ) cations , in the case of CrCAX1 [10] . Previous characterisation of PfCAX by expression in Xenopus oocytes reported Ca2+/H+ exchange and , using competition assays , data implicating transport of other divalent cations , including Mg2+ and transition metal ions [7] . Here , we developed an alternative approach to characterise PfCAX and TgCAX , taking advantage of the yeast expression system that is most commonly used to study CAXs ( e . g . [10] ) . Expression of PfCAX and ( N-terminally truncated ) sPfCAX and sTgCAX in a yeast strain lacking vacuolar Ca2+ transporters increased tolerance of yeast to extracellular Ca2+ . Transport measurements in yeast vacuolar membrane preparations confirmed that PfCAX is a Ca2+/H+ exchanger and is localised to the tonoplast when expressed in yeast . In a previous study [7] there was no physiological interaction of PfCAX with Na+ , as would be predicted because sequences that maybe associated with Na+ transport [10] are lacking in the PfCAX sequence unlike CrCAX1 . In agreement with previous work demonstrating that the N-terminal region of plant CAXs is autoinhibitory [29] , N-terminal truncation of PfCAX improved Ca2+ tolerance and Ca2+ transport rates . Truncation of the protein may also improve delivery to the vacuole and provide an alternative explanation for improved functionality . While this paper was under revision , a report describing PfCAX/CHA expression in yeast lacking the endogenous CAX gene ( vcx1 ) and containing a calcium biosensor ( apoaequorin ) was published [30] . Although no new functional data were reported and only indirect transport measurements were possible , unlike the data presented here ( see Figure 3 ) , the data produced , using this system , are consistent with PfCAX functioning as a Ca2+/H+ exchanger and lend support to the present findings . While functional characterisation of proteins , whether in situ or in heterologous expression systems , provides an understanding of what a protein does , localisation of proteins in situ adds physiological context . Most plant and fungal CAXs localise to the tonoplast and provide efflux pathways to remove Ca2+ from the cytosol . In the case of PfCAX , Xenopus oocyte [7] and yeast ( this study and [30] ) heterologous expression data suggest plasma membrane and tonoplast localisation , respectively . However , tagging studies indicate that PfCAX is located in the parasite's mitochondrial inner membrane [7] , a novel CAX localisation . Given these complexities , additional expression and localisation studies were performed in T . gondii and P . berghei parasites . While proteomic analysis suggests that TgCAX is present in intracellular T . gondii tachyzoites ( e . g . [17] ) , tagging of the gene under control of the endogenous promoter did not allow localisation , as expression may be limited at this stage . However , use of a strong promoter by transient transfection enabled localisation of TgCAX , mainly to a large vesicular-like region , with similar size , shape and location to the PLV , as described previously [23] . Considering that the staining pattern covered the entire region of the vesicle-like structure , rather than just the edge ( as might be expected for integral proteins such as CAXs ) , the lack of co-localisation with VP1 ( a marker of both the PLV and acidocalcisomes ) and the loss of this staining pattern over time , it may be that these staining results are an artifact of overexpression . There was also no evidence for mitochondrial localisation of TgCAX ( see Figure 8 ) , as there is for PfCAX [7] . Stably transfected parasites contained punctate staining for tagged TgCAX . The punctate staining pattern with TgCAX-Ty , although consistent with what is observed for acidocalcisomes , did not co-localise with VP1 . This suggests that TgCAX-Ty may localise to a separate novel vacuolar compartment such as one containing cathepsin L [31] and will benefit from further studies . Similar to TgCAX , there is proteomic evidence that PfCAX is expressed during the asexual blood stage ( e . g . [17] ) . Here , the expression of PbCAX-GFP was barely detectable in asexual blood stages . This is consistent with the low level of expression of TgCAX and the apparently inessential nature of PbCAX and TgCAX at similar life cycle stages . PbCAX-GFP expression was much greater in gametocytes ( predominantly females ) and this expression profile was maintained throughout the sexual stages . Localisation altered during sexual development . In activated female gametes , the data are consistent with a build-up of PbCAX-GFP within a large membranous network that is most likely the endoplasmic reticulum ( although other secretory vesicles cannot be ruled out ) , with little if any mitochondrial localisation . In zygotes and ookinetes ( stages when PbCAX is likely essential , as demonstrated here ) , PbCAX-GFP localisation was dispersed intracellularly across the parasite , with the majority not localised to mitochondria . At these stages a proportion of the observed PbCAX-GFP signal may associate with its production and trafficking , or with mislocalised protein , although if PbCAX-GFP was mislocalised during ookinete conversion then it would not allow development to proceed ( in keeping with the knock-out results presented here ) . PfCAX has a predicted mitochondrial targeting sequence at residues 11–18 ( YVRRTISQ ) , consistent with mitochondrial localisation [7] , and this is conserved throughout the apicomplexan CAXs . Interestingly , this protein sequence has been identified in phosphoproteomic studies [17] , using preparations derived from mature trophozoite-infected erythrocytes , and two of these residues , T15 and S17 , are putative phospho-acceptor sites ( with ascores of 1000 , suggesting the annotations have a high degree of confidence ) . Phosphorylation of the S17 residue of PbCAX has also been reported , in a similar study , using ookinete preparations ( available on GeneDB ) . The homologous protein region of TgCAX was also identified in tachyzoite preparations , although it contained no phosphorylated sites [17] . However , neighbouring residues at positions S26 , S27 and T46 were identified as being phosphorylated , albeit with lower ascores of 19 , 13 and 6 , respectively . Previous work has demonstrated that phosphorylation of the mitochondrial signal sequence of 2′ , 3′-cyclic nucleotide-3′-phosphodiesterase 2 ( CNP2 ) alters its localisation so that it is retained in the cytoplasm [32] and its possible a similar mechanism could alter the location of apicomplexan CAXs . However , the localisation evidence presented here provides little evidence for mitochondrial function of CAXs in apicomplexan parasites . Neither TgCAX nor PbCAX are required for asexually reproducing parasites . Most active Ca2+ transport pathways , including in apicomplexan parasites , play critical roles in providing Ca2+ tolerance , by refilling internal Ca2+ stores , and by shaping cytosolic free Ca2+ transients or oscillations that act as intracellular signals [33] . Previously identified Ca2+ efflux pathways include the plasma membrane and internally localised P . falciparum Ca2+-ATPase , PfATP4 [34] , [35] , the SERCA-type pumps at the endoplasmic reticulum in P . falciparum and T . gondii , PfATP6 and TgSERCA , respectively [5] , [6] , and the non SERCA-type T . gondii Ca2+-ATPase , TgA1 , that localises to acidocalcisomes [36] , [37] . Currently , the in vivo function of each Ca2+ efflux transporter ( with respect to the cytosol ) is unclear , although it is likely that there would be some overlap in the functions of these pathways . Thus , loss of one active Ca2+ transport mechanism might be compensated for by others , particularly during asexual development stages when the parasites are mainly intracellular , and do not face large extracellular Ca2+ concentrations [38] . In yeast and plants , Ca2+/H+ exchangers play major roles in providing tolerance to excess cytosolic Ca2+ , due to their ability to function as low affinity , high capacity Ca2+ transporters . For example , in yeast the Ca2+/H+ exchanger Vcx1 efficiently sequesters Ca2+ into the vacuole when cytosolic Ca2+ levels are high without requiring the vacuolar Ca2+-ATPase [39] . However , deletion of vacuolar Ca2+/H+ exchange activity , such as in the Arabidopsis cax1 knock-out , leads to up-regulation of vacuolar Ca2+-ATPase activity , as a compensatory mechanism [40] that may also manifest itself in apicomplexans . Interestingly , PbCAX was essential during ookinete conversion , with knock-out lines failing to transform from zygotes and over an extended time period ( in excess of 8 h ) becoming necrotic or possibly apoptotic . Given that this phenotype could be reversed by the removal of extracellular Ca2+ , this suggests that PbCAX's primary role is to provide a tolerance mechanism to environmental Ca2+ at a time when the parasite is extracellular and , at least initially , exposed to plasma concentrations of free Ca2+ of approximately 1 . 2 mM ( it should be noted that our understanding of changes in the ionic environment in the mosquito midgut is very limited , presently ) . This is consistent with its primary role in plants and fungi [14] . Evidence is accumulating that Ca2+ exchangers play a role in directly regulating cytosolic Ca2+ elevations and in modulating cellular signaling associated with stress responses [33] . In the case of the sexual stages of Plasmodium parasites , it is well known that Ca2+ and Ca2+ interacting proteins regulate important parasite functions such as gametocyte differentiation into gametes , ookinete development and motility [41]–[45] . As PbCAX knock-out lines failed to transform from zygotes and long before becoming necrotic , it is tempting to speculate that PbCAX directly regulates the signaling pathways involved in zygote differentiation but further studies are required to answer this hypothesis . Regulation of Ca2+/H+ exchangers , via protein modification or protein interaction , is consistent with their role as Ca2+ modulators . The Arabidopsis Ca2+/H+ exchanger AtCAX1 can be regulated via an autoinhibitory domain that is present on its long hydrophilic N-terminal tail [29] . Transport activity of AtCAX1 may be activated by phosphorylation or interaction with an activator protein at this domain [29] , [46] . CrCAX1 from Chlamydomonas may also share this mechanism of regulation . This CAX protein has an extended hydrophilic N-terminal tail which when truncated leads to increased Ca2+/H+ exchange activity [10] . The CAX sequences from P . falciparum and other apicomplexan organisms likewise have a long N-terminal tail ( Figure 1 and S1 ) and N-terminal truncation of PfCAX ( to give sPfCAX ) led to an increased Ca2+ tolerance phenotype in yeast ( Figure 2 ) and enhanced Ca2+/H+ exchange activity ( Figure 3 ) . This may indicate that PfCAX ( and by association other apicomplexan CAXs ) has the potential to be regulated by an analogous mechanism . Phosphoproteomic data support this by suggesting the N-termini of PfCAX , PbCAX and TgCAX are phosphorylated ( [17] and GeneDB ) . The S . cerevisiae Ca2+/H+ exchanger Vcx1 does not have an extended N-terminal tail and is regulated by the Ca2+-dependent phosphatase calcineurin ( PP2B ) [11] . Functional calcineurin is absent in higher plants which may explain why regulatory mechanisms differ between plant and yeast CAX proteins , but a calcineurin phosphatase which requires Ca2+/calmodulin and is inhibited by immunosuppressant drugs ( FK506 and cyclosporine A ) is present in P . falciparum [47] . Some of the potential roles of calcineurin in P . falciparum are being elucidated [48] , although its signaling roles and possible involvement in Ca2+ homeostasis including Ca2+ transporter regulation are unknown , but suggest alternative mechanisms of regulation and are worthy of investigation . Dysregulation of CAXs may prove equally as detrimental to plasmodial parasite survival at other life cycle stages , as PbCAX deletion does during ookinete conversion , and to other apicomplexans . The data presented here confirm the functional Ca2+/H+ exchange activity of apicomplexan CAXs , shed light on possible CAX regulation , fail to support substantial mitochondrial localisation , demonstrate that CAXs are neither expressed to a high level nor required during asexual apicomplexan parasite development but PbCAX is expressed to a measureable level during sexual stages of development and is likely essential during ookinete conversion by protecting parasites against extracellular Ca2+ . In a new age of malaria eradication there is renewed interest in the development of transmission blocking therapies [49] . Most antimalarial treatments fail to kill circulating sexual parasite stages and thus do not stop transmission . Therefore , there is increasing interest in developing drugs and vaccines that can stop the transmission process . Here , PbCAX has been shown to be essential for parasite transmission . Furthermore , plasmodial CAXs are single copy genes with no close paralogues and CAXs are not found in higher animals . At present there are no potent and specific CAX inhibitors , KB-R7943 ( a first-generation inhibitor of Na2+/Ca2+ exchangers ) being the best known [7] . Our study has identified a new transmission blocking target , which is required for ookinete development by protecting the parasite from environmental Ca2+ , and provides the experimental tools necessary to aid development of this therapeutic strategy .
All animal work has passed an ethical review process and was approved by the United Kingdom Home Office . Work was carried out in accordance with the United Kingdom Animals ( Scientific Procedures ) Act 1986 and in compliance with European Directive 86/609/EEC for the protection of animals used for experimental purposes . All procedure were performed under Home Office licence number 40/3344 . A synthetic version of pfcax cDNA was synthesised ( Geneart ) and codon optimised for expression in S . cerevisiae ( Figure S3 ) . The optimisation resulted in an increase in average GC content from 28% to 32% . This pfcax cDNA was sub-cloned into the XbaI and SacI sites of the yeast expression vector piHGpd . N-terminal truncated pfcax and TgCAX variants ( spfcax and sTgCAX ) , encoding proteins lacking the first 60 and 98 amino acids , respectively , were generated by PCR amplification using synthetic pfcax or T . gondii cDNA as template and the primers sPfCAXF/sTgCAXF and PfCAXR/TgCAXR ( Table S1 ) . The spfcax cDNA was cloned into the pGEM-T vector , while sTgCAX cDNA was cloned into the Strataclone vector for propagation and sequencing . spfcax was then sub-cloned into piHGpd , as above , while sTgCAX was subcloned into piUGpd . Plasmids were transformed into the S . cerevisiae strain K665 ( pmc1::TRP1 vcx1Δ ) [11] , using the lithium acetate/polyethylene glycol method [50] . Transformed yeast colonies were grown in synthetic defined medium minus histidine or uracil and tryptophan for selective growth of the plasmid and maintenance of the insertional mutation . Expression of pfcax in yeast was confirmed by RT-PCR following extraction of yeast total RNA , using the acid phenol method [51] . Extracted RNA was further purified by phenol/chloroform/isoamyl alcohol extraction and isopropanol precipitation . First strand cDNA was produced from 1 µg of DNase-treated total RNA , using Superscript II reverse transcriptase ( Invitrogen ) and an oligo-dT/PfCAXR primer mix . PCR was performed using sPfCAXF/PfCAXR primers and yeast actin primers ( Table S1 ) , as a constitutive control . Transformation of sTgCAX in yeast was confirmed by PCR following extraction of genomic DNA . PCR was performed using the primers PUGF and PUGR ( found approximately 200 base pairs either side of the multicloning site in piUGpd ) and yeast actin primers ( Table S1 ) . Ca2+ tolerance assays of K665 yeast expressing the synthetic pfcax and wild-type sTgCAX plasmids were performed on solid growth media and in liquid media . The previously characterised Ca2+/H+ exchanger sCrCAX1 [10] was used as a positive control for comparison with PfCAX and sPfCAX . Serial dilutions of yeast were grown at 30°C on solid YPD medium containing a 50 mM CaCl2 concentration . For determination of yeast growth rate in liquid media , yeast strains of the same starting cell density were inoculated in YPD medium containing a range of CaCl2 concentrations and grown at 30°C , shaking for 16 h in 24-well flat bottomed plates , and cell growth was determined by measuring absorbance at 600 nm . Vacuolar membrane vesicles were isolated from yeast cells expressing cax plasmids , by purification of the microsomal fraction through a two-step sucrose gradient , as described previously [52] . By testing hydrolytic activity of the V-type H+-ATPase ( V-ATPase ) the isolated tonoplast fractions were all demonstrated to have no measurable contamination from other membrane fractions , as only the V-ATPase inhibitor bafilomycin ( a tonoplast marker ) inhibited H+-ATPase activity in these vesicles , while inhibitors of other membrane-localised H+-ATPases did not inhibit H+-ATPase activity ( data not shown ) . Ca2+/H+ exchange activity was determined by measuring pH gradient-dependent 45Ca2+ uptake into membrane vesicles , as described previously [53] . Transfection experiments were performed on P . berghei ANKA strain 2 . 34 parasites , as previously described [54] . The pbcax knock-out vectors were constructed for a double cross-over homologous recombination in the pBS-DHFR plasmid that contains a Toxoplasma gondii dhfr/ts cassette conferring resistance to pyrimethamine [55] , [56] . The knock-out construct was generated by inserting 507 bp of the pbcax 5′ untranslated ( UTR ) region upstream and 497 bp of the pbcax 3′ UTR region downstream of the dhfr cassette ( sequences of primers , N0431-4 , used to amplify fragments from P . berghei genomic DNA are given in Table S1 ) . The final knock-out construct was digested with ApaI and NotI to release the fragment for transfection . Transfection of P . berghei parasites with the knock-out construct was carried out in both wild-type parasites and in a line that constitutively expresses cytosolic GFP [18] . To generate a pbcax-gfp construct for a single cross-over homologous recombination , a 0 . 9 kb region of the pbcax gene without the stop codon was inserted in frame and upstream of the gfp sequence in the plasmid p277 containing the human dhfr cassette and conveying resistance to pyrimethamine [19] . Prior to transfection , the final construct was digested with PacI . This cuts the plasmid in the middle of the insert , which is optimal for the homologous recombination event . For the analysis of PbCAX-GFP localisation during the parasite's life cycle , images of GFP-expressing parasites were captured with a Zeiss AxioImager M2 ( Carl Zeiss , Inc ) microscope fitted with an AxioCam ICc1 digital camera ( Carl Zeiss , Inc ) . Hoechst 33342 ( Sigma ) was used for nuclear staining of all stages and the P28 cy3-labelled antibody was used as a marker for female gametes/zygotes/ookinetes . High resolution live cell imaging was performed using an Olympus-based personal Delta Vision work station at ×100 ( numerical NA 1 . 4 , oil ) . Subsequent off-line image preparation was carried out using Applied Precision software and finalised with Adobe Photoshop . Images presented are 2D projections of 0 . 1 µm stepped Z-stacks . For ookinete conversion assays , blood was taken by cardiac puncture from P . berghei-infected mice on day 4 post-infection into heparinised syringes , mixed with ookinete culture medium ( RPMI1640 culture medium containing 25 mM HEPES , 25% ( v/v ) fetal bovine serum , 10 mM sodium bicarbonate , 50 µM xanthurenic acid , pH 7 . 6 ) and cultured at 19°C for a further 21–24 h before assessment of conversion . For direct immunolabelling ( to aid identification ) , cultured cells were pelleted for 2 min at 800× g and then labelled for 10 min on ice in 50 µl of ookinete medium containing Hoechst 33342 and Cy3-conjugated mouse monoclonal antibody specific for P28 [21] . For exflagellation assays , mice were infected as described above . On day 4 to 5 post-infection , 10 µl of infectious tail snip blood was added to 40 µl of ookinete medium and incubated at room temperature for 15 mins . Gametocytaemia was 5–8% . The data were generated from the 15 min time point from 1 spot of tail blood performed in triplicate ( from 3 different infected mice ) . An aliquot of 10 µl of tail blood was also analysed between 7 and 10 min in triplicate to ensure no exflagellation events were missed prior to microgametogenesis and in each case none were observed . Exflagellation was counted by measuring the number of exflagellation centres in 8 fields under magnification ×40 on a Zeiss Primostar microscope . Parasite transmission to mosquitoes in vivo was assessed using mosquitoes fed directly on P . berghei-infected mice , as described previously [56] . Briefly , infected mice were offered to overnight-starved A . stephensi ( SD 500 strain ) mosquitoes in groups of approximately 100 for 30 min . Unfed mosquitoes were removed the next day , and the remaining mosquitoes were maintained at 19°C and 80% relative humidity on a 12-h light/dark cycle , being fed on a 2% ( w/v ) D-glucose solution that was replenished every 2 to 3 days . At 14 day post-feeding , mosquito midguts were dissected and analysed by fluorescence microscopy . T . gondii tachyzoites ( RHhxgprt- and ΔKU80hxgprt- strains [24] , [25] ) were grown in confluent human foreskin fibroblasts ( HFF ) maintained in Dulbecco's Modified Eagle's Medium ( DMEM , GIBCO , Invitrogen ) supplemented with 10% ( v/v ) fetal calf serum , 2 mM glutamine and 25 µg/ml gentamicin . Genomic DNA was prepared from tachyzoites ( RH strain ) , using the Wizard SV genomic DNA purification system ( Promega ) . Total RNA was isolated from tachyzoites , using Trizol ( Invitrogen ) , and then total cDNA was generated by RT-PCR , using Superscript II reverse transcriptase ( Invitrogen ) according to the manufacturer's instructions . All amplifications were performed with LaTaq or ExTaq ( TaKaRa ) and the primers used are listed in Table S1 . The full-length cDNA of TgCAX was amplified with primers pairs TgCAX-1/TgCAX-2 , cloned into the pGEM-Teasy vector ( Promega ) and then sub-cloned into pTUB8MycGFPPfMyoAtailTy-HX [57] between the EcoRI and NsiI sites to create the pTUB8TgCAX-Ty vector . For the knock-in vector , a fragment of genomic DNA corresponding to the C-terminal part of the gene was amplified using primers TgCAX-3/TgCAX-2 and cloned between KpnI and NsiI sites of the pTUB vector [57] modified to introduce 3 Ty-tags at the C-terminal end of the insert . This construct was then used to generate the knock-out vector by introducing a fragment of the 5′UTR of TgCAX amplified with primers TgCAX-4/TgCAX-5 and cloned between the BamHI and NotI sites of the knock-in vector . Parasite transfections were performed by electroporation as described previously [58] . The hypoxanthine-xanthine-guanine phosphoribosyl transferase ( hxgprt ) gene was used as a positive selectable marker in the presence of mycophenolic acid ( 25 mg/ml ) and xanthine ( 50 mg/ml ) as described before [59] . For immunofluorescence assays ( IFA ) , parasite-infected HFF cells were fixed with 4% ( v/v ) paraformaldehyde/0 . 05% ( v/v ) glutaraldehyde ( PFA/GA ) in phosphate buffered saline ( PBS ) and processed as described previously [60] . For generation of specific antibodies against TgCAX , the peptide NH2-GAPSRQLHLGLLSEGW-COOH was used to immunise two guinea pigs ( Eurogentec ) according to their standard protocol . For plaque assays , host cells were infected with parasites for 7 days before fixation with PFA/GA . Giemsa staining was then performed as described previously [61] . For intracellular growth assays , HFF cells were inoculated with parasites and grown for 24 h before fixation with PFA/GA . IFAs were performed using α-gliding associated protein ( GAP ) 45 antibodies ( a surface parasite marker ) . The number of parasites per vacuole was determined by counting the parasites in 100 vacuoles in duplicate in three independent experiments . For egress assays , host cells were inoculated with freshly released parasites and grown for 30 h . Parasite-infected host cells were then incubated for 5 min at 37°C with DMEM containing 0 . 06% ( v/v ) dimethyl suphoxide ( DMSO ) or 3 µM of the Ca2+ ionophore A23187 from Streptomyces chartreusensis ( Calbiochem ) before fixation . IFAs were performed using α-GAP45 antibodies and the average number of egressed vacuoles was determined by counting 100 vacuoles for each condition in at least 2 independent experiments . Unpaired , two-tailed Student's t-tests or one-way analysis of variance ( ANOVA ) with Dunnett's post-hoc test were performed , as noted in the text , using the GraphPad Prism software program ( version 5 for PC ) .
|
Calcium is vital to all living organisms . It is used within cells to regulate many essential processes and , because of this , its cellular concentration is tightly controlled . To change cellular calcium levels , cells use calcium transport proteins . These proteins can alter calcium concentration by moving calcium into or out of the cell or specialised calcium storage compartments within the cell . We know little about how single-celled apicomplexan parasites , including Plasmodium ( the causal agent of malaria ) and Toxoplasma ( the causal agent of toxoplasmosis ) , regulate their calcium levels . Here , we have demonstrated that removing apicomplexan genes for a protein that exchanges calcium for protons across membranes ( a Ca2+/H+ exchanger ) and a member of the cation exchanger ( CAX ) family , does not affect the survival of parasites during those stages when they live within host cells . It is , however , lethal for the mouse malaria P . berghei when the parasite is free living within its mosquito vector . When we removed calcium from around the parasites at this stage they were able to develop normally , suggesting that the protein provides a mechanism for the parasite to tolerate environmental calcium . Learning how this calcium transport protein impacts on the development of apicomplexan parasites may lead to the development of novel anti-parasitic interventions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"toxoplasma",
"gondii",
"microbiology",
"plasmodium",
"falciparum",
"parasitic",
"diseases",
"parasitology",
"parasite",
"physiology",
"parastic",
"protozoans",
"calcium",
"signaling",
"infectious",
"diseases",
"signaling",
"in",
"cellular",
"processes",
"biology",
"signal",
"transduction",
"tropical",
"diseases",
"(non-neglected)",
"protozoology",
"malaria",
"molecular",
"cell",
"biology"
] |
2013
|
The Plasmodium berghei Ca2+/H+ Exchanger, PbCAX, Is Essential for Tolerance to Environmental Ca2+ during Sexual Development
|
For genome-wide association studies in family-based designs , we propose a new , universally applicable approach . The new test statistic exploits all available information about the association , while , by virtue of its design , it maintains the same robustness against population admixture as traditional family-based approaches that are based exclusively on the within-family information . The approach is suitable for the analysis of almost any trait type , e . g . binary , continuous , time-to-onset , multivariate , etc . , and combinations of those . We use simulation studies to verify all theoretically derived properties of the approach , estimate its power , and compare it with other standard approaches . We illustrate the practical implications of the new analysis method by an application to a lung-function phenotype , forced expiratory volume in one second ( FEV1 ) in 4 genome-wide association studies .
During the analysis phase of genome-wide association studies , one is confronted with numerous statistical challenges . One of them is the decision about the “right” balance between maximization of the statistical power and , at the same time , robustness against confounding . In family-based designs , the possible range of analysis options spans from a traditional family-based association analysis [1]–[4] , e . g . TDT , PDT , FBAT , to the application of population-based analysis methods that have been adapted to family-data [1]–[3] . While , by definition , the first group of approaches is completely immune to population admixture and model misspecification of the phenotype , and can be applied to any phenotype that is permissible in the family-based association testing framework ( FBAT [4]–[6] ) , the second category of approaches maximizes the statistical power by a population-based analysis . The phenotypes are modeled as a function of the genotype , and population-based methods such as genomic control [7] , [8] , STRUCTURE [9] and EIGENSTRAT [10] , are applied to account for the effects of population admixture and stratification . Hybrid-approaches that combine elements of both population-based and family-based analysis methods , e . g . VanSteen algorithm [11] and Ionita weighting-schemes [12] , [13] have been suggested to bridge between the 2 types of analysis strategies . Contrary to the other methods that combine family data and unrelated samples [14]–[17] , such hybrid testing strategies maintain the 2 key features of the family-based association tests: The robustness against confounding due to population admixture and heterogeneity , and the analysis flexibility of the approach with respect to the choice of the target phenotype . Such 2-stage testing strategies utilize the information about the association at a population-level , the between-family component , to prioritize SNPs for the second step of the approach in which they are tested formally for association with a family-based test . The hybrid approaches can achieve power levels that are similar to approaches in which standard population-based methods are applied to family-data , but the optimal combination of the 2 sources of information ( the between-family component and the within-family component ) is not straightforward in the hybrid approaches . In this communication , we propose a new family-based association test for genome-wide association studies that combines all sources of information about association , the between and the within-family information , into one single test statistic . The new test is robust against population-admixture even though both components , the between and the within-family components , are used to assess the evidence for association . The approach is applicable to all phenotypes or combinations of phenotypes that can be handled in the FBAT-approach , e . g . binary , continuous , time-to-onset , multivariate , etc [4]–[6] , [18] . While the correct model specification for the phenotypes will increase the power of the proposed test statistic , misspecification of the phenotypic model does not affect the validity of the approach . Using extensive simulation studies , we verify the theoretically derived properties of the test statistic , assess its power and compare it with other standard approaches . An application to the Framing heart study ( FHS ) illustrates the value of the approach in practice . A new genetic locus for the lung-function phenotype , FEV1 ( forced expiratory volume in the first second ) is discovered and replicated in 3 independent , genome-wide association studies .
In order to construct a family-based association test that incorporates both the within and the between-family information , the Z-statistics that correspond to the p-values of FBATi and Ti are computed . The statistic Zα* is the α quantile of standard normal distribution . pFBATi and pTi are the p-value of the FBAT-statistics and one sided p-value of the screening statistic where the direction of the one sided screening statistic is defined by the directionality of FBATi . Based on the statistical independence of FBATi and Ti [11] under the null-hypothesis , we can obtain an overall family-based association test statistic Zi by combining both Z-statistics in a weighted sum , where the parameters wFBAT and wT are standardized weights so that the overall family-based association test Zi has a normal distribution with mean 0 and variance 1 , i . e . wFBAT2+wT2 = 1 . In the literature , this approach of combining two test statistics is known as the Liptak-method [24] . However , the Liptak-method varies here from its standard application in that the 2 test statistics have to be combined so that confounding in the screening statistic Ti cannot affect the validity of the overall family-based association test statistic Zi . In the context of a genome-wide association study ( GWAS ) , we are able to achieve this goal by using rank-based p-values for the screening statistic Ti instead of their asymptotic p-values . The “screening statistics” Ti are sorted based on their evidence for association so that T ( m ) denotes the screening statistic with the least amount of evidence for association and T ( 1 ) the screening statistic with the largest amount of evidence for association . The rank-based p-value , ( i – 0 . 5 ) /m , is then assigned to the ordered screening test statistics T ( i ) . If there is a tie , then the average of the ranks will be used for the computation of the rank-based p-value for the ith marker . Since the null-hypothesis will be true for the vast majority of the SNPs in a GWAS , the rank-based p-values provide an alternative way to assess the significance of the population-based screening statistic Ti . The overall association test Zi is then computed based on the Z-score for the asymptotic p-value of the FBAT-statistic and the Z-score for the ranked-based p-value of the screening statistic Ti . In Text S1 we show that the overall association test Zi maintains the global significance level α , under any situations including population admixture and stratification . This can be understood intuitively as well . The smallest rank-based p-value is 0 . 5/m . Using the Bonferroni-correction to adjust for multiple testing , the individual , adjusted significance level is α/m which will always be smaller than the smallest rank-based p-value , 0 . 5/m , unless the pre-specified global significance level α is great than 0 . 5 . This implies that the overall family-based association test can never achieve genome-wide significance just based on the rank-based p-values alone . The FBAT-statistic has to contribute evidence for the association as well in order for the overall family-based association test to reach genome-wide significance . Finally , we have to address the specification of the weights wFBAT and wT in the overall family-based association test statistic Zi . While any combination of weights wFBAT and wT will provide a valid test statistic Zi , the most powerful overall statistic Zi is approximately achieved when the ratio of the weights is equal to the ratio of the standardized effect sizes , the expected effect size of the regression coefficient divided by its ( estimated ) standard deviation . For quantitative traits in unascertained samples , one can show that optimal power levels are achieved for equal weights , i . e . wFBAT = wT . In general , the equal weighting scheme seems to provide good power levels for any disease mode of inheritance and for different trait types , e . g . binary traits , time-to-onset , etc . The theoretical derivation of optimal weighting schemes for such scenarios is ongoing research and will be published subsequently . Furthermore , it is important to note that , instead of the Liptak-method , Fisher's method for combining p-values could have been used as well to construct an overall family-based association test which would have the same robustness properties as the overall-test based on the Liptak-method . However , simulation studies ( data not shown ) suggest that the highest power levels are consistently achieved with the Liptak method . We therefore omit the approach based on Fisher's method here .
In the first part of the simulation study , the type-1 error of the proposed family-based association test denoted as LIP was assessed in the absence and in the presence of population admixture , and we use the Wald test based on the conditional mean model [22] with between-family component for pTi in our all simulations . For various scenarios , we verified that the proposed overall family-based association test maintains the α-level . For simplicity , we assume in the simulation studies that the random samples are given , i . e . no ascertainment , and that the parental genotypes are known . Assuming Hardy-Weinberg equilibrium , the parental genotypes are generated by drawing from Bernoulli distributions defined by the allele frequencies . The offspring genotypes are obtained by simulated Mendelian transmissions from the parents to the offspring . For the jth trio , the offspring phenotype Yj is simulated from a Normal distribution with mean aXj and variance 1 , i . e . N ( aXj , 1 ) , where the parameter a represents the genetic effect size and the variable Xj denotes the offspring genotype . Under the null-hypothesis of no association , the genetic effect size parameter a will be set to 0 . For scenarios in which population admixture is present , we assume that the admixture is created by the presence of 2 subpopulations whose phenotypic means differ by 0 . 2 . The allele frequencies for each marker in the two subpopulations are generated by the Balding-Nichols model [25] . That is , for each marker , the allele frequency in an ancestral population is generated from a uniform distribution between 0 . 1 and 0 . 9 , U ( 0 . 1 , 0 . 9 ) . Then , the marker allele frequencies for the two subpopulations are independently sampled from the beta distributions ( p ( 1−FST ) /FST , ( 1−p ) ( 1−FST ) /FST ) for the whole markers in each replicate of the simulated GWAS . A survey reported FST estimates with a median of 0 . 008 and 90th percentile of 0 . 028 among Europeans , and the corresponding values are 0 . 027 and 0 . 14 among Africans , and 0 . 043 and 0 . 12 among Asians [26] . The value for Wright's FST was assumed to be 0 . 05 , 0 . 1 , 0 . 2 , or 0 . 3 . Each trio was assigned to the one of the 2 subpopulations with 50% probability . In the absence and presence of the population stratification ( FST = 0 . 05 , 0 . 1 , 0 . 2 , and 0 . 3 ) , Table 1 shows the empirical type-1 error rates of the overall association test statistic Zi for a GWAS with 500 , 000 SNPs . The estimates for the empirical significance levels in Table 1 are based on 2 , 000 replicates . The empirical genome-wide significance level is calculated as the proportion of replicates for which the minimum p-values among the 500 , 000 markers is less than 0 . 05/500 , 000 . We consider the proposed equal weights for wFBAT and wT , for genome-wide significance level 0 . 05 and Table 1 shows that the type-1 error rate is preserved well . For different significance levels , we calculate in Table 2 the empirical proportions of SNPs for which the overall family-based association test Zi is significant at the α-levels of 0 . 05 , 0 . 01 , 10−3 , 10−4 and 10−5 . The simulation studies are conducted in the absence and in the presence of population admixture . Table 2 does not provide any evidence for a departure of the empirical significance levels from the theoretical levels , both in the absence and presence of population substructure . These results confirm our theoretical conclusions that Zi is robust against population stratification and maintains correct type-1 error . In the next set of simulation studies , we assess the effects of the local population stratification on the overall family-based association test . We generate local population stratification under the following assumptions: there are two subpopulations , G1 and G2 which distinguish themselves from each other in 2 marker regions . We assume that a subject can be from all possible 4 combinations at the 2 particular regions , e . g . ( G1 , G1 ) , ( G1 , G2 ) , ( G2 , G1 ) and ( G2 , G2 ) . Both regions consist of 10K SNPs and 90K SNPs respectively and if subjects are from the same subpopulation in each genetic region , their assumed allele frequencies of the markers in the corresponding region are equal . For example , the allele frequencies of each marker in the marker region 1 are the same for samples in ( G1 , G1 ) and ( G1 , G2 ) , but they are different for ( G1 , G1 ) and ( G2 , G2 ) . In the simulation study , we generate the parental genotypes based on these allele frequency assumptions and obtain the offspring genotypes based on simulated Mendelian transmissions . Using the Balding-Nichols model we considered FST's of 0 . 001 , 0 . 005 , 0 . 01 and 0 . 05 in the simulation studies . The offspring's phenotype was generated under the null hypothesis , but we assumed that each sub-population strata had a different phenotypic mean: 0 for ( G1 , G1 ) , 0 . 2 for ( G1 , G2 ) , 0 . 4 for ( G2 , G1 ) and 0 . 6 for ( G2 , G2 ) . Each replicate consists of 2 , 000 trios with equal number of trios for all 4 possible combinations . The data was analyzed with the proposed overall family-based association test and with standard linear regression after adjusting population admixture with EIGENSTRAT [10] . For EIGENSTRAT , we applied the principal component analysis to the mean of the paternal and maternal genotypes at each locus because parents of each offspring are from the same subpopulation , and then the residuals obtained from regressing offspring genotypes and phenotypes with eigenvectors respectively are used to calculate the generalized Armitage trend test [27] . Table 3 provides the empirical type-1 error for both analysis approaches based on 2 , 000 replicates . While EIGENSTRAT exhibits an inflated type-1 error , the proposed overall family test maintains the theoretical significance level . For the analysis of quantitative traits , Table 4 provides the empirical power for 500K GWAS from 2000 replicates when there is no population stratification . Under the assumption of an additive disease model for a quantitative trait , the genetic effect , a , is given as a function of the heritability , h2 , the minor allele frequency pDı and the phenotypic variance , σ2 , by: a = σh/[2p ( 1−p ) ( 1−h2 ) ]0 . 5 . In the simulation study , we assume heritabilities of h2 = 0 . 001 , 0 . 005 , 0 . 01 and 0 . 015 for 2 , 000 , 2 , 500 and 3 , 000 trios . The allele frequency of the disease locus , pDı , is 0 . 3 and the phenotypic variance is 1 . We compare the achieved power levels of the proposed overall family-based association test , Zi , with the weighting approach by Ionita-Laza et al [12] , the original VanSteen approach [11] , the QTDT approach [28] and population-based analysis , i . e . using linear regression of the phenotype Y on the genotype X . Bonferroni correction is used to adjust for multiple testing in the population-based analysis , FBAT , QTDT and the proposed method . The results in Table 4 suggest that the proposed association test achieves power levels that represent a major improvement over the existing methods for family-based association tests ( VanSteen [11] or Ionita-Laza [12] ) . Our approach reaches the same power levels as the population-based analysis . For the power comparisons that are shown in Figure 1 , Figure 2 , and Figure 3 , the number of trios is assumed to be 1 , 000 in 500K GWAS and the empirical powers are calculated based on 10 , 000 replicates at an α-level of 0 . 001 for the all genetic models . The results confirm that the Liptak's method combining Ti and FBATi has similar power to the population-based method , and the choice of equal weights performs well . The simulation results in Table 4 also suggest that QTDT [28] approach achieves similar power levels as the standard FBAT approach , which is consistent with previously reported findings in the literature [29] . However , both standard FBAT and QTDT are still much less powerful than the proposed overall family-based association test . Table 5 shows the empirical power for a GWAS with 100 , 000 SNPs in the presence of population stratification . For the parameters of this simulation study , we assume FST = 0 . 001 , 0 . 005 , 0 . 01 , and 0 . 05 , and the additive mode of inheritance at the disease locus with values for the heritability of h2 = 0 . 005 , 0 . 01 and 0 . 015 . The disease allele frequency pDı in the ancestral population is assumed to be 0 . 3 . The phenotypic data is simulated so that their phenotypic means for two subpopulations are 0 and 0 . 2 respectively . Each individual/trio is assigned to either subpopulation with probability 0 . 5 . The parental genotypes are used to estimate the ancestry for EIGENSTRAT as before . Various methods have been suggested to adjust the population stratification in a population-based designs and we compare the proposed methods with the EIGENSTRAT approach [10] . In order to maximize the power of the proposed method , we apply the EIGENSTRAT approach to the population-based component Ti of our approach , i . e . principal component analysis based on the parental genotypes and the offspring's phenotype is integrated into the generalized Armitage test for Ti [27] . To keep the power comparisons unbiased , the population-based components of the approaches by VanSteen and Ionita-Laza are also adjusted for population admixture , using the EIGENSTRAT approach . The results in Table 5 show that the proposed test statistic Zi is considerably more powerful than population-based analysis adjusted with EIGENSTRAT . QTDT is slightly more powerful than FBAT , but it is much less powerful than LIP as is in Table 4 . This suggests that EIGENSTRAT should be applied only to between-family component in family-based association studies . Our unpublished work showed that the proposed approach can be less powerful than the combination of population-based analysis and EIGENSTRAT if pTi is calculated from the conditional mean model [11] , [22] without adjusting population stratification . For the assessment of the severity of pulmonary diseases , the lung volume of air that a subject can blow out within one second after taking a deep breath is an important endo-phenotype . It is referred to as the forced expiratory volume in one second ( FEV1 ) . FEV1 is an important measure for lung function and we apply the proposed method to a family-based GWAS of FEV1 . The proposed method is applied to 550K GWAS Framingham Heart Study ( FHS ) data set for FEV1 , and then we confirm whether the selected SNPs are replicated in the British 1958 Birth Cohort ( BBC ) , another population sample , as well as two samples of asthmatics in the the Childhood Asthma management program ( CAMP ) [30] and an Afro-Caribbean group of families from Barbados ( ACG ) [31] . In FHS , 9 , 274 subjects were genotyped and 10 , 816 subjects of those had at least one FEV1 measurement . Of the 8637 participants with genotyping and FEV1 measures , only those with a call rate of 97% or higher were included . We adjusted the covariates , age , sex and the quadratic term of height that are known to be associated with FEV1 . For within-family components , the FBAT statistic for quantitative trait was applied . Markers were excluded from the analysis if the number of informative families was less than 20 , or the minor allele frequency was less than 0 . 05 . In total , 306 , 264 SNPs were used for analysis and , based on the number of SNPs , rank-based empirical p-values , pTi , and the genome-wide significance level was obtained with Bonferroni correction . When we let n and ninf be the total number of individuals and the number of informative trios respectively , ninf: ( 2n−ninf ) are used for the weights of Zi because some of parental phenotypes are available . Table 6 shows the p-values for the top 10 SNPs from the proposed method . In our analysis , the genome-wide significance level at 0 . 05 is 1 . 636×10−7 and our results show that only the first ranked SNP , rs805294 , is significant at the genome-wide level 0 . 2 with Bonferroni correction . For rs805294 , we also checked the significance in other data sets , BBC , CAMP [30] and ACG [31] . In CAMP , 1215 subjects in 422 families were genotyped and there are 488 informative trios for rs809254 and in ACG , there were only 33 informative trios ( Table 7 ) . In the BBC , 1372 unrelated subjects were genotyped with the Affymetrix chip and 1323 unrelated subjects genotyped with the Illumina chip . In CAMP and ACG , age , sex and the quadratic terms of heights were adjusted and in the BBC , age , sex , height , recent chest infection and nurse were adjusted . Table 7 also shows that rs805294 is significant and their directions are same for the considered studies except for the ACG sample . In particular , in the ACG study , the MAF of the SNP is different from other studies , which indicates a different local LD structure; The ACG sample is from an Afro-Caribbean population , contrary to the other studies which only include Caucasian study subjects . In addition , the ACG sample lacks statistical power for this particular SNP , i . e . there are only 33 informative trios in this sample . Thus , the inconsistent finding in the ACG study could be attributable to genetic heterogeneity , i . e . different local LD structure/flip-flop phenomena [32] , or insufficient statistical power . For meta analysis , the sample sizes are used as weights for Liptak's method and we use 13∶13∶5∶1 = FHS∶BBC∶CAMP∶ACG as weights because the between-family information is used only for FHS . If the p-value from Illumina gene chip in BBC and the p-values from FHS , CAMP and ACG are combined , then the p-values by Liptak's method using proposed weights and Fisher's method are 1 . 534×10−8 and 1 . 081×10−7 respectively , and they become 4 . 625×10−9 and 3 . 554×10−8 if the p-values from one-tailed tests are used for BBC , CAMP and ACG with the same direction of FHS . If the p-value from the Affymetrix gene chip in BBC is combined with the other studies , then they are 3 . 787×10−8 ( Liptak's method ) and 1 . 890×10−7 ( Fisher's method ) for two-tailed tests , and 1 . 098×10−8 ( Liptak's method ) and 6 . 236×10−8 ( Fisher's method ) for one-tailed tests . As a result we can conclude that rs805294 is significantly associated with FEV1 at a genome-wide scale and the gene , LY6G6C , associated with rs805293 will be investigated in further studies .
Genome-wide association studies have become one of the most important tools for the identification of new disease loci in the human genome . However , even though advances in genotyping technology have enabled a new generation of genetic association studies that provide robust and replicable findings , population stratification/genetic heterogeneity and the multiple testing problems continue to be the major issues in the statistical analysis that have to be resolved in each study . While family-based association tests provide analysis results that are completely robust against confounding due to population-substructures , the analysis approach is not optimal in terms of statistical power . Numerous approaches have been suggested to minimize this disadvantage of family-based association tests but the previous approaches had to compromise either in terms of robustness or in terms of efficiency . In this communication , we develop an approach that efficiently utilizes all available data , while maintaining complete robustness against confounding due to population substructure . The proposed methods combines the p-values of the family-based tests ( the within-component ) with the rank-based p-values for population-based analysis ( the between component ) to achieve optimal power levels . The use of rank-based p-values for the population-based component is similar in spirit to the genomic control approach . In principle , the genomic control functions as rescaling the variance inflated due to population stratification under the assumption of the constant FST . Rank-based p-value directly rescales the statistics based on their ranks , which always generates the uniformly distributed p-value and provides validity even for varying FST due to local population stratification etc . Although our simulations are limited to independent unascertained samples and quantitative traits , the proposed work can be easily extended to ascertained samples , large pedigree , or different trait types , etc . By replacing the parental genotypes with the sufficient statistics by Rabinowitz&Laird [19] , the FBAT-statistic and the screening-statistic can be adopted straight-forwardly to designs with extended pedigrees [23] . Similarly , parental phenotypes can be incorporated into the conditional mean model [23] or its non-parametric extensions [33] as additional outcome variables . The optimal weights can vary between the different scenarios and further theoretical investigation is currently ongoing , but limited initial simulation studies suggest that equal weights , while not always the most powerful choice in such situation , will always result in more powerful analysis than currently used methods .
|
In genome-wide association studies , the multiple testing problem and confounding due to population stratification have been intractable issues . Family-based designs have considered only the transmission of genotypes from founder to nonfounder to prevent sensitivity to the population stratification , which leads to the loss of information . Here we propose a novel analysis approach that combines mutually independent FBAT and screening statistics in a robust way . The proposed method is more powerful than any other , while it preserves the complete robustness of family-based association tests , which only achieves much smaller power level . Furthermore , the proposed method is virtually as powerful as population-based approaches/designs , even in the absence of population stratification . By nature of the proposed method , it is always robust as long as FBAT is valid , and the proposed method achieves the optimal efficiency if our linear model for screening test reasonably explains the observed data in terms of covariance structure and population admixture . We illustrate the practical relevance of the approach by an application in 4 genome-wide association studies .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"genetics",
"and",
"genomics/complex",
"traits"
] |
2009
|
On the Analysis of Genome-Wide Association Studies in Family-Based Designs: A Universal, Robust Analysis Approach and an Application to Four Genome-Wide Association Studies
|
Break-induced replication ( BIR ) is a mechanism to repair double-strand breaks ( DSBs ) that possess only a single end that can find homology in the genome . This situation can result from the collapse of replication forks or telomere erosion . BIR frequently produces various genetic instabilities including mutations , loss of heterozygosity , deletions , duplications , and template switching that can result in copy-number variations ( CNVs ) . An important type of genomic rearrangement specifically linked to BIR is half-crossovers ( HCs ) , which result from fusions between parts of recombining chromosomes . Because HC formation produces a fused molecule as well as a broken chromosome fragment , these events could be highly destabilizing . Here we demonstrate that HC formation results from the interruption of BIR caused by a damaged template , defective replisome or premature onset of mitosis . Additionally , we document that checkpoint failure promotes channeling of BIR into half-crossover-initiated instability cascades ( HCC ) that resemble cycles of non-reciprocal translocations ( NRTs ) previously described in human tumors . We postulate that HCs represent a potent source of genetic destabilization with significant consequences that mimic those observed in human diseases , including cancer .
Double-strand DNA breaks ( DSBs ) are dangerous because they can lead to chromosomal rearrangements or cell death . DSBs may result from a cell's exposure to various DNA-damaging agents , such as radiation and various chemicals , including anti-cancer drugs . In addition , problems with DNA metabolism can also result in DSB formation . DSB-induced changes to the genome have been implicated in promoting various human diseases , including cancer , which emphasizes the importance of proper repair of such lesions . Multiple pathways of DSB repair have evolved ( reviewed in [1] , [2] ) . Non-homologous end joining ( NHEJ ) is a repair mechanism in which two non-homologous broken ends of a DNA molecule fuse together , often producing small DNA insertions and deletions that can be destabilizing . Alternatively , homologous recombination ( HR ) mechanisms repair DSBs through recombination , where broken DNA ends initiate copying of a homologous sequence elsewhere within the genome . The most efficient pathway of HR is gene conversion ( GC ) , where both ends of a DSB use a homologous sequence to copy lost DNA in order to repair the DSB lesion . Alternatively , break-induced replication ( BIR ) is an HR mechanism that employs only a single end of a DSB for repair . During BIR , a single broken DNA end invades a homologous region within the genome to initiate extensive DNA synthesis that can copy large portions of a donor molecule through its telomere ( reviewed in [3]–[6] ) . BIR is a primary pathway to repair broken replication forks and eroded telomeres . Also , it has been observed that gap repair can proceed through BIR [7] and , for reasons that are not entirely clear yet; the frequency of BIR is known to increase in aged cells [8] . BIR is initiated by strand invasion , which occurs with kinetics similar to those of the GC pathway [7] . However , after strand invasion , progress stalls and DNA synthesis is delayed by 4 or more hours [9] . The exact reason for this pause is not known , but several possibilities have been proposed , including slow replication fork assembly , unstable D-loop formation , and the existence of a “recombination checkpoint” , discouraging BIR repair ( reviewed in [3] , [5] , [6] ) . The delay in BIR initiation leads to the establishment of a checkpoint-mediated G2/M arrest , which allows cells to complete BIR prior to cell division . Consistently , it was observed that a defective checkpoint ( achieved by deletion of RAD9 ) decreased BIR efficiency and also increased chromosome loss [9] . Once DNA synthesis associated with BIR is initiated , it is fast and processive , similar to normal S-phase DNA replication [9] . It has been demonstrated that the initiation of BIR DNA synthesis involves the majority of proteins required for initiation of S-phase DNA replication [10] . Also , Polδ , a main replicative polymerase , plays a crucial role in DNA synthesis in BIR [11]–[13] . However , the role of two other replicative polymerases , Polε and Polα , in BIR remains unclear [11] , ( also reviewed in [5] , [6] ) . In addition , the mechanism of BIR is drastically different from S-phase replication as it proceeds via migrating DNA bubble leading to conservative inheritance of newly synthesized strands [14] , [15] , [16] . While the end result of BIR is repair of the DSBs , the mechanism of BIR increases the likelihood of a variety of deleterious outcomes that may have destabilizing consequences in the genome . Among these are loss of heterozygosity , deletions , duplications , translocations , copy-number variations , and a significantly elevated mutation rate [9] , [17]–[25] . In addition , half- crossovers ( HCs ) , which are chromosome fusions initially identified in rad51Δ and rad52Δ mutants [13] , [26]–[29] , were recently demonstrated to occur in wild type and various mutants following initiation of BIR [12] , [13] . BIR-induced HC formation is initiated by strand invasion , but the resulting intermediate ruptures prior to repair to yield a rearranged chromosome consisting of fused pieces of the recipient and donor molecules , as well as a destabilized fragment of the broken donor . Accordingly , HC formation requires proteins involved in the strand invasion step of BIR , however the impairment of proteins involved in BIR after strand invasion promotes HCs [12] . Thus , HCs are markedly elevated in yeast bearing mutations pol32Δ and pol3-ct , which interfere with successful initiation of DNA synthesis [12] , [13] . It was proposed that the failure to initiate DNA synthesis in these mutants promotes resolution of the Holliday junction ( HJ ) formed during strand invasion . The exact mechanism of HJ resolution remains unknown , though the resolvase Mus81 has been implicated as one protein capable of resolving HJs and therefore may contribute to HC formation [13] . HCs result in the breakage of a previously intact donor chromosome , and this can have deleterious consequences by initiating recurrent cycles of genetic instability . Analogous cycles ( called NRTs , for non-reciprocal translocations ) have been described in mammalian tumors where broken chromosomes initiate recombination with an intact donor , which in turn leads to breakage of the donor [30] . While the molecular mechanism of NRTs remains undefined , we have previously proposed [12] that cycles of NRTs are mediated by cascades of HCs that continue until the donor fragments are either stabilized or lost . Thus , further investigation into HC formation and the possible cascades of genetic instability that may result is warranted . To further define mechanisms of HC formation and the effects of HCs on genetic instability , we hypothesized that various factors that interrupt ongoing BIR replication may induce HC formation in a manner similar to mutations that prevent initiation of BIR replication . We show that interruption of BIR synthesis by exposure of cells to DNA damaging agents or due to a defective replisome results in a dramatic increase in HCs . Moreover we demonstrate that a disruption of BIR imposed by premature onset of mitosis increase HC formation . Finally , we document the occurrence of half-crossover-initiated instability cascades ( HCCs ) that closely resemble NRT cycles observed in cancer cells .
Half-crossovers ( HCs ) are chromosome fusions resulting from aberrant processing of BIR intermediates . It has been proposed that HCs lead to deleterious consequences by initiating cascades of genetic instabilities . Here we aimed to identify genetic factors that promote the channeling of BIR into HC formation . To assay the efficiency of BIR and the frequency of half-crossovers in DSB repair , we employed our disomic experimental system in yeast , Saccharomyces cerevisiae , wherein a galactose-induced DSB is initiated at the MATa locus on the truncated copy of chromosome III ( recipient chromosome ) ( Fig . 1A ) [12] . The second full copy of chromosome III contains the uncleavable MATα-inc allele and serves as a template for DSB repair ( donor chromosome ) . Upon induction of the DSB , DNA is repaired predominantly by BIR ( Fig . 1B ) because only one end of the DSB has large homology to the full-length donor copy of chromosome III . The ends of both chromosomes that participate in BIR repair are marked by ADE1 , LEU2 , ADE3 or HPH; such that repair outcomes can be determined using appropriate selective media . Also , a NAT cassette was used to replace a region 30 kb centromere-proximal to MATa that contained two Ty1 elements ( FS2 ) in the recipient chromosome [21] . Using these genetic markers , it was determined that more than 75% of DSB repair outcomes displayed an Ade+Leu− phenotype , indicating BIR repair of the galactose-induced DSB ( Fig . 1B , Fig . S1 ) . Approximately 14% of the DSB repair outcomes were Ade+Leu+ , indicating the DSB was repaired by gene conversion ( GC ) ( Fig . 1C , Fig . S1 ) . Other colonies had an Ade−rLeu− phenotype ( were ADE1-deficient and red ( as described in [12] ) and resulted from failure of the chromosome to repair the DSB leading to chromosome loss ( CL ) ( Fig . 1D ) . Also , a small number of colonies were Ade−wLeu− ( were ADE3-deficient and white ) , which represented HC events resulting from fusion of the ADE1-containing segment of the recipient chromosome with the HPH-containing segment of the donor chromosome and concurrent loss of the ADE3 and LEU2 segments of the donor and recipient chromosomes , respectively ( Fig . 1E ) . We tested the effect of mutations that impair DNA polymerases on HC formation by plating yeast on a galactose-containing medium [12] . Although each of the Polδ mutations tested here had varying effects on BIR efficiency , they all stimulated HC formation . In particular , pol3-Y708A , a mutation that affects the catalytic subunit of Polδ [31] , dramatically decreased BIR efficiency , and increased chromosome loss ( P<0 . 0001; Fig . S1 ) . Also , pol3-Y708A increased the number of colonies containing HCs to 17% compared to approximately 5% in wild type ( P<0 . 0001; Fig . 2A; see also Fig . S1; note that Fig . 2 presents the fraction of colonies that are fully or partially HCs , while Fig . S1 shows the fraction of HCs among all DSB repair events ) . Similarly , the pol31-WRRGW mutation , which disrupts the Pol31-Pol32 interaction [32] , displayed similar effects and elevated HCs to 26% ( P<0 . 0001; Fig . 2A; Fig . S1 ) . These phenotypes were similar to those previously observed in pol32Δ and pol3-ct mutants [12] , [13] , suggesting that HCs in these mutants are promoted predominantly by failure to initiate BIR . In strains bearing the pol3-t mutation known to compromise the processivity of Polδ during S-phase DNA replication [33] , [34] , HCs were also elevated ( P<0 . 0001 ) , even though these cells frequently successfully completed BIR repair ( Fig . 2A; Fig . S1 ) . Therefore , the increase of HCs in pol3-t might be explained by interruptions of ongoing BIR rather than by problems in BIR initiation . Thus , an intact Polδ appears to be necessary to prevent HC formation . Conversely , mutations impairing Polε showed no effect on HCs . Thus , no increase compared to wild type was observed in either pol2-1 [35] mutants with a truncated catalytic subunit or in pol2-Y831A mutants [31] with a mutation in the same conserved catalytic motif as pol3-Y708A mutants ( Fig . 2A ) . Interestingly , the pol1-1 mutation [36] , which impairs Polα ( a part of the primase complex ) , decreased HCs to less than 0 . 3% ( P<0 . 0001; Fig . 2A ) . Due to the increased HCs observed in mutants with decreased polymerase processivity , we hypothesized that BIR pausing induced by damage in the template DNA used for BIR could promote HC formation . This idea was tested by analyzing DSB repair in our BIR system in the presence of DNA damaging agents . Following induction of BIR in galactose-containing liquid cultures , cells were exposed to either the alkylating agent MMS or 4-NQO ( UV mimetic ) for seven hours while BIR repair occurred . Cells were plated on YEPD and the resulting colonies were analyzed using selective media ( Fig . 2B; Fig . S2 ) . Cell viability was calculated by plating cells on YEPD ( see Materials and Methods for details ) . Both drugs were deactivated ( see Materials and Methods ) prior to being plated on YEPD . In cells treated with 2 . 4 or 6 mM MMS , the percent of colonies with HC outcomes was increased approximately 3- and 4-fold , respectively , compared to cells treated with galactose but no drug ( P<0 . 0001 ) , while the percent of HC outcomes approximately doubled in cells treated with 5 . 3 µM 4-NQO compared to the same control ( P<0 . 0001; Fig . 2B; see also Fig . S2 ) . Significantly , the level of HCs in cells exposed to damage alone ( without DSB induction ) was not significantly increased in comparison to no-damage control ( Fig . 2B , no-DSB ) . Thus , our data suggest that the observed increase in HCs is promoted not by DNA damage per se , but by base damage in the chromosomal region undergoing BIR repair . Our data also identifies a previously unknown synergy between BIR and DNA damage that dramatically increases the rate of HC formation . We note that the levels of DNA damaging drugs used in these experiments , as expected , decrease viability ( see Materials and Methods ) . Overall , since both MMS and 4-NQO induce base damage capable of blocking DNA polymerases ( and therefore replication ) [37]–[39] , our results suggest that interruption of ongoing BIR leads to the aberrant processing of BIR intermediates resulting in HC formation . Previously , we demonstrated that initiation of DNA synthesis during BIR is a very slow process ( takes up-to 4 hours ) and leads to the establishment of a checkpoint-mediated G2/M cell cycle arrest [7] , [9] that prevents mitotic division and thus allowing cells to complete BIR . Given our data that interruptions in BIR due to decreased processivity of polymerases or DNA damage promote HC formation , we hypothesized that checkpoint deficiency may stimulate HC formation due to an interruption in BIR progression by the premature onset of mitosis . To test this hypothesis , we analyzed BIR outcomes in mutants lacking Rad9 or Rad24 , which are required for the DNA damage response ( reviewed in [40] ) . Consistent with previous observations in these mutants [9] , checkpoint deficiency led to a high frequency of multi-sectored colonies ( colonies containing ≥3 different repair sectors ) ( Fig . 3A , Fig . 4C ) . Formation of multi-sectored colonies likely resulted from premature onset of mitosis , and DSB repair that occurred after subsequent cell divisions and took place only in a fraction of the daughter calls . Consistent with this idea were the results of FACS analyses that confirmed full G2/M arrest in wild type cells between 4 and 8 hours after DSB induction , with only partial arrest observed at these time points in rad9Δ and rad24Δ mutants ( Fig . 3B ) . We observed that the percentage of colonies with at least one HC sector was extremely high , 71% and 65% among multi-sectored colonies of rad24Δ and rad9Δ mutants respectively . This was a significant increase ( P<0 . 0001 ) compared to wild type where the frequency of colonies with HC was only 4% ( Fig . 4D ) . Also , when calculated as a fraction of all sectors , HC sectors comprised approximately 20% in rad9Δ and rad24Δ mutants ( Fig . 4E; left ) . Among simple colonies ( with no more than 2 sectors or events evident ) , no notable difference in HC frequency was observed between the checkpoint-deficient and wild type strains ( Fig . 4D ) . We propose that the increase in HCs in checkpoint-deficient mutants results from premature onset of mitosis that may occur either during the first cell division following DSB induction or during subsequent cell divisions , as further explored in the following sections . Genetic analysis of repair outcomes in checkpoint-deficient mutants revealed increased chromosome loss ( P<0 . 0001; Fig . 4A , 4E; middle ) and a decreased level of Ade+Leu−outcomes , which normally represent BIR ( P<0 . 0001; Fig . 4A , 4E; right ) . This was consistent with our previous results [9] and most likely reflected failed DSB repair in these strains . The decreased BIR efficiency in checkpoint-deficient mutants was further supported by PFGE analysis of cells undergoing DSB repair over a 10-hour time course ( performed similar to [12] ) , where the amount of BIR repair product was significantly reduced in rad9Δ and rad24Δ mutants compared to the wild type ( Fig . 3C–F ) ) . Based on the decreased efficiency of BIR , we hypothesized that a fraction of Ade+Leu− events in checkpoint deficient mutants might in fact represent not BIR , but GCRs resulting from abnormal stabilization of the broken molecules ( similar to discussed in [21] , [41] ) . To detect possible GCRs and to characterize their contribution to heterogeneity of the colonies , we employed PFGE to analyze repair outcomes in 23 individual rad24Δ and 11 rad9Δ Ade+/− multi-sectored colonies , which comprised Leu+ and/or Leu− clones ( see , for example , colonies in Fig . 5A , B ) . In this analysis , we focused on individual colonies with at least one HC sector because they represented the majority of all multi-sectored colonies in both mutants . All Ade+Leu− sectors as well as a representative number of Ade+Leu+ ( GC ) , Ade−wLeu− ( HC ) and Ade−rLeu− ( CL ) sectors from each colony were cloned out and analyzed ( see , for example , Fig . 5C for PFGE analysis of all sectors from the colony shown in Fig . 5A ) . PFGE analyses of colony sectors with Ade+Leu+ , Ade−wLeu− and Ade−rLeu− phenotypes confirmed that they resulted from GC , HC , and CL , respectively , as predicted ( Fig . 5C and see also Fig . 4A , B ) . PFGE analysis of the Ade+Leu− outcomes from individual colonies revealed three categories of outcomes ( Fig . 5D ) . In the first group , which comprised 64% and 35% of all Ade+Leu− events in rad24Δ and rad9Δ , respectively , the chromosome structure was similar to the one observed in true BIR outcomes . However , the high fraction of HC events in the respective colonies makes it highly likely that many of these Ade+Leu− outcomes resulted not from BIR , but from segregation of a HC repair product with an intact copy of the full-length chromosome III ( similar to events described in [12] , [13]; see Materials and Methods for details ) . Other Ade+Leu− outcomes were represented by events where DSB repair resulted in formation of GCRs . Thus , approximately 13% of Ade+Leu− from multi-sectored colonies of each of the checkpoint-deficient mutants were chromosomal rearrangements ( called CRs ) , where the broken recipient chromosome was aberrantly stabilized by de novo telomere formation or through ectopic recombination between a Ty or delta element in the MATa-containing chromosome and Ty or delta element located in an ectopic position ( Fig . 1F , Fig . 5D; similar to previously demonstrated [21] ) . These CRs carried an unchanged donor chromosome ( a 356 kb band that hybridized to the ADE3-specific probe ) and a recipient band of any size ( different from 346 kb ) that hybridized to ADE1 ( Fig . 4A , 4B; CR ) . Previously , the structure of similar CR events was characterized by array-CGH [21] and it was determined that CRs often result from ectopic BIR initiated by strand invasion of Ty or delta elements of the broken recipient chromosome into Ty or delta elements at ectopic positions . A significant fraction of Ade+Leu− outcomes represented a new type of GCR that contained a single BIR-sized ( 346 kb ) recipient chromosome and a rearranged donor ( a band other than 356 kb that hybridized to ADE3 ( Fig . 4A , 4B; half-crossover-initiated cascades ( HCC ) , Fig . 5D ) ) . We posited that these repair outcomes likely arose from the rupture of the donor chromosome during HC formation , resulting in an ADE3-containing broken fragment that was stabilized by ectopic recombination . The possibility of such HCC events has been previously discussed [12] , but never demonstrated . Here we found that 61% and 73% of rad24Δ and rad9Δ multi-sectored colonies contained at least one Ade+Leu− sector that represented a HCC event ( Fig . 5D ) , the molecular structure of which was further analyzed by array-CGH ( see below ) . We conclude that premature onset of mitosis resulting from a defective checkpoint leads to aberrant processing of BIR intermediates resulting in frequent HCs and other GCRs . We observed that more than 74% and 91% of all analyzed colonies in rad24Δand rad9Δ , respectively , contained at least one CR or HCC sector . Also , both HCCs and CR events were frequently observed among Ade+Leu− events obtained from unselected ( random ) colonies in both rad9Δ and rad24Δ mutants , but were very rare in the wild type strains ( Fig . 6A , 6B , 6F , and [9] , [12] ) . In addition , the analysis of strains containing the pol3-t mutation , which increased the frequency of HC formation ( discussed in the previous section ) also revealed DSB-induced HCC outcomes ( Fig . S3 ) . As indicated above , the majority of multi-sectored colonies in rad9Δ and rad24Δ mutants contained at least one HCC event characterized by the presence of a 346 kb band that hybridized to ADE1 , as well as a second band of varying size ( other than 356 kb ) that hybridized to an ADE3-specific probe and represented a GCR that resulted from breakage and stabilization of the donor chromosome ( Fig . 4B; HCC ) . We used comparative genomic hybridization ( array-CGH ) to further characterize the nature of 13 stabilized donor chromosomes obtained from HCC events identified in rad24Δ mutants . ( Fig . 7A ) Based on array-CGH , the stabilized donor chromosomes resulting from HCCs were divided into three main classes that accounted for all 13 analyzed outcomes: isochromosomes ( Class I ) , translocations ( Class II ) , and secondary BIR events ( Class III ) ( Table S2 ) . Class I rearrangements included 9 of the 13 HCC events analyzed by array-CGH ( Fig . 7A , Table S2 ) . Each of these events had in common a deletion of sequences in the right arm of chromosome III and a duplication of sequences from the opposite chromosome arm ( Fig . 7B , 7C; Table S2 ) , thus forming an isochromosome . ( Class I is subdivided into Class Ia , Ib , Ic , and Id depending on the point of recombination and other details of the process; see Table S2 ) . We propose that the formation of outcomes Ia , Ib , and Ic was initiated by invasion of the broken recipient into the full donor chromosome III , which led to the formation of an HC represented by a 346 kb band hybridized to ADE1-specific probe ( Fig . S4 , S5 ) . The resulting broken ADE3-containing fragment was then resected and subsequently repaired by non-allelic homologous recombination between a Ty or delta element located in the right arm of chromosome III and a Ty or delta element located in the left arm of chromosome III . For example , in the case of Class Ia outcomes ( H7 and H8 ) , the recombination occurred between the Ty1γ element in FS1 and a delta element YCLCδ1 ( Fig . 7A , Table S2 , Fig . S4 ) . The predicted size of such an isochromosome ( calculated based on [19] , [42] , and also based on the data from SGD ) was approximately 245 kb ( Table S2 ) , which was consistent with the size of the ADE3-hybridizing band observed by PFGE analysis ( Fig . 7A ) . Our proposed molecular structure was further confirmed through a detailed Southern analysis using the restriction enzymes EcoO1091 and EciI and Probe 1 ( FS1-specific ) and Probe 4 ( specific to the region of chromosome III located centromere-distal to YCLCδ1 ( Fig . 7B , Fig . S4 , Table S3 ) . Analogously , in the case of Class Ib outcomes ( H10 , H11 , H12 and H13 ) , recombination occurred between delta elements YCRCδ6 and YCLWTy1-1 [19] , which corresponds to YCLWδ15in SGD . The predicted size of such an isochromosome was approximately 219 kb ( Table S2 ) , which was consistent with the size of the ADE3-hybridizing band observed by PFGE analysis ( Fig . 7A ) . The structures of H10 , H11 , and H13 were further confirmed by Southern analysis using the restriction enzymes EciI and PshAI and also Probe 4 and Probe 5 ( specific to region of chromosome III located centromere proximal to YCRCδ6 ) ( Fig . 7C , Fig . S5 , Table S3 ) . Also see Text S1 and Fig . S6 for a detailed description of Class Ic HCC outcomes ( H4 and H5 ) . The formation of the HCC outcome H2 ( Class Id ) can be explained similarly to other Class I events , but likely involved two half-crossover events ( see Text S1 and Fig . S7 ) . Class II included only one of the 13 HCC outcomes , outcome H3 . H3 was determined to result from a deletion in chromosome III between positions of FS1 ( a tandem repeat of Ty1 elements ) and MAT , and a duplication of all sequences located on chromosome V distal to a solo delta element YERCdelta16 ( Fig . S8 and Table S2 ) . We propose that the formation of H3 was initiated by HC formation between the left arm of the recipient and the right arm of the donor chromosomes ( similar to Class I ) , which resulted in a broken ADE3-containing fragment . This fragment was subsequently repaired by recombination between a Ty element in FS1 and a delta element located in the right arm of chromosome V , which led to the formation of translocation . ( See Text S1 , Table S2 and Fig . S8 for the details of structural analysis of H3 ) . Class III rearrangements included 3 of the 13 analyzed HCC outcomes and were further divided into IIIa and IIIb . Class IIIa was represented by the outcome H9 . Array-CGH analysis of this outcome indicated a duplication of chromosome III sequences from MAT through the telomere indicative of BIR repair; however , PFGE analysis revealed that both ADE1 and ADE3-hybridizing chromosomes were equal in size ( 346 kb long; Fig . S9 , Table S2 ) . We hypothesized that , similar to other HCC events , the formation of H9 was initiated by HC , which led to the formation of an ADE3-containing broken fragment . This broken donor fragment was stabilized through invasion into the HC product centromere proximal to NAT followed by BIR that copied the right arm of HC ( Fig . S8 ) . Therefore , we named this outcome a “secondary BIR event” . Importantly , 8 of 24 HCCs that were originally identified by PFGE showed a pattern similar to H9 ( both ADE1- and ADE3-hybridizing bands were approximately 346 kb; data not shown ) , which suggests that all of them most likely represented secondary BIR events , even though only H9 was analyzed by array-CGH . Therefore , it appears that secondary BIR events are relatively common among BIR outcomes in checkpoint-deficient mutants . The array-CGH analysis of the outcomes H6 and H1 ( class IIIb ) also showed a duplication of chromosome III sequences from MAT through the telomere indicative of BIR , which made them similar to secondary BIR events . However , these events included additional rearrangements ( Table S2; see also Text S1 for details ) . Overall , we conclude that interruptions during BIR repair in checkpoint-deficient mutants lead to frequent breakage of the donor chromosome that results in further cascades of DNA instabilities . It has been demonstrated that the absence of Rad9 increases the rate of resection at a DSB , which could contribute to the increased frequency of chromosome loss and GCRs we observed in rad9Δ mutants [43] , [44] . Therefore , we tested whether sgs1Δ and rad50Δ known to decrease the efficiency of DSB resection ) [45]–[47] affected the distribution of repair outcomes in rad9Δ and rad24Δ . We observed that deletion of SGS1 , which is required for long-range 5′-strand resection , in rad24Δ , dramatically reduced the frequency of chromosome loss ( P<0 . 0001 ) ( Fig . 4E ) , and virtually eliminated all multi-sectored colonies ( Fig . 4C ) . The majority of colonies formed in rad24Δsgs1Δ were fully Ade+Leu− , and their PFGE analysis indicated that they contained normal BIR events ( Fig . 4E; Fig . 6D , 6F ) , even though they can also represent secondary BIR events . Deletion of SGS1 in rad9Δ also affected the distribution of repair outcomes . Compared to rad9Δalone , the frequency of chromosome loss was decreased ( P<0 . 0001 ) , while HCC were increased in the double mutant ( P<0 . 0001 ) ( Fig . 4E; Fig . 6E , 6F ) . In addition , we observed that deletion of RAD50 , which is involved in end processing near the DSB site [45] , did not affect the distribution of repair outcomes in rad9Δ or rad24Δ ( Fig . 4E ) . Importantly , the deletion of SGS1 gene alone ( in strains with functional checkpoint response ) led to only a modest change in distribution of DSB repair outcomes ( Fig . 4C , 4E , 6C , 6F . ) . Overall , our data suggest that deletion of SGS1 significantly affects the distribution of repair outcomes in the absence of a functional checkpoint response .
BIR is a critical mechanism to repair broken chromosomes . Normally , BIR is initiated by a DSB produced in such a way that only one end of the broken molecule is available for repair ( Fig . 8A ) . It thus initiates with a single invasion into a homologous template ( Fig . 8B ) followed by initiation of DNA synthesis ( Fig . 8C ) that proceeds to the telomere ( Fig . 8D ) . Increased HC formation and chromosome loss was previously demonstrated in pol32Δ and pol3-ct mutants during BIR repair where strand invasion was successful , but DNA synthesis could not be ( or was poorly ) initiated [12] , [13] . Likewise , here we report a similar phenotype in strains containing other mutations in Polδ , pol3-Y708A ( a mutation affecting the catalytic subunit of Polδ ) [31] and pol31-WRRGW ( a mutation in the Pol31 subunit of Polδ ) [32] . For each of these cases , we propose that HC formation results from resolution of HJ structures that persist when BIR DNA synthesis is not initiated ( Fig . 8B , 8L ) . It has been suggested that Mus81 resolves HJs and therefore may contribute to HC formation [13] . We demonstrate that BIR interrupted at various stages of its progression , for example during replication , also stimulates HCs . This most likely occurs in pol3-t mutants because the processivity of Polδ is compromised ( Fig . 8C , 8L ) . Similarly , results from another recent study [15] demonstrate that the deletion of PIF1 , which encodes a DNA helicase specifically required for DNA synthesis during BIR [48] , also leads to more frequent HCs . We speculate that in these mutants , DNA synthesis is successfully initiated but proceeds with frequent stops , thereby promoting HC formation . We postulate that a similar mechanism of paused DNA synthesis can explain the increase in HC formation we observed in wild type cells exposed to the DNA damaging agents ( MMS or 4-NQO ) during BIR repair ( Fig . 8C , 8L ) . Regardless of the mechanism that leads to paused replication , these data support our hypothesis that interruptions in DNA replication during BIR induce HCs . Interestingly , mutations affecting Polε that were investigated so far did not promote HCs . This might be explained by limited participation of Polε in BIR [11] . Curiously , the mutation in Polα ( pol1-1 ) led to decreased HCs compared to wild type , which may indicate that pol1-1 delays accumulation of BIR intermediates that are resolved to produce HCs . We previously demonstrated that successful completion of BIR replication requires checkpoint machinery to maintain cell cycle arrest until repair is completed [9] . Consistently , here we observed that the premature onset of mitosis in checkpoint-deficient cells undergoing BIR repair led to an increased frequency of HCs ( Fig . 8F , 8L ) . Formation of HC molecules could result from a signal from the cell to resolve the HJ structure as previously discussed , but mechanical rupture of BIR intermediates initiated by chromosomal segregation is also a possibility . In checkpoint-deficient mutants , we also frequently observed chromosome loss and translocations , which we propose result from strand dissociation that can be stimulated by HJ resolution ( Fig . 8G , 8H ) . 5′-to-3′ DNA resection following dissociation may lead to chromosome loss ( Fig . 8K ) or to ectopic strand invasion at positions of DNA repeats ( Fig . 8I ) resulting in translocations ( Fig . 8J ) . Alternatively , elevated chromosome loss and translocations may result from increased 5′-to-3′ resection of DSB ends prior to strand invasion ( Fig . 8A , 8K ) or following unwinding of a D-loop ( Fig . 8B , 8H ) An important outcome of this study is the discovery of HC-induced cascades ( HCC ) . The existence of HCCs has been previously hypothesized [12] , but , until now , had not been demonstrated . HCCs represent DSB repair outcomes that contain a HC product along with a rearranged donor chromosome ( Fig . 8N ) . We propose that HCCs are initiated by a single HC that leads to breakage of the donor chromosome ( Fig . 9A , B ) . The new DSB in the donor molecule undergoes 5′ to 3′ resection ( Fig . 9C ) , and the resulting 3′ DNA end invades a homologous DNA molecule at an ectopic position ( Fig . 9E ) in the newly formed HC ( Fig . 9F ) , or in the sister chromatid ( Fig . 9G ) . This initiates recombination and can stabilize the broken donor chromosome if repair proceeds through BIR; conversely , this intermediate may also result in HC formation , thereby continuing the cascade of genetic instability . Even when the donor fragment successfully stabilizes through BIR , if an ectopic site such as a Ty or delta element is used for recombination , translocations will occur ( Fig . 9E; see also Figs . S4 , S5 , S6 , S7 , S8 ) . In our system , repair of the broken donor chromosome often proceeded by BIR using the recently formed HC ( Fig . 9F ) or the sister chromatid ( Fig . 9G ) as a template . We termed these events “secondary BIR” . Among all HCC events analyzed by PFGE , approximately half showed a pattern suggestive of HCC resulting in invasion of the broken donor into the initial HC , even though only one of these cases was analyzed by array-CGH ( case H9; Fig . S9 ) . Additionally , 15% of the HCCs analyzed by CGH could be explained by secondary BIR associated with complex rearrangements ( cases H1 and H6 ) . It should be noted that all identified secondary BIR events were initiated by strand invasion that occurred centromere proximal to FS2 which resulted in a change in the size of the donor and therefore allowed the detection of these events . It remains possible; however , that many additional secondary BIR events are initiated by strand invasion between FS2 and MAT . This is expected to result in chromosome structures and phenotypes indistinguishable from classic BIR . Therefore , we propose that the actual frequency of secondary BIR events maybe higher than currently estimated . This is significant because secondary BIR events could be more deleterious than classic BIR events . For example , we speculate that secondary BIR may result in homozygous mutations that result when a mutation occurs during DNA synthesis associated with HC formation ( Fig . 9F ) and is then copied during the repair of the broken donor using the initial HC as a template . Homozygous mutations could be more deleterious than heterozygous ones ( reported in association with classic BIR [18] ) because they can lead to the manifestation of recessive phenotypes including those leading to cancer . A significant finding was the formation of multi-sectored colonies consisting of broad genotypic variations by checkpoint-deficient mutants . More than 70% of the multi-sectored colonies contained at least one sector with chromosomal rearrangements of recipient or donor chromosomes , with HCC being a major class of these rearrangements . In addition , we observed that deletion of SGS1 , which is known to reduce long-range DSB resection , led to a significant increase in HCC frequency in rad9Δ mutants . Possibly , reduced resection stabilizes the broken donor chromosome , thus giving it more chances to repair by invading a homologous template ( Fig . 9E–G ) . Alternatively , it is possible that faster initiation and/or progression of BIR that was previously documented in the absence of Sgs1 [7] contributes to the increased stabilization of the donor chromosome and therefore to the increased level of HCC . Interestingly , in rad24Δsgs1Δ , the multiple sectoring of colonies was completely eliminated , and the majority of outcomes were indistinguishable from normal BIR . We propose that these events are likely to be secondary HCC resulting from secondary BIR . In addition to being frequently observed when BIR is induced in checkpoint-defective mutants , HCCs occurred in cases of compromised BIR in polymerase-deficient mutants ( pol3-t ) and when BIR proceeds in the presence of MMS ( Sakofsky et al , manuscript submitted ) . HCCs were also observed in BIR-defective pif1Δ mutants [15] . Overall , we propose that ongoing cycles of genetic instability are a ubiquitous outcome of HC formation . We propose that HCs and HCCs may be a mechanism for genetic destabilization leading to various diseases in humans . In particular , we propose HCCs to be a mechanism capable of producing non-reciprocal translocations ( NRTs ) that have been described in mammalian tumor cells . NRT is a pathway of telomere acquisition by broken chromosomes that results in the donor molecule losing genetic information , including its telomere , and becoming unstable [30] . This destabilization of the donor makes NRTs especially devastating because the events are self-perpetuating and result in cascades of genomic destabilization , including chromosome loss and multiple rearrangements . We propose that the cycles of NRTs can be explained by initiation of BIR followed by its interruption leading to HCCs in tumor cells . Importantly , we suggest that initiation of HCC can be facilitated by checkpoint deficiency , which is frequent in cancer cells . [49] , [50] . In addition , our data suggest that cycles of HCCs could also contribute to clonal variations in pre-cancerous cells . Finally , our observation of increased HCs and HCCs resulting from the exposure of cells undergoing BIR to DNA damaging agents , could be of significant relevance to current cancer therapeutic strategies where anti-replication drugs are used in combination with agents that induce DSBs ( for example , X-ray and gamma irradiation ) . We propose that DSBs induced by these agents may initiate BIR , which will frequently lead to NRT events in the presence of cancer drugs that inhibit replication . Because NRT events often initiate cascades of genetic instability , promoting such events in cancer cells could be one mechanism of rapid GCR formation that could result in negative oncotherapy outcomes such as secondary tumorigenesis and drug resistance .
All yeast strains ( Table S1 ) were isogenic to AM1003 [12] which is a chromosome III disome with the following genotype: hmlΔ::ADE1/hmlΔ::ADE3 MATa-LEU2-tel/MATα-inc hmrΔ::HPH FS2Δ::NAT/FS2 leu2/leu2-3 , 112 thr4 ura3-52 ade3::GAL::HO ade1 met13 . In this strain , the HO endonuclease-induced DSBs introduced at MATa are predominantly repaired by BIR because the portion of the chromosome centromere-distal to MATa is truncated to leave only 46 bp of homology with the donor sequence [9] , [12] . The majority of single-gene deletion mutants were constructed by transformation with a PCR-derived KAN-MX module flanked by terminal sequences homologous to the sequences flanking the open reading frame of each gene [51] . The resulting constructs were confirmed by PCR and by phenotype . sgs1Δ mutants were constructed by PCR-amplification of sgs1ΔURA3 from the strain yWH239 [48] using oligonucleotides complementary to sequences flanking SGS1 . To disrupt RAD50 , plasmid pNKY83 [52] was digested simultaneously with BglII and EcoRI and transformed into recipient strains that were subsequently screened for a Ura+ phenotype ( rad50::hisG::URA3::hisG ) . Several polymerase-deficient mutants were constructed using the “pop-in-pop-out” method . The “pop-in” step involved transformation of strain AM1003 with the following plasmids: 1 ) HpaI-linearized p170-Y708 containing the pol3-Y708 allele [31] , 2 ) AgeI-linearized p173-Y831 containing the pol2-Y831 allele [31] , 3 ) SacI-linearized pCM54 containing the pol1-1 allele [36] , 4 ) MfeI-linearized pRS306-pol31-WRRGW containing the pol31-WRRGW allele [32] , 5 ) HpaI-linearized p171 containing the pol3-t allele [34]; and 6 ) EcoRI-linearized p2A5 containing the pol2-1 allele [35] . In every case , the “pop-in” step was followed by a “pop-out” step that involved growth of transformants on rich medium for two days followed by selection of Ura− outcomes on 5-FOA . Rich medium ( yeast extract-peptone-dextrose [YEPD] ) and synthetic complete medium , with bases and amino acids omitted as specified , were made as described [53] . YEP-lactate ( YEP-Lac ) and YEP-galactose ( YEP-Gal ) media contained 1% yeast extract and 2% Bacto peptone supplemented with 3 . 7% lactic acid ( pH 5 . 5 ) or 2% ( w/v ) galactose , respectively . Yeast cultures were grown at 30°C or at 20°C ( in the case of yeast strains bearing polymerase mutations , which rendered them temperature-sensitive ) . As indicated , MMS or 4-NQO ( Sigma Aldrich ) was added to rich medium for some experiments . The kinetics of DSB repair was examined in time-course experiments as described previously [12] . YEP-Lac ( 500 to 1000 ml was inoculated with approximately 2×106 cells/ml . Cultures were grown at 30°C overnight to reach a concentration of approximately 5×106 cells/ml . HO endonuclease was induced by the addition of galactose to achieve a final concentration of 2% . For PFGE gel electrophoresis , 50 ml aliquots were removed , and sodium azide was added to achieve a concentration of 0 . 1% to stop DNA repair processes . Extraction of DNA embedded in 0 . 55% agarose plugs was performed as described [9] . For fluorescence-activated cell sorter ( FACS ) analyses , 5 ml aliquots were removed , cells were spun , diluted , fixed by the addition of 70% ethanol , and stored at 4°C . FACS analysis was performed using propidium iodide with a Becton Dickinson fluorescence-activated cell analyzer , similar to [9] . PFGE was performed by running genomic DNA embedded in agarose plugs at 6 V/cm , for 40 hours ( initial switch time 10 s; final switch time 35 s ) followed by Southern blotting and hybridization using 32P- labeled DNA probes containing either an ADE1 ( SalI fragment from pJH879 ) ( similar to [9] , [12] ) or ADE3 ( obtained by PCR amplification of chromosome VII from 907979–908735 bp ) sequence . The images were analyzed using GE Healthcare Typhoon FLA 9500 . The kinetics of accumulation of BIR product was measured using an ADE1-specific fragment as a probe . To account for variation in DNA loads , intensities of the bands corresponding to the intact chromosome III , as well as to the repaired chromosome III , were normalized to intensities of the bands corresponding to chromosome I , which also hybridizes to the ADE1-specific probe . The efficiency of BIR repair , presented as the percentage of truncated chromosome III that was converted to BIR product , was calculated by dividing the normalized intensity of a repair band by the normalized intensity of uncut , truncated chromosome III . Results of three time-course experiments were used to calculate the average ± SD BIR efficiency for each strain . Cell viability following exposure to the DNA damaging agents was determined as a ratio of the number of colony forming units ( CFU ) observed experimentally and the number that was predicted based on the cell concentration determined using hemocytometer prior to plating . The viability of cells following treatment with DNA damaging agents was as follows: for DSB repair in the presence of MMS: ( 20±7 ) % and ( 14±7 ) % for 2 . 4 mM and 6 mM MMS respectively; for DSB repair in the presence of 5 . 3 µM 4-NQO: ( 10±5 ) %; following MMS treatment without DSB: ( 23±8 ) % and ( 8±3 ) % for 2 . 4 mM and 6 mM MMS respectively; and following 5 . 3 µM 4-NQO treatment without DSB: ( 19±4 ) % . To monitor the repair of HO-induced DSBs in individual colonies , we harvested logarithmically growing cells grown in YEP-Lac at 30°C and plated them on YEP-Gal . The resulting colonies were then replica plated onto omission media to examine the ADE1 , ADE3 , LEU2 , and NAT markers . When temperature-sensitive strains bearing pol3-t or pol1-1 polymerase mutations were used , the cells were grown in YEP-Lac at 20°C . Following plating on YEP-Gal , the cells were incubated at 30°C for 24 hours ( a length of time sufficient to complete BIR ) , and then incubated at 20°C until the colonies were full-grown . To test the effect of DNA damage , the cells were grown to log phase in YEP-Lac medium at 30°C , incubated in galactose-containing media for 30 minutes ( to induce DSB ) , and then incubated with or without a DNA damaging agent ( MMS or 4-NQO ) for 7 hours . ( The time for incubation with DNA damaging agents was selected based on the known kinetics of BIR [12] ) . The DNA damaging agents were then deactivated by treatment with 10% sodium thiosulfate prior to serial dilution and plating of cells onto YEPD . Repair events were identified by a phenotypic analysis after replica plating onto omission media , and also by PFGE . Gene conversion ( GC ) outcomes displayed an Ade+Leu+ phenotype and contained two copies of chromosome III: a 356 kb chromosome that hybridized to an ADE3-specific probe and a short ( 217 kb ) chromosome that hybridized to an ADE1-specific probe ( Fig . 1C ) . The absence of repair led to CL , which was detected by formation of Ade−rLeu− colonies containing a single , 356 kb chromosome III , which hybridized to the ADE3-specific probe ( Fig . 1D ) . Formation of Ade−wLeu− colonies or colony sectors indicated formation of HCs . These cells contained a single , 346 kb chromosome III that hybridized to the ADE1-specific probe ( Fig . 1E ) . Ade+Leu− phenotypes could result from several repair outcomes: BIR , HC ( when it co-segregates with an intact copy of the donor chromosome during mitosis [12] ) , from CRs [21] or from HCC events . CRs and HCC events were identified by PFGE . CRs carried a 356 kb band that hybridized to the ADE3-specific probe and a band of any size ( different from 346 kb ) that hybridized to ADE1 ( Fig . 1F ) . HCC contained a single , 346 kb band that hybridized to an ADE1-specific probe in addition to a band of varying size that hybridized to ADE3 ( Fig . 1G ) . BIR usually carried a 356 kb band that hybridized to ADE3 and a 346 kb band that hybridized to ADE1 ( Fig . 1B ) . Occasionally , BIR resulted from long 5′-3′ resection beyond NAT , followed by strand invasion that occurred centromere proximal to NAT . This led to the formation of BIR outcomes that were Ade+Nat−Leu− ( BIR-LR ) with donor and recipient chromosomes of identical ( 356 kb ) sizes ( Fig . 5C ) . PFGE could not distinguish between BIR and events where a HC co-segregated with an intact donor chromosome . Therefore , we assumed the number of Ade+Leu− HCs to be equal to the number of Ade−wLeu− HCs based on the idea that an HC product should co-segregate with an intact copy of the donor chromosome in half of the cases of HC formation . Overall , the formula to calculate the number of BIR events was as follows: BIR = ( number of Ade+Leu− ) − ( GCR+HCC+HC ) . The distribution of various types of repair among all repair events was determined differently for simple colonies ( containing <3 repair sectors ) and for multi-sectored colonies ( containing ≥3 repair sectors ) . The frequency among simple colonies was determined as previously described in [9] . The frequency of each repair outcome in multi-sectored colonies was determined as the sum of all sectors belonging to this phenotypic class divided by the total number of sectors analyzed . Importantly , since yeast strains used in this study did not clump during plating , we proposed that multi-sectored colonies originated from single cells , where DSB repair was preceded by mitotic divisions . This idea was also supported by our experiments where individual checkpoint-deficient cells were micro-manipulated on galactose containing plates , which led to the formation of multi-sectored colonies ( data not shown ) . In total , the following number of colonies were scored in experiments aimed to determine the effect of defective polymerases on half-crossover formation for each subsequent strain: Pol+ ( wt ) – 1192 colonies; pol3Y-708A – 2428 colonies; pol3-t – 1240 colonies; pol31-WRRGW – 776 colonies; pol2-Y831A – 2491 colonies; pol2-1 – 896 colonies; and pol1-1 – 287 colonies . In addition , the following number of colonies were scored in experiments aimed to determine the effect of DNA damage on half-crossover formation for each subsequent condition: DSB+no damage: – 2583 colonies; DSB+5 . 3 µM 4-NQO – 2645 colonies; DSB+2 . 4 mM MMS – 526 colonies; DSB+6 mM MMS – 1186 colonies; no DSB+5 . 3 µM 4-NQO – 1629 colonies; no DSB+2 . 4 mM MMS – 1407 colonies; no DSB+6 mM MMS – 778 colonies; no DSB+no damage - 1072 colonies . The number of colonies scored in experiments aimed to determine the distribution of repair outcomes in checkpoint-deficient mutants was as follows: Rad+ ( wt ) – 718 colonies , rad24Δ – 756 colonies; rad9Δ – 465 colonies; rad24Δsgs1Δ - 339 colonies; rad9Δsgs1Δ - 338colonies; and rad9Δrad50Δ - 340 colonies . Finally , the number of simple ( s ) and multiple ( m ) repair events scored during analysis of the effect of checkpoint- deficient mutants was as follows for each strain background: wild type ( wt ) : 1353 s; rad24Δ: 671 s and 1782 m; rad9Δ: 473 s and 946 m; rad24Δsgs1Δ: 677 s; rad9Δsgs1Δ: 450 s and 515 m; rad9Δrad50Δ: 508 s and 346 m; sgs1Δ: 1504 s and 49 m . The CGH analysis was conducted as described recently [54] . Briefly , genomic DNA was prepared from the agarose-embedded full length chromosome material . DNA from the parental strain was labeled with dUTP-Cy3 and DNA from the derivative strains carrying genome rearrangements was labeled with dUTP-Cy5 . The labeled DNAs were mixed and competitively hybridized to custom Agilent 60-mer oligonucleotide microarrays . The arrays were scanned , the images were analyzed , and the CNV regions were identified using GenePix 6 . 0 and Nexus Copy Number software , respectively . For Southern analysis , genomic DNA of DSB repair outcomes was purified by glass bead/phenol method as described [55] , digested with the appropriate restriction enzymes , and the resulting fragments were separated on a 0 . 8% agarose gel . Southern blotting was carried out by standard procedures using 32P-labeled DNA probes that were generated by PCR amplification using 24 bp primers ( available as Table S3 ) and genomic DNA of AM1003 as a template . The locations of these probes on chromosome III are as follows: ( 1 ) Probe 1 , 148247–148549 bp ( SRD1 ( FS1 ) -specific ) ; ( 2 ) Probe 3 , 167594–167893 bp ( RHB1-specific ) ; ( 3 ) Probe 4 , 82015–82365 bp ( KCC4-specific ) ; and 4 ) Probe 5 , 123682–123981 bp ( CIT2-specific ) . The location of the YER134C-specific probe on chromosome V is 436745–437044 bp ( Probe 2 ) . For all probes mentioned above , the starting and ending coordinates on the corresponding chromosomes are derived from the Saccharomyces Genome Database ( SGD ) . All mutants were analyzed for their effect on BIR repair in at least three independent plating experiments . Results from these independent experiments were pooled if it was determined that the distributions of all events were statistically similar to each other using a Chi-square test . The effects of individual mutations on DSB repair were determined by comparing the resulting pooled distributions of repair outcomes obtained for mutants to the distribution obtained for the wild-type strain ( AM1003 ) by Chi-square tests . Specifically , to determine the effect of various mutations on the frequency of HC , all repair outcomes were divided into two groups: HC ( AdewLeu− outcomes ) and others ( combining all other groups ) . Comparison of the distributions between these two classes in specific mutants vs . wild type was used to determine whether a mutation affected the frequency of HC . The effect of mutations on other DSB repair outcomes and the effects of exposure to various DNA damaging agents were determined similarly .
|
Maintaining genomic stability is important to prevent birth defects , genetic disorders and other diseases , including cancer . Double-strand DNA breaks ( DSBs ) , which can result from exposure of living cells to ionizing radiation and various chemicals , threaten genomic integrity , thus making DSB repair essential . The choice of DSB repair pathway is important because some pathways confer destabilizing consequences . Break-induced replication ( BIR ) is a mechanism of DSB repair that is often associated with deleterious events that can threaten genetic stability . One such deleterious event is the formation of half-crossovers ( HCs ) , which occurs when two chromosomes physically interacting during BIR repair fuse . Here we employed a yeast-based system to unravel the genetic factors promoting HC formation . We demonstrate that the interruption of BIR due to problems in DNA synthesis or checkpoint control , promote HCs . Additionally , we document that disruption of BIR promotes half-crossover-initiated cascades ( HCC ) that can significantly destabilize the genome and could be accounted as a potential mechanism responsible for cycles of non-reciprocal translocations contributing to cancer in humans .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biochemistry",
"nucleic",
"acids",
"dna",
"dna",
"repair",
"biology",
"dna",
"recombination",
"molecular",
"genetics",
"computational",
"biology"
] |
2014
|
Cascades of Genetic Instability Resulting from Compromised Break-Induced Replication
|
Nipah virus ( NiV ) is a highly pathogenic re-emerging virus that causes outbreaks in South East Asia . Currently , no approved and licensed vaccine or antivirals exist . Here , we investigated the efficacy of ChAdOx1 NiVB , a simian adenovirus-based vaccine encoding NiV glycoprotein ( G ) Bangladesh , in Syrian hamsters . Prime-only as well as prime-boost vaccination resulted in uniform protection against a lethal challenge with NiV Bangladesh: all animals survived challenge and we were unable to find infectious virus either in oral swabs , lung or brain tissue . Furthermore , no pathological lung damage was observed . A single-dose of ChAdOx1 NiVB also prevented disease and lethality from heterologous challenge with NiV Malaysia . While we were unable to detect infectious virus in swabs or tissue of animals challenged with the heterologous strain , a very limited amount of viral RNA could be found in lung tissue by in situ hybridization . A single dose of ChAdOx1 NiVB also provided partial protection against Hendra virus and passive transfer of antibodies elicited by ChAdOx1 NiVB vaccination partially protected Syrian hamsters against NiV Bangladesh . From these data , we conclude that ChAdOx1 NiVB is a suitable candidate for further NiV vaccine pre-clinical development .
Nipah virus ( NiV ) is a highly pathogenic emerging virus in the family Paramyxoviridae , genus Henipavirus . The virus was first detected in 1998 when it caused an outbreak of severe , rapidly progressing encephalitis in pig farmers in Malaysia and Singapore , with a case-fatality rate of 38% [1] . Pigs were most likely infected after eating fruit contaminated by infected fruit bats of the genus Pteropus , the animal reservoir of NiV [2] . This was then followed by pig-to-human infection , but only very limited human-to-human transmission was reported [3] . Since then , NiV has caused near annual outbreaks in Bangladesh and India . Reported outbreaks in these countries involve fewer patients , with a higher case-fatality rate ( average 75% ) [4] and , in contrast to the Malaysia outbreak , there are well-documented cases of human-to-human transmission [5] . It has been hypothesized that the source for NiV infection in index cases in India and Bangladesh is palm sap contaminated with bat urine [6–9] . The most recent outbreak of NiV occurred in the Indian state Kerala in May 2018 where 19 patients were infected resulting in 17 fatalities . Although this state had not seen NiV infections before 2018 , bats of the genus Pteropus are prevalent in this area . Sequence analyses have demonstrated that isolates originating from Malaysia and Bangladesh represent two different genetic lineages [10–12] . A second member of the Henipavirus genus is Hendra virus ( HeV ) , which is characterized by a similar pathology and has caused infections in humans in Australia [13] . NiV-caused disease is characterized by the onset of non-specific symptoms such as fever , headache , dizziness , and myalgia . Hereafter patients may develop severe encephalitis and pulmonary disease . Pulmonary disease is observed more frequently in patients infected with NiV Bangladesh . A unique potential complication is the late onset or relapsing encephalitis , which has been documented up to 11 years after NiV infection [14] . The host range of NiV is broad , facilitated by the use of the conserved ephrin-B2 and–B3 as cellular receptors [15] , raising the possibility of further outbreaks resulting from transmission from infected livestock or domestic animals . The current lack of licensed vaccines or treatments has prompted the WHO to identify NiV as a pathogen requiring urgent investment into development of countermeasures [16] . Given the sporadic nature of NiV outbreaks , the aim is to develop a vaccine that demonstrates protective efficacy in animal models and acceptable safety and immunogenicity profiles in phase I and II clinical trials . Vaccines which meet these criteria will be stockpiled and may then be used in the event of an outbreak , following clinical trial protocols for use prepared in advance . Apart from demonstrating efficacy against challenge in animal models , other desirable characteristics for a vaccine to be stockpiled are the availability of large-scale manufacturing processes , thermostability , and safety in all sections of the population including the youngest , oldest , and immunocompromised patients [17] . ChAdOx1-vectored vaccines fulfil all these requirements , making this a promising platform . The ChAdOx1 vector is a replication-deficient simian adenovirus vector which has been used to produce several vaccines which are now in clinical development . A common feature of these vaccines is their low reactogenicity , strong immunogenicity , and the absence of vector replication after immunization , an important safety feature . In pre-clinical studies a single dose of ChAdOx1 vectored vaccines has been shown to be protective against infection with Rift Valley Fever Virus , Middle East respiratory syndrome coronavirus , Mycobacterium tuberculosis and Zika virus [18–21] . Large scale manufacturing has been performed for replication-deficient adenoviral vectored vaccines for Ebola , with one vaccine now licensed and another in advanced clinical development [22 , 23] . Further , a simple thermostabilization process allows for vaccine storage at ambient temperatures [24] , removing the need for a cold chain for storage and shipping . We now report on pre-clinical immunogenicity and efficacy testing of ChAdOx1 NiVB .
Animal experiment approval was received from the Institutional Animal Care and Use Committee ( IACUC ) at Rocky Mountain Laboratories . Experiments were performed in an Association for Assessment and Accreditation of Laboratory Animal Care-approved facility by certified staff , following the guidelines and basic principles in the NIH Guide for the Care and Use of Laboratory Animals , the Animal Welfare Act , United States Department of Agriculture and the United States Public Health Service Policy on Humane Care and Use of Laboratory Animals ( Protocol # 2017-033E and 2018-035E ) . The Institutional Biosafety Committee ( IBC ) approved work with infectious NiV and Hendra virus ( HeV ) strains under BSL4 conditions and sample inactivation was performed according to IBC-approved standard operating procedures for removal of specimens from high containment . Henipavirus isolates were obtained from the Special Pathogens Branch of the Centers for Disease Control and Prevention , Atlanta , GA or Public Health Agency , Winnipeg , Canada . NiV Bangladesh ( GenBank no . AY988601 ) , NiV Malaysia ( GenBank no . AF212302 ) , and HeV ( GenBank no . AF017149 ) have been passaged three , four , and three times in VeroE6 cells respectively . All virus propagation in this manuscript was performed in VeroE6 cells in Dulbecco’s modified Eagle’s medium ( DMEM , Sigma ) supplemented with 2% fetal bovine serum ( Gibco ) , 1 mM L-glutamine ( Gibco ) , 50 U/ml penicillin ( Gibco ) , and 50 μg/ml streptomycin ( Gibco ) ( 2% DMEM ) . VeroE6 cells were maintained in DMEM supplemented with 10% fetal bovine serum , 1 mM L glutamine , 50 U/ml penicillin and 50 μg/ml streptomycin . The glycoprotein ( G ) gene from Nipah virus ( Bangladesh outbreak 2008–2010 , Genbank accession number: JN808864 . 1 ) was codon optimized for humans and synthesized by GeneArt ( Thermo Fisher Scientific ) . The synthesized G gene was cloned into a transgene expression plasmid comprising a modified human cytomegalovirus immediate early promoter ( CMV promoter ) with tetracycline operator ( TetO ) sites and the polyadenylation signal from bovine growth hormone ( BGH ) . The resulting expression cassette was inserted into the E1 locus of a genomic clone of ChAdOx1 using site-specific recombination [25] . The virus was rescued and propagated in T-REx-293 cells ( Invitrogen ) . Purification was by CsCl gradient ultracentrifugation , and the virus was titered as previously described [26] . Doses for vaccination were based on infectious units ( IU ) . Female Golden Syrian hamsters ( 4–6 weeks old ) were purchased from Envigo . Animals were vaccinated I . M . with 50 μl of 108 IU of vaccine or injected I . M . with 50 μl of saline , in each thigh ( 100 μl total volume ) . For the homologous challenge vaccine experiment , animals were vaccinated at D-70 and/or D-42 . For the heterologous challenge experiment , animals were vaccinated at D-28 . Three days prior to vaccination and virus challenge animals were bled via orbital sinus puncture . All animals were challenged with 1000LD50 of virus in 500 μl DMEM via I . P . inoculation: NiV Bangladesh = 5 . 3 x 105 TCID50; NiV Malaysia = 6 . 8 x 104 TCID50; HeV = 6 . 0 x 103 TCID50 . We chose the I . P . route as a uniformly lethal challenge route and to be able to compare with previously conducted vaccine experiments [27] . For each study group , 10 hamsters were utilized . Of these , four animals were euthanized 4 ( HeV ) or 5 ( NiV ) days post inoculation and the remaining six animals were followed for 28 days post challenge . Weight was recorded daily up to 10 days post infection , and oropharyngeal swabs were taken daily up to 7 days post inoculation in 1 mL of DMEM . Animals were euthanized when >20% of weight loss was recorded , or severe disease signs ( e . g . difficulty breathing or paralysis ) were observed . Upon euthanasia , blood and tissues were collected and subsequently analyzed for virology and histology as approved by IACUC . Female Golden Syrian hamsters ( 4–6 weeks old ) were purchased from Envigo . Fifteen animals were vaccinated with either ChAdOx1 NiVB or ChAdOx1 GFP as described above at 56 and 28 days before serum collection . Serum was collected via cardiac puncture , pooled per vaccine group and IgGs were purified using the MAbtrap kit ( Sigma ) according to manufacturer’s instructions from 10 mL of serum . Purified IgGs were filtered through an 0 . 45μm filter and diluted to 4 . 5 mL in sterile PBS . Ten hamsters were immunized via I . P . injection using 400 μl per hamster . All animals were challenged as described above one day post treatment . For each study group , 10 hamsters were utilized . Of these , four animals were euthanized 5 days post challenge and the remaining six animals were followed for 56 days post challenge . Weight was recorded daily up to 10 days post challenge , and oropharyngeal swabs were taken daily up to 7 days post inoculation in 1 mL of DMEM . Animals were euthanized when >20% of weight loss was recorded , or severe disease signs ( e . g . difficulty breathing or paralysis ) were observed . Upon euthanasia , blood and tissues were collected and subsequently analyzed for virology and histology as approved by IACUC . Virus titrations were performed by end-point titration in VeroE6 cells , which were inoculated with tenfold serial dilutions of virus swab media or tissue homogenates . After 1hr incubation at 37°C and 5% CO2 , tissue homogenate dilutions were removed , washed twice with PBS and replaced with 100 μl 2% DMEM . Cytopathic effect was scored at 5 dpi and the TCID50 was calculated from 4 replicates by the Spearman-Karber method [28] . Sera was heat-inactivated ( 30 min , 56°C ) and two-fold serial dilutions were prepared in 2% DMEM . Hereafter , 100 TCID50 of NiV was added . After 1hr incubation at 37°C , virus was added to VeroE6 cells and incubated at 37°C and 5% CO2 . At 5 dpi , cytopathic effect was scored . The virus neutralization titer was expressed as the reciprocal value of the highest dilution of the serum which still inhibited virus replication . NiV-G Malaysia ( residues E144—T602 , gene accession number NC_002728 ) was cloned into the pHLSEC mammalian expression vector [29] and NiV-F Malaysia ( residues G26—D482 , gene accession number AY816748 . 1 ) was cloned into the pHLSEC vector containing a C-terminal GCNt trimerization motif [30] . The constructs were transiently expressed in human embryonic kidney ( HEK ) 293T cells in roller bottles , as described previously [29] . Supernatant was harvested 96 hours after transfection and diafiltrated using the AKTA FLUX system ( GE Healthcare ) against either PBS , pH 7 . 4 ( NiV-G ) or buffer containing 10 mM Tris and 150 mM NaCl , pH 8 . 0 ( NiV-F ) . The proteins were further purified by Ni-NTA immobilized metal-affinity chromatography using His-Trap HP columns ( GE Healthcare ) followed by size exclusion chromatography . NiV-G was purified using a Superdex 200 10/300 Increase GL column ( GE healthcare ) equilibrated in PBS pH 7 . 4 and NiV-F was purified using a Superose 6 Increase 10/300 GL column ( GE healthcare ) equilibrated in 10 mM Tris and 150 mM NaCl pH 8 . 0 . Maxisorp plates ( Nunc ) were coated overnight at 4°C with 5 μg of G or F protein per plate in Carb/Bicarb binding buffer ( 4 . 41 g KHCO3 and 0 . 75 g Na2CO3 in 1 L distilled water ) . After blocking with 5% milk in PBS with 0 . 01% tween ( PBST ) , serum ( 2x serial diluted starting at 100x dilution ) in 5% milk in PBST was incubated at RT for 1 hr . Antibodies were detected using affinity-purified antibody peroxidase-labeled goat-anti-hamster IgG ( Fisher , 14-22-06 ) in 5% milk in PBST and TMB 2-component peroxidase substrate ( Seracare ) and read at 450 nm . All wells were washed 3x with PBST in between steps . Prior to using F and G proteins based on NiV Malaysia , we established that cross-reactivity with NiV Bangladesh antibodies was sufficient for usage in ELISA by testing sera known to be positive for NiV Bangladesh antibodies . Necropsies and tissue sampling were performed according to IBC-approved protocols . Harvested tissues were fixed for a minimum of 7 days in 10% neutral-buffered formalin and subsequently embedded in paraffin . Hematoxylin and eosin ( H&E ) staining and in situ hybridization ( ISH ) were performed on tissue sections and cell blocks . Detection of NiV and HeV viral RNA was performed using the RNAscope FFPE assay ( Advanced Cell Diagnostics Inc . , Newark , USA ) as previously described [31] and in accordance with the manufacturer’s instructions . Briefly , tissue sections were deparaffinized and pretreated with heat and protease before hybridization with target-specific probes for NiV or HeV . Ubiquitin C and the bacterial gene , dapB , were used as positive and negative controls , respectively . Whole-tissue sections for selected cases were stained for NiV and HeV viral RNA , UBC and dapB by the RNAscope VS FFPE assay ( RNAscopeVS , Newark , USA ) using the Ventana Discovery XT slide autostaining system ( Ventana Medical Systems Inc . , Tucson , USA ) . A board-certified veterinary anatomic pathologist evaluated all tissue slides . Statistical analysis was performed by the Log-rank ( Mantel-Cox ) test to compare survival curves , and by Welch-corrected one-tailed unpaired student’s t-test to compare infectious virus titers in tissue . SEM was calculated for all samples . P-values < 0 . 05 were significant .
To determine efficacy of the ChAdOx1 NiVB vaccine , we vaccinated groups of 10 hamsters with either a single dose at D-42 or a prime-boost regime at D-70 and D-42 . As control groups , we either injected hamsters with ChAdOx1 GFP at D-70 and D-42 or saline at D-42 ( Fig 1A ) . Virus neutralizing antibodies could be detected after a single dose of ChAdOx1 NiVB and increased upon a secondary dose ( average VN titer ± SEM = 30 . 5 ± 5 . 7 after single dose , 91 ± 21 after boost ) . In contrast , no neutralizing antibodies could be detected in serum obtained from the control groups ( Fig 1B ) . All hamsters were challenged with a lethal dose of NiV Bangladesh ( 1000 LD50 ) via intraperitoneal inoculation on D0 ( Fig 1A ) . All vaccinated animals survived challenge and did not show signs of disease , such as weight loss , at any stage throughout the experiment . This was in contrast to the control groups in which all animals succumbed to disease between D6 and D10 and exhibited weight loss ( Fig 1C and 1D ) , as well as respiratory and/or neurological signs , including labored breathing and paralyzed hind legs . Statistical analysis demonstrated that survival in the vaccinated groups was significant compared to both control groups ( P < 0 . 0001 ) . Oropharyngeal swabs were taken daily and assessed for infectious virus by limiting dilution titrations . None of the vaccinated animals shed virus at any timepoint . In contrast , control animals from both groups were found to shed virus at D5 and D6 ( Fig 1E ) . Four animals of each group were euthanized at D5 and lung and brain tissue were harvested . Infectious virus could only be detected in lung tissue of animals from both control groups ( average titer ± SEM = 3 . 3 x 104 ± 2 . 5 x 104 TCID50/g of tissue ) and was not detected in any tissue of the vaccinated animals ( Fig 1F ) . We did not observe any differences between the two control groups . Lung and brain tissue harvested at D5 were then evaluated for pathological changes . None of the vaccinated animals displayed pulmonary pathology and no viral RNA was detected in lung tissue by ISH . Control animals developed pulmonary lesions that were indistinguishable between the two groups . These hamsters developed bronchointerstitial pneumonia that was characterized by multifocal inflammatory nodules that were centered on terminal bronchioles and extend into adjacent alveoli . The nodules were composed of large numbers of foamy macrophages and fewer neutrophils and lymphocytes admixed with small amounts of necrotic debris . In most cases hemorrhage , fibrin and edema admixed with inflammatory cells was observed . Edema and fibrin often were extended into surrounding alveoli . Alveoli that were adjacent to areas of inflammation were thickened by fibrin , edema and small numbers of macrophages and neutrophils as previously observed in NiV infected hamsters [32] . There was abundant viral RNA demonstrated by ISH in areas of inflammation ( brown staining ) . The viral RNA was predominantly found in type I pneumocytes but was also multifocally present in vascular and bronchiolar smooth muscle and endothelial cells ( Fig 2 ) . To determine efficacy of ChAdOx1 NiVB against NiV Malaysia and HeV , groups of 10 hamsters were vaccinated with a single dose of ChAdOx1 NiVB or a single dose of ChAdOx1 GFP at D-28 ( Fig 3A ) . As before , virus neutralizing antibodies could be detected after vaccination with ChAdOx1 NiVB but not upon injection with ChAdOx1 GFP ( Average VN titer ± SEM = 68 . 6 ± 13 . 6 ) ( Fig 3B ) . Subsequently , hamsters were challenged with either NiV Malaysia or HeV ( 1000 LD50 ) via intraperitoneal inoculation on D0 ( Fig 3A ) . All vaccinated animals challenged with NiV Malaysia survived with no signs of disease such as weight loss at any stage throughout the experiment . In contrast , animals challenged with NiV Malaysia that received ChAdOx1 FGP all succumbed to infection between D5 and D6 . These animals experienced weight loss and respiratory and neurological signs ( Fig 3C and 3D ) . Statistical analysis demonstrated that survival in the vaccinated group was significantly different from the control group ( P = 0 . 0012 ) . Oropharyngeal swabs were taken daily and assessed for infectious virus . None of the vaccinated animals challenged with NiV Malaysia shed virus at any timepoint . In contrast , control animals challenged with NiV Malaysia were found to shed virus at D5 and D6 ( Fig 3E ) . Four animals from both groups were euthanized at D5 and lung and brain tissue were harvested . Infectious virus could only be detected in lung and brain tissue of animals from the control group ( average virus titer lung ± SEM = 1 . 5 x 105 ± 5 . 2 x 104 TCID50/g , brain ± SEM = 6 . 8 x 101 ± 4 . 4 x 101 TCID50/g ) and was not detected in any tissue of the vaccinated animals ( Fig 3F ) . Four out of six vaccinated animals challenged with HeV succumbed to disease between D5 and D7 . The two survivors showed minimal weight loss ( <2% ) and no signs of disease . Animals that received ChAdOx1 FGP all succumbed to HeV infection between D4 and D6 . These animals showed weight loss as well as respiratory and neurological signs ( Fig 3C and 3D ) . Log-rank ( Mantel-Cox ) test demonstrated that survival in the vaccinated group was significant ( P = 0 . 0476 ) compared to the control group . Oropharyngeal swabs were taken daily and assessed for infectious virus . None of the vaccinated animals challenged with HeV shed virus at any timepoint . In contrast , control animals challenged with HeV were found to shed virus at D4 , D5 and D6 ( Fig 3E ) . Four animals from both groups were euthanized at D4 and lung and brain tissue were harvested . Infectious virus was detected in three out of four lungs of the vaccinated animals and all lungs of the control animals ( average virus titer ± SEM = 5 . 2 x 105 ± 3 . 6 x 105 and 4 . 4 x 106 ± 2 . 2 x 106 TCID50/g tissue for vaccinated and control animals , respectively ) . No statistical difference in infectious virus titer was found between the two groups using an unpaired one-tailed Student’s t-test ( P = 0 . 0674 ) . Infectious virus was only detected in brain tissue of animals from the control group ( average titer ± SEM = 4 . 6 x 102 ± 2 . 0 x 102 TCID50/g ) and not in vaccinated animals ( Fig 3F ) . Harvested lung tissue was then evaluated for pathological changes . All four groups of hamsters developed pulmonary lesions . All animals challenged with HeV and control animals challenged with NiV Malaysia developed bronchointerstitial pneumonia which was indistinguishable from the lesions described for the control animals in the homologous challenge study . Vaccinated hamsters challenged with NiV Malaysia developed mild to moderate bronchointerstitial pneumonia and did not display any evidence of pulmonary edema , fibrin or hemorrhage . ISH demonstrated viral RNA predominantly in type I pneumocytes and rarely in vascular and bronchiolar smooth muscle and endothelial cells in animals challenged with HeV and control animals challenged with NiV Malaysia . In vaccinated animals challenged with NiV Malaysia , however; there was very little RNA present and only in type I pneumocytes in areas of inflammation ( Fig 4 ) . Finally , we wanted to assess the protective effect of antibodies elicited after ChAdOx1 NiVB vaccination . Two groups of 15 hamsters were either vaccinated with ChAdOx1 NiVB or injected with ChAdOx1 FGP at D-56 and D-28 . All animals were bled at D0 and we collected 13 and 15 mL respectively . IgG was purified from 10 mL pooled serum . Ten animals per group were then injected peritoneally with purified IgG . Animals were challenged with a lethal dose of NiV Bangladesh ( 1000 LD50 ) one day post passive transfer ( Fig 5A ) . We were unable to detect neutralizing antibodies in serum obtained at D5 from four hamsters from each group . However , serum from animals treated with NiV antibodies was positive by ELISA against NiV G protein , albeit with a lower reciprocal titer than antibodies in serum obtained from single-dose vaccinated animals ( Fig 5B ) . One out of six animals treated with NiV antibodies succumbed to disease on D11 . No weight loss was observed , however the animal showed severe neurological signs . None of the other NiV antibody-treated animals experienced weight loss or signs of disease . Four out of six animals treated with GFP antibodies succumbed to disease between D6 and D8 . These animals showed weight loss and respiratory or neurological signs . The two surviving animals did not show any signs of disease throughout the experiment . One of these animals did not seroconvert as measured by ELISA against NiV F and G protein , and it was suspected this animal was not infected . Therefore , this animal was excluded from the survival curve . The log-rank ( Mantel-Cox ) test demonstrated that survival in the treated group was significant ( P = 0 . 0168 ) compared to the control group ( Fig 5C and 5D ) . Oropharyngeal swabs were taken daily and assessed for infectious virus . Shedding was minimal and found in one animal treated with NiV antibodies on D5 , and five animals treated with GFP antibodies between D4 and D6 ( Fig 5E ) . Four animals from both groups were euthanized at D5 and lung and brain tissue were harvested . Infectious virus could only be detected in lung tissue of animals treated with GFP antibodies and was not detected in any tissue of the animals treated with NiV antibodies ( Fig 5F ) . Lung tissue harvested at D5 was then evaluated for pathological changes . Both groups of hamsters developed pulmonary lesions similar to those described in the homologous challenge study , however; the NiV antibody-treated hamsters developed mild to moderate pulmonary lesions whereas the control animals developed severe lesions . Additionally , none of the NiV antibody-treated hamsters displayed any pulmonary fibrin , edema or hemorrhage . ISH demonstrated viral RNA in type I pneumocytes in areas of inflammation . Abundance of viral RNA was notably less in animals treated with NiV antibodies ( Fig 6 ) .
NiV is a re-emerging infectious disease which causes outbreaks with a high case-fatality rate . No licensed vaccine against NiV currently exists , and it is therefore key that a safe and effective vaccine be developed . Several vaccine candidates have been explored in different animal models . These can be categorized as subunit vaccines or live-vectored vaccines that target the NiV outer membrane proteins G and/or F . Protection against disease and lethality has been shown in hamsters [27 , 33] , pigs [34 , 35] , African green monkeys [36–38] , cats [39] , and ferrets [40 , 41] . Efficacy is thought to be mediated by neutralizing antibodies , as passive transfer of antibodies in naive animals also results in protection against disease [27 , 42] . These approaches are promising , but no vaccine candidates have so far been moved into clinical trials . In the studies presented here , we tested the efficacy of a vaccine based on NiV Bangladesh G protein in a replication-deficient simian adenovirus vector in Syrian hamsters . A prime-only as well as a prime-boost regime protected Syrian hamsters against challenge with a lethal dose of NiV Bangladesh and NiV Malaysia , and partially protected against HeV challenge . Furthermore , antibodies elicited by vaccination alone provided partial protection against a NiV Bangladesh challenge . Two genetic lineages of NiV have been described; NiV Malaysia and NiV Bangladesh [10–12] . Although NiV Malaysia has not caused an outbreak in humans since 1999 , the virus was isolated from Pteropus vampyrus , Pteropus hypomelanus and Pteropus lylei in Malaysia and Cambodia [43–45] and another spillover event could occur . Having one vaccine that protects against both lineages of NiV virus would be the easiest and cheapest countermeasure . A single-dose vaccination with ChAdOx1 NiVB , which is based on NiV Bangladesh , fully protected Syrian hamsters against lethal disease caused by NiV Malaysia . The G proteins of the NiV strains used in this study are 95 . 5% pairwise identical on the amino acid level , with 27 amino acid differences scattered throughout the protein . Although we did not see sterile protection against NiV Malaysia , none of the vaccinated animals showed signs of disease and all were protected against lethal disease . These results suggest that ChAdOx1 NiVB could protect against both lineages of NiV . Like NiV , HeV is a species in the Henipavirus genus and thus we investigated cross-protection of ChAdOx1 NiVB against a lethal challenge with HeV in Syrian hamsters . The G protein of the HeV strain used in this study was 78 . 2% identical to the ChAdOx1 NiVB G protein; 133 amino acids differ between the two proteins . ChAdOx1 NiVB only protected partially against HeV challenge; four out of six animals did not survive challenge . We observed a non-significant decrease in infectious HeV titer in lung and brain tissue of vaccinated animals compared to control animals . It is possible that disease progression in vaccinated animals is delayed compared to control animals . This is supported by the delay in time to death; whereas the average time to death is 5 days in control animals , it is 6 days in vaccinated animals . Cross-protection of NiV or HeV vaccines has been studied by other groups as well . An adeno-associated virus vaccine expressing NiV G protein offered 50% protection against a lethal challenge with HeV in hamsters [46] . In contrast , vaccines based on HeV provide full protection against NiV in the ferret and NHP model [36 , 41 , 47] . Likewise , high levels of cross-protective antibodies were found in sera from HeV-infected individuals , whereas cross-protective antibodies were limited in NiV-infected individuals [48] . This might be caused by induction of a more robust and cross-reactive immune response by native HeV protein compared to NiV protein , as suggested by Bossart et al . [48] . Human cases of HeV are associated with direct contact with infected horses , the intermediate animal host of HeV , and direct contact with bats or their products has not yet been associated with HeV infection in humans [49] . It is therefore likely that prevention of HeV in horses will completely prevent human cases . Currently , a HeV vaccine ( Equivac ) is available for horses and fully protects against HeV [50] . Furthermore , the total number of human cases that contracted HeV is relatively low at 7 [13] . Thus , the requirement of a human vaccine for HeV is therefore less urgent than that of a NiV vaccine . Previous work has shown that the humoral immune response to NiV vaccination is sufficient to protect Syrian hamsters against a lethal challenge with NiV [27 , 42] . Likewise , administration of a human neutralizing monoclonal antibody ( m102 . 4 ) provided full protection against both HeV and NiV in multiple animal models [51 , 52] . Administration of purified IgG obtained from ChAdOx1 NiVB vaccinated hamsters provided partial protection against NiV challenge . Furthermore , infectious virus could only be detected in the lungs of control animals and not in the lungs of vaccinated animals , and thus as in previous studies , ChAdOx1 NiVB-elicited antibodies are able to provide protection against a lethal challenge with NiV . Although we were able to detect NiV G protein-specific antibodies in serum obtained from NiV antibody-treated animals , the reciprocal titer was much lower than that detected in serum from Syrian hamsters after a single dose of ChAdOx1 NiVB . It is possible that administering a higher dose of IgG would have led to uniform protection . Two animals treated with IgG purified from animals which received injections with ChAdOx1 FGP survived a lethal challenge with NiV Bangladesh . Occasional survival has been observed in the Syrian hamster model [33] . The increased survival rate might however also reflect a non-specific effect of treatment with IgG , which has been reported previously [53] . As the survival rate was significantly different between the NiV IgG-treated group and the control IgG-treated group , the passive transfer experiment shows that antibodies elicited by ChAdOx1-NiVB are sufficient for protection against a lethal challenge with NiV . Animals in the passive transfer experiment were observed for 56 days , to ensure that the two animals that survived would not succumb to disease after 28 days . The Syrian hamster is a suitable initial small animal model to investigate the efficacy of NiV vaccines , followed by the African green monkey [54] . The immune system of African green monkeys is more like humans than that of hamsters and is therefore seen as a more relevant animal model to test NiV vaccines . Based on the results presented in the current manuscript , future studies are planned to test ChAdOx1 NiVB in African green monkeys , supported by the Coalition for Epidemic Preparedness Innovations ( CEPI ) . We show that ChAdOx1 NiVB provides complete protection against lethal disease in Syrian hamsters challenged with NiV Bangladesh . Furthermore , ChAdOx1 NiVB vaccination results in complete survival but with limited evidence of viral replication after NiV Malaysia challenge , and partial protection against HeV . Passive transfer of antibodies elicited by ChAdOx1 NiVB vaccination provide partial protection against lethal challenge with NiV Bangladesh .
|
Nipah virus was discovered in 1998 after an outbreak in Malaysia . Since then , several outbreaks have been reported in Bangladesh and India . Although most outbreaks are relatively small , a very high case-fatality rate is reported ( 75% ) . Furthermore , human-to-human transmission has been reported . Currently , no approved vaccine or countermeasure exist . In this manuscript , we discuss a vaccine based on a chimpanzee adenovirus . Importantly , the vaccine vector ( ChAdOx1 ) is in clinical trials . In the work presented here , we show that this vaccine is fully protective against both genotypes of Nipah virus . Furthermore , we observe partial protection against Hendra virus , a related virus . Antibodies produced upon vaccination with our vaccine alone are partially protective against Nipah virus . This is an important step forwards towards the development of an approved vaccine for Nipah virus .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"enzyme-linked",
"immunoassays",
"respiratory",
"infections",
"immunology",
"geographical",
"locations",
"vertebrates",
"animals",
"mammals",
"pulmonology",
"vaccines",
"preventive",
"medicine",
"infectious",
"disease",
"control",
"antibodies",
"vaccination",
"and",
"immunization",
"immunologic",
"techniques",
"bangladesh",
"research",
"and",
"analysis",
"methods",
"public",
"and",
"occupational",
"health",
"hamsters",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"proteins",
"immunoassays",
"people",
"and",
"places",
"biochemistry",
"rodents",
"eukaryota",
"asia",
"physiology",
"biology",
"and",
"life",
"sciences",
"malaysia",
"amniotes",
"organisms"
] |
2019
|
A single-dose ChAdOx1-vectored vaccine provides complete protection against Nipah Bangladesh and Malaysia in Syrian golden hamsters
|
Chloroplast genomes in land plants contain approximately 100 genes , the majority of which reside in polycistronic transcription units derived from cyanobacterial operons . The expression of chloroplast genes is integrated into developmental programs underlying the differentiation of photosynthetic cells from non-photosynthetic progenitors . In C4 plants , the partitioning of photosynthesis between two cell types , bundle sheath and mesophyll , adds an additional layer of complexity . We used ribosome profiling and RNA-seq to generate a comprehensive description of chloroplast gene expression at four stages of chloroplast differentiation , as displayed along the maize seedling leaf blade . The rate of protein output of most genes increases early in development and declines once the photosynthetic apparatus is mature . The developmental dynamics of protein output fall into several patterns . Programmed changes in mRNA abundance make a strong contribution to the developmental shifts in protein output , but output is further adjusted by changes in translational efficiency . RNAs with prioritized translation early in development are largely involved in chloroplast gene expression , whereas those with prioritized translation in photosynthetic tissues are generally involved in photosynthesis . Differential gene expression in bundle sheath and mesophyll chloroplasts results primarily from differences in mRNA abundance , but differences in translational efficiency amplify mRNA-level effects in some instances . In most cases , rates of protein output approximate steady-state protein stoichiometries , implying a limited role for proteolysis in eliminating unassembled or damaged proteins under non-stress conditions . Tuned protein output results from gene-specific trade-offs between translational efficiency and mRNA abundance , both of which span a large dynamic range . Analysis of ribosome footprints at sites of RNA editing showed that the chloroplast translation machinery does not generally discriminate between edited and unedited RNAs . However , editing of ACG to AUG at the rpl2 start codon is essential for translation initiation , demonstrating that ACG does not serve as a start codon in maize chloroplasts .
The evolution of chloroplasts from a cyanobacterial endosymbiont was accompanied by a massive transfer of bacterial genes to the nuclear genome , and by the integration of chloroplast processes into the host’s developmental and physiological programs [1] . In multicellular plants , chloroplasts differentiate from non-photosynthetic proplastids in concert with the differentiation of meristematic cells into photosynthetic leaf cells . This transformation is accompanied by a prodigious increase in the abundance of the proteins that make up the photosynthetic apparatus , which contribute more than half of the protein mass in photosynthetic leaf tissue [2] . Both nuclear and chloroplast genes contribute subunits to the multisubunit complexes that participate in photosynthesis . The expression of these two physically separated gene sets is coordinated by nucleus-encoded proteins that control chloroplast gene expression , and by signals emanating from chloroplasts that influence nuclear gene expression [1 , 3] . Beyond these general concepts , however , little is known about the mechanisms that coordinate chloroplast and nuclear gene expression in the context of the proplastid to chloroplast transition . Furthermore , a thorough description of the dynamics of chloroplast gene expression during this process is currently lacking . Despite roughly one billion years of evolution , the bacterial ancestry of the chloroplast genome is readily apparent in its gene organization and gene expression mechanisms . Most chloroplast genes in land plants are grouped into polycistronic transcription units [4] that are transcribed by a bacterial-type RNA polymerase [5] and translated by 70S ribosomes that strongly resemble bacterial ribosomes [6] . As in bacteria , chloroplast ribosomes bind mRNA at ribosome binding sites near start codons , sometimes with the assistance of a Shine-Dalgarno element [6] . Superimposed on this ancient scaffold are numerous features that arose post-endosymbiosis [7] . For example , a phage-type RNA polymerase collaborates with an RNA polymerase of cyanobacterial origin [5] , and chloroplast RNAs are modified by RNA editing , RNA splicing , and other events that are either unusual or absent in bacteria [8] . Ribosome profiling data from E . coli revealed that the rate of protein output from genes encoding subunits of multisubunit complexes is proportional to subunit stoichiometry , and that proportional synthesis is typically achieved by differences in the translational efficiency of genes residing in the same operon [9 , 10] . As the majority of chloroplast gene products are components of multisubunit complexes , it is of interest to know whether similar themes apply . Furthermore , the gene content of polycistronic transcription units in chloroplasts has diverged from that in the cyanobacterial ancestor . Has “tuned” protein output been maintained in chloroplasts despite this disrupted operon organization ? If so , what mechanisms achieve this tuning in light of the new gene arrangements and the new features of mRNA metabolism ? In this work , we used ribosome profiling to address these and other questions of chloroplast gene regulation in the context of the proplastid to chloroplast transition . For this purpose , we took advantage of the natural developmental gradient of the maize seedling leaf blade , where cells and plastids at increasing stages of photosynthetic differentiation form a developmental gradient from base to tip [11] . By using the normalized abundance of ribosome footprints as a proxy for rates of protein synthesis , we show that the rate of protein output from many chloroplast genes is tuned to protein stoichiometry , and that tuned protein output is achieved through gene-specific balancing of mRNA abundance with translational efficiency . This comprehensive analysis revealed developmentally programmed changes in translational efficiencies , which superimpose on programmed changes in mRNA abundance to shift the balance of protein output as chloroplast development proceeds .
We analyzed tissues from the same genetic background and developmental stage as used in previous proteome [2] and nuclear transcriptome [12 , 13] studies of photosynthetic differentiation in maize . Four leaf sections were harvested from the third leaf to emerge in 9-day old seedlings ( Fig 1A ) : the leaf base ( segment 1 ) , which harbors non-photosynthetic proplastids; 3–4 cm above the base ( segment 4 ) , representing the sink-source transition and a region of active chloroplast biogenesis; 8–9 cm above the base ( segment 9 ) , representing young chloroplasts; and a section near the tip ( segment 14 ) harboring mature bundle sheath and mesophyll chloroplasts [2 , 12] . The developmental transitions represented by these fractions are illustrated in the immunoblot assays shown in Fig 1B . The mitochondrial protein Atp6 is most abundant in the two basal sections , subunits of photosynthetic complexes ( AtpB , PetD , PsaD , PsbA , NdhH , RbcL ) are most abundant in the two apical sections , and a chloroplast ribosomal protein ( Rpl2 ) exhibits peak abundance in the two middle sections . These developmental profiles are consistent with prior proteome data [2] . To explore the contribution of differential chloroplast gene expression to the distinct proteomes in bundle sheath and mesophyll cells , we also analyzed bundle sheath and mesophyll-enriched fractions from the apical region of seedling leaves . Standard protocols for the separation of bundle sheath and mesophyll cells involve lengthy incubations that are likely to cause changes in ribosome position . We used a rapid mechanical fractionation method that minimizes the time between tissue disruption and the generation of ribosome footprints ( see Materials and Methods ) . Markers for each cell type were enriched 5- to 10-fold in the corresponding fraction ( Fig 1C ) . This degree of enrichment is comparable to that of the fractions used to define mesophyll and bundle sheath-enriched proteomes in maize [14] . We modified our previous method for preparing ribosome footprints from maize leaf tissue [15] to reduce the amount of time and tissue required , and to reduce contamination by non-ribosomal ribonucleoprotein particles ( RNPs ) . In brief , leaf tissue was flash frozen and ground in liquid N2 , thawed in a standard polysome extraction buffer , and treated with Ribonuclease I to liberate monosomes . Ribosomes were purified by pelleting through a sucrose cushion under conditions that leave chloroplast group II intron RNPs ( ~600 kDa ) [16] in the supernatant ( S1A Fig ) . RNAs between approximately 20 and 35 nucleotides ( nt ) were gel purified and converted to a sequencing library with a commercial small RNA library kit that has minimal ligation bias [17] . rRNA contaminants were depleted after first strand cDNA synthesis by hybridization to biotinylated oligonucleotides designed to match abundant contaminants detected in pilot experiments ( S1 Table ) . Approximately 35 million reads were obtained for each “Ribo-seq” replicate , roughly 50% of which aligned to mRNA ( S2 Table ) . RNA-seq data was generated from RNA extracted from aliquots of each lysate taken prior to addition of RNAse I . Replicate RNA-seq and Ribo-seq assays showed high reproducibility ( Pearson correlation of >0 . 98 , S2 Fig ) . Almost all plastid genes were represented by at least 100 reads per replicate in all datasets ( S3 Fig ) . Several clusters of low abundance reads mapped to small unannotated ORFs , but further investigation is required to evaluate which , if any , of these are the footprints of translating ribosomes . Ribosomes in the cytosol , mitochondria , and chloroplasts have distinct genetic origins . Accordingly , the ribosome footprints from each compartment displayed different size distributions ( Fig 2A ) . The cytosolic ribosome footprints showed a minor peak at 23 nucleotides and a major peak at 31 nucleotides , similar to observations in yeast [18] . The mitochondrial data showed a major peak at 28–29 nucleotides and a minor peak at 36 nucleotides , similar to the 27 and 33-nt peaks reported for human mitochondria [19] . The plastid ribosome footprints had a broad size distribution suggestive of two populations , with peaks at approximately 30 and 35 nucleotides . A similar distribution was observed in pilot experiments involving the gel purification of RNAs up to 40-nt ( S1B Fig ) indicating that the peak at 35-nt was not an artifact of our gel purification strategy . A broad and bimodal size distribution was also observed for chloroplast ribosome footprints from the single-celled alga Chlamydomonas reinhardtii , albeit with peaks at slightly different positions [20] . The two prior reports of ribosome footprint size distributions in plants [21 , 22] did not parse the data from the three compartments , but the 31-nucleotide modal size reported in those studies is consistent with our data . Our data show the 3-nucleotide periodicity expected for ribosome footprints ( Fig 2B and 2C ) . Interestingly , the degree of periodicity varies with footprint size ( S4 Fig ) . The reads are largely restricted to open reading frames in the cytosol ( Fig 2C ) and chloroplast ( Fig 2D ) . Taken together , these results provide strong evidence that the vast majority of the Ribo-seq reads come from bona-fide ribosome footprints . The placement of ribosome P and A sites with respect to ribosome footprint termini has not been reported for any organellar ribosomes or for cytosolic ribosomes in maize . A meta analysis of our data showed that the position of the 3’ end of ribosome footprints from initiating and terminating ribosomes in chloroplasts and mitochondria is constant with respect to start and stop codons , respectively , regardless of footprint size; however , the position of the 5’ ends varies with footprint size ( Fig 2E , S4C Fig ) . Therefore , the positions of the A and P sites in organellar ribosomes can be inferred based on the 3’-ends of their footprints , as is also true for bacterial ribosomes [23 , 24] . The modal distance between the start of the P site in chloroplast ribosomes and the 3’-ends of chloroplast ribosome footprints is 7 nucleotides . By contrast , cytosolic ribosome footprints are approximately centered on the P site regardless of footprint size ( S4B Fig ) . The partitioning of ribosome footprints among the three genetic compartments shifts dramatically during the course of leaf development ( Fig 2G ) . The contribution of cytosolic translation drops from 99% at the leaf base to 57% in the apical leaf sections due to the increasing contribution of ribosome footprints from chloroplasts . This shift of cellular resources towards chloroplast translation corresponds with the massive increase in the content of photosynthetic complexes harboring plastid-encoded subunits ( Rubisco , PSII , PSI , cytochrome b6f , ATP synthase , NDH ) ( Fig 1 ) . Ribosome footprints from mitochondria accounted for a very small fraction of the total at all stages . However , our protocol was not optimized for the quantitative recovery of mitochondrial ribosomes so these data may not reflect the total mitochondrial ribosome population . In the discussion below we define the “translational output” of a gene as the abundance of ribosome footprints per kb per million reads mapped to nuclear coding sequences ( RPKM ) , and we use this value to compare rates of protein synthesis among genes on a molar basis . This is a typical interpretation of Ribo-seq data , and it is based on evidence that the bulk rate of translation elongation on all ORFs is similar under any particular condition , despite the fact that ribosome pausing can lead to the over-representation of ribosomes at specific positions [9 , 25] . Although this may be an over simplification in some instances , this interpretation of our data produced results that are generally coherent with current understanding of chloroplast biogenesis ( see below ) . Group II introns interrupt eight protein-coding genes in maize chloroplasts . These present a challenge for data analysis because the unspliced transcripts make up a substantial fraction of the RNA pool [16] and translation can initiate on unspliced RNAs and terminate within introns [15] . We therefore calculated translational output based solely on the last exon ( normalized to exon length ) . Data summaries presented below include RNA-seq data only for that subset of intron-containing genes for which multiple methods of analysis provided consistent values for the abundance of spliced RNA isoforms ( see Materials and Methods ) . Fig 3 summarizes the abundance of Ribo-seq and RNA-seq reads from protein-coding chloroplast genes in each of the four leaf segments . To display the low values from Segment 1 , they are replotted with a smaller Y-axis scale in S5 Fig . The abundance of mRNA from genes in the same transcription unit ( Fig 3A and S5A Fig , bracketed arrows ) is typically similar , but the protein output of co-transcribed genes varies considerably . Translational efficiency ( translational output /mRNA abundance ) varies widely among genes ( Fig 3A and S5A Fig , bottom ) . The atpH mRNA is the most efficiently translated of any chloroplast mRNA at all four developmental stages , surpassing even psbA , whose product is the most rapidly synthesized protein in photosynthetic tissues [26] . Prodigious psbA expression results from very high mRNA abundance in combination with a translational efficiency that is comparable to that of other photosystem genes . When the data are grouped according to gene function , correlations between function and translational output become apparent ( Fig 3B ) . For example , the translational output of genes encoding subunits of ribosomes and the NDH complex are consistently very low , whereas the translational output of genes encoding subunits of PSI , PSII , the ATP synthase , and the cytochrome b6f complex are consistently much higher . These trends mirror the abundance of these complexes as inferred from proteome data [27] . The data for complexes whose subunits are not found in a 1:1 ratio show further that translational output is tuned to subunit stoichiometry . For example , the chloroplast-encoded subunits of the ATP synthase ( AtpA , AtpB , AtpE , AtpF , AtpH , AtpI ) are found in a 3: 3: 1: 1: 14: 1 molar ratio in the complex [28 , 29] . The translational output of their genes mirrors this stoichiometry quite well , whereas mRNA abundance does not ( Fig 4A ) . These genes are distributed between two transcription units ( Fig 4A ) . A single mRNA encodes AtpB and AtpE , whose rates of synthesis are tuned via differences in translational efficiency . The atpI-atpH-atpF-atpA primary transcript is processed to yield various smaller isoforms [30] but the abundance of RNA from each gene is nonetheless quite similar ( Fig 4A ) . The translational output of the atpH gene is boosted relative to that of its neighbors primarily through exceptionally high translational efficiency ( Fig 4A bottom ) . In a second example , the unequal stoichiometry of subunits of the plastid-encoded RNA polymerase ( PEP ) ( 2 RpoA:1 RpoB:1 RpoC1:1 RpoC2 ) [5] is mirrored by the relative translational output of the corresponding genes ( Fig 4B ) . In this case , however , tuning occurs primarily at the level of mRNA accumulation . The plastid-encoded subunits of PSI , PSII , the cytochrome b6f complex , the NDH complex , and chloroplast ribosomes are found in equal numbers in their respective complex . Genes encoding subunits of each of these complexes are distributed across multiple transcription units , many of which also encode subunits of other complexes . This gene organization sometimes results in considerable disparity in mRNA level among subunits of the same complex ( Fig 3B bottom ) . In general , such differences are buffered by opposing changes in translational efficiency , such that translational outputs more closely reflect protein stoichiometry than does mRNA abundance ( see , for example , the NDH complex in S6B Fig ) . In the case of PSI ( Fig 4C ) , the structural genes ( psaA , psaB , psaC , psaJ , psaI ) exhibit an approximately three-fold range of translational output , but all of these genes vastly out produce two genes encoding PSI assembly factors ( ycf3 and ycf4 ) [31–33] . The psaI and ycf4 genes are adjacent in the same polycistronic transcription unit ( Fig 4C bottom ) , and their difference in translational output is programmed primarily by a difference in translational efficiency . The translational output of psbN , which encodes a PSII assembly factor [34] , is likewise much less than that of structural genes for PSII ( Fig 4D ) . Taken together , this body of data shows that the tuning of translational output to protein stoichiometries is accomplished via trade-offs between mRNA level and translational efficiency , with this balance differing from one gene to the next . Where mRNA abundance closely matches protein stoichiometry , differences in translational efficiency make only a small contribution ( as observed for rpoA , rpoB , rpoC1 and rpoC2 ) . Where mRNAs are severely out of balance with protein stoichiometry , differences in translational efficiency compensate . The translational output of PSII structural genes is well matched , with the notable exception of psbA ( Fig 4D ) , whose output vastly exceeds that of other genes in photosynthetic leaf segments ( segments 9 and 14 ) . This behavior is consistent with the known properties of the psbA gene product , whose damage and rapid turnover during active photosynthesis is compensated by a high rate of synthesis to support PSII repair [26] . Setting psbA aside , the relative translational outputs of other genes only approximate the stoichiometries of their products: several-fold differences between relative output and stoichiometry are common among subunits of a particular complex , suggesting that proteolysis of unassembled subunits serves to fine-tune protein stoichiometries . It is also possible that the calculated translational outputs do not perfectly reflect rates of protein synthesis due to differences in translation elongation rates among mRNAs . That said , instances in which translational outputs are particularly discordant among subunits of the same complex are worthy of note , as this may reflect physiologically relevant behaviors . For example , the translational output of ndhK is balanced with other ndh genes in non-photosynthetic leaf segments but ndhK substantially out produces the other ndh genes in mature chloroplasts ( S6B Fig ) . This behavior is reminiscent of psbA , and suggests that NdhK may be damaged and replaced during active photosynthesis . To explore the dynamics of chloroplast gene expression during the proplastid to chloroplast transition , we calculated standardized values for translational output , mRNA abundance and translational efficiency such that developmental shifts can be compared despite large differences in signal magnitude . This analysis shows that the developmental dynamics of translational output varies widely among genes ( Fig 5A top ) . The standardized values were used as the input for hierarchical clustering , which produced four clusters from the translational output data , four from the mRNA data , and five from the translational efficiency data ( Fig 5B , S7 Fig ) . The genes in each cluster are identified by color in Fig 5A . Although the transitions between clusters are not marked by obvious distinctions , the distinct trends defining each cluster are clear in the plots in Fig 5B . Genes whose translational output and mRNA abundance peak early in development ( segment 4 ) generally encode components of the chloroplast gene expression machinery ( rpl , rps , rpo , matK ) ( Fig 5A and 5C ) . Most genes encoding components of the photosynthetic apparatus ( psb , psa , atp , pet genes ) have peak mRNA and translational output in young chloroplasts ( segment 9 ) . A handful of photosynthesis genes either maintain or increase translational output and mRNA in mature chloroplasts ( segment 14 ) ( Fig 5A and 5C ) . There is considerable similarity among the clusters produced from the translational output and mRNA data ( Fig 5A and 5B ) , implying that programmed changes in mRNA abundance underlie the majority of developmental shifts in translational output . However , changes in translational efficiency also influence the developmental shifts in translational output ( Fig 5A bottom ) . In general , ORFs encoding proteins involved in photosynthesis are more efficiently translated later in development and those encoding gene expression factors are more efficiently translated early in development , albeit with numerous exceptions ( Fig 5A bottom , 5C right ) . Transcription units that encode both photosynthesis and gene expression factors provide revealing examples of distinct translational dynamics . In the psaA-psaB-rps14 transcription unit , for example , rps14 is found in a translational output cluster with other genes involved in gene expression , whereas psaA and psaB reside in a translational output cluster with other photosynthesis genes ( Fig 5A top ) . This results from distinct developmental shifts in translational efficiency: the rps14 ORF is translated more efficiently early in development whereas psaA and psaB are more efficiently translated later in development ( Fig 5D ) . The psaI-ycf4-cemA-petA transcription unit provides a second example . The translational output of psaI , cemA , and petA show similar developmental dynamics , but ycf4 clusters with different genes due to more efficient translation earlier in development ( Fig 5D ) . Again , these distinct patterns correlate with function , as psaI and petA encode components of the photosynthetic apparatus , whereas ycf4 encodes an assembly factor for PSI [31 , 32] . Many polycistronic RNAs in chloroplasts are processed to smaller isoforms . Although the impact of processing on translational efficiencies remains unclear [35 , 36] , it is plausible that programmed changes in the accumulation of processed isoforms could uncouple the expression of cotranscribed genes during development . To address this possibility , we used RNA gel blot hybridization to analyze transcripts from two transcription units that include genes whose translational efficiencies exhibit distinct developmental dynamics: psaI-ycf4-cemA-petA and psaA-psaB-rps14 transcription units ( S8 Fig ) . Processed rps14-specific transcripts accumulate preferentially in immature chloroplasts ( segment 4 ) , correlating with the stage at which rps14 is most efficiently translated . Analogously , a monocistronic psaI isoform accumulates preferentially in segments 4 and 9 where psaI is most efficiently translated . Various cause and effect relationships may underlie these correlations , as is discussed below . In maize and other C4 plants , photosynthesis is partitioned between mesophyll ( M ) and bundle sheath ( BS ) cells . Three protein complexes that include plastid-encoded subunits accumulate differentially in the two cell types: Rubisco and the NDH complex are enriched in BS cells whereas PSII is enriched in M cells [2 , 14] . Differential accumulation of several chloroplast mRNAs in the two cell types has been reported [37–41] , but a comprehensive comparison of chloroplast gene expression in BS and M cells has been lacking . To address this issue we performed RNA-seq and Ribo-seq analyses of BS- and M- enriched leaf fractions . The translational output of genes encoding subunits of Rubisco , PSII , and the NDH complex ( Fig 6A ) correlated well with the relative abundance of subunits of these complexes in the same sample preparations ( Fig 1C ) , and with quantitative proteome data [2] . Cell-type specific differences in mRNA accumulation ( Fig 6B ) can account for many of the differences in translational output ( Fig 6A ) , indicating that differences in transcription and/or RNA stability make a strong contribution to preferential gene expression in one cell type or the other . However , the data suggest that differences in translational efficiency contribute in certain instances ( Fig 6C ) . Four genes encoding PSII core subunits ( psbA , psbB , psbC , psbD ) provide the most compelling examples , as their translational output is considerably more biased toward M cells than are their mRNA levels . Organellar RNAs in land plants are often modified by an editing process that converts specific cytidine residues to uridine [42 , 43] . Some sites are inefficiently edited , which raises the question of whether the translation machinery discriminates between edited and unedited RNAs . The protein products of several unedited mitochondrial RNAs have been detected in plants [44 , 45] . We used our Ribo-seq and RNA-seq data to examine this issue for chloroplast RNAs . Fig 7 summarizes the data for those sites of editing that are represented by at least 100 reads in both the Ribo-seq and RNA-seq data in at least two replicates ( 17 of the 28 edited sites in the maize chloroplast transcriptome ) . In general , the percent editing was similar in the RNA-seq and Ribo-seq data , implying little discrimination between edited and unedited RNAs by the translation machinery . There were , however , two major exceptions: rpl2 ( nt 2 ) and ndhA ( nt 563 ) . In these cases a large fraction of the RNA-seq reads came from unedited RNA , whereas virtually all of the Ribo-seq reads came from edited sites . These two sites have unusual features that can account for the preferential translation of the edited RNAs . Editing at the ndhA site is linked to the splicing of the group II intron in the ndhA pre-mRNA: the site is not edited in unspliced transcripts and it is fully edited in spliced transcripts [46–48] . Failure to edit unspliced RNA is presumably due to the position of the intron between the edited site and the cis-element that specifies it . Translation that initiates on unspliced ndhA RNA would terminate at an in-frame stop codon within the intron . Thus , exon 2 is translated only from spliced RNAs , and these are 100% edited . In the case of rpl2 , the editing event creates an AUG start codon from an ACG precursor; this is the only editing event in maize chloroplasts that creates a canonical start codon . Although it has been reported that ACG can function as a start codon in chloroplasts [49 , 50] , our data show that this particular ACG is strongly discriminated against by initiating ribosomes . The fact that the Ribo-seq data show the expected strong bias toward edited rpl2 and ndhA ( 563 ) instills confidence that valid conclusions can be made from our data for other edited sites . Approximately 40% of the petB and ndhA ( nt 50 ) sequences are unedited in both the RNA-seq and Ribo-seq data , indicating that these unedited sequences give rise to a considerable fraction of the translational output of the corresponding genes . Editing of the petB site is essential for the function of its gene product ( cytochrome b6 ) [51] . It seems likely that the product of this unedited RNA is either unstable or selected against during complex assembly , as has also been suggested for the products of two unedited transcripts in mitochondria [52 , 53] . The remaining sites show almost complete editing in the RNA-seq data and , as expected , in the Ribo-seq data as well . That said , there is an overall trend toward less representation of unedited sequences in the Ribo-seq data than in the RNA-seq data . This may simply be a kinetic effect as would be expected if ribosome binding is slow in comparison to editing , such that ribosomes generally translate older ( and therefore more highly edited ) mRNAs .
Ribosome profiling data from bacteria revealed a striking correspondence between the stoichiometry of subunits of multisubunit complexes and their relative rates of synthesis [9 , 10] . Our results show that the relative translational outputs of chloroplast genes likewise approximate the relative abundance of the gene products . This tuning is apparent when comparing sets of genes encoding different complexes ( e . g . compare genes encoding the low abundance NDH complex to genes encoding the highly abundant PSI and PSII complexes ) ( Fig 3B ) , and when comparing genes encoding subunits of the same complex ( e . g . the PEP RNA polymerase and the ATP synthase ) ( Fig 4A and 4B ) . Our calculations of translational output rest on the assumption that the rate of translation elongation on all mRNAs is similar under any particular condition . This same assumption produced remarkable concordance between protein stoichiometry and inferred translational output in bacteria [9 , 10] . Although our results show a clear trend toward “proportional synthesis” , they also suggest that the tuning of protein output to stoichiometry is less precise in chloroplasts than it is in bacteria . Subunits of photosynthetic complexes are subject to proteolysis when their assembly is disrupted [55] , and a similar ( albeit wasteful ) mechanism could contribute to balancing stoichiometries when proteins are synthesized in excess under normal conditions . That said , instances in which inferred translational outputs are particularly incongruent with protein stoichiometries may reflect physiologically informative behaviors . The most prominent examples of “over-produced” proteins in our data are PsbA and PsbJ in PSII , PsaC and PsaJ in PSI , NdhK in the NDH complex , Rps14 in ribosomes , and PetD , PetL and PetN in the cytochrome b6f complex ( Fig 4 and S6 Fig ) . Disproportionate synthesis of PsbA is well known , and compensates for its damage and proteolysis during photosynthesis [26] . The other proteins suggested by our data to be produced in excess may likewise be subject to more rapid turnover than their partners in the assembled complex . A proteomic study in barley demonstrated that subunits of each photosynthetic complex generally turn over at similar rates [56] , but data for these particular proteins were not reported . Interestingly , the inferred rates of synthesis of PsbA , PetD , and NdhK are well matched to those of their partner subunits early in development , but outpace those of their partners in mature chloroplasts ( Fig 4D , S6 Fig ) . This feature of psbA expression coincides with the need to replace its gene product , D1 , following photo-induced damage and proteolysis [26] . By extension , the developmental dynamics of petD and ndhK expression suggest that their gene products may turn over more rapidly than their partners as a consequence of photosynthetic activity . In bacteria , proportional synthesis of subunits within a complex is achieved largely through the tuning of translational efficiencies among ORFs on the same mRNA [9 , 10] . In chloroplasts , genes encoding subunits of the same complex are generally distributed among multiple transcription units [4] and RNA segments within a transcription unit often accumulate to different levels [8] . It is interesting to consider how this shift in the gene expression landscape is reflected in the mechanisms that balance protein output among genes . In the case of the four genes encoding the PEP RNA polymerase , relative translational outputs closely match the 2:1:1:1 protein stoichiometry , and this is programmed primarily at the level of mRNA abundance ( Fig 4B ) . By contrast , widely varying translational efficiencies are superimposed on small variations in mRNA abundance to tune translational output to protein stoichiometry in the ATP synthase complex ( Fig 4A ) . Genes for ribosomal proteins are distributed among ten transcription units , several of which also encode proteins involved in photosynthesis ( see Fig 3A ) . For example , rps14 is cotranscribed with genes encoding the reaction center proteins of PSI ( psaA/psaB ) , and translational outputs within this transcription unit are balanced by large differences in translational efficiency ( Fig 4C ) . Similarly , the psaI transcription unit encodes subunits of the abundant PSI and cytochrome b6f complexes , a low abundance PSI assembly factor ( Ycf4 ) and a protein of unknown function ( CemA ) ; large differences in translational efficiency adjust the translational outputs to meet these different needs ( Fig 4C bottom ) . For complexes harboring plastid-encoded subunits in equal stoichiometries ( ribosomes , NDH , PSI , PSII , cytochrome b6f ) , compensating differences in translational efficiency generally buffer differences in mRNA level . Taken together , these results imply that mRNA abundance and translational efficiencies have coevolved in chloroplasts to produce proteins in close to the optimal amounts . In some instances , mRNA levels are sharply out of balance with protein stoichiometries , in which case differential translational efficiencies compensate . In other instances , mRNA levels approximate protein stoichiometries , and translational efficiencies are similar . These observations further suggest that for most genes in maize chloroplasts , mRNA levels and translational efficiencies are poised such that they limit the rate of protein synthesis to a similar extent . This view is further supported by the developmental dynamics discussed below . In Chlamydomonas chloroplasts , synthesis of subunits within the same photosynthetic complex is coordinated through assembly-dependent auto-regulatory mechanisms [57] . By contrast , current data for angiosperm chloroplasts suggest that translational efficiencies are generally independent of the assembly status of the gene products [15 , 58] . It seems likely that translational efficiencies are dictated by the interplay between the sequence and structure of RNA proximal to start codons and the proteins that bind this region . Translation initiation in chloroplasts sometimes involves a Shine-Dalgarno interaction and is facilitated by an unstructured translation initiation region [6 , 59] . Additionally , the translation of some chloroplast ORFs requires the participation of gene-specific translation activators [15 , 60–75] . Such proteins provide a means for tuning protein synthesis within and between transcription units . The atpH ORF and its nucleus-encoded translational activator PPR10 exemplify this mechanism . The exceptionally high translational efficiency of atpH ( Fig 3A ) boosts its translational output to match the high stoichiometry of AtpH in the ATP synthase complex ( Fig 4A ) ; this high translational efficiency requires the binding of PPR10 adjacent to the atpH ribosome binding site , an interaction that prevents the formation of inhibitory RNA structures involving the translation initiation region [15 , 30 , 62] . Our results provide a comprehensive view of the dynamics of chloroplast mRNA abundance and translation during the proplastid to chloroplast transition . The majority of genes involved in chloroplast gene expression exhibit peak mRNA abundance and translational output in developing chloroplasts ( segment 4 ) whereas the majority of genes encoding subunits of the photosynthetic apparatus exhibit peak mRNA abundance and translational output in young chloroplasts ( segment 9 ) ( Fig 5C ) . That said , even in proplastids ( segment 1 ) , genes involved in photosynthesis are generally represented by more mRNA and a higher translational output than are those involved in chloroplast gene expression ( S5 Fig ) . Our data show that programmed changes in translational efficiency combine with changes in mRNA abundance to produce developmental shifts in translational output ( Fig 5A ) . In general , translational efficiency is lowest at the leaf base , reflecting the low ribosome content in proplastids . The translational efficiency of most ORFs peaks in young chloroplasts ( segment 9 ) . In this context , it is intriguing that one subset of genes exhibit peak translational efficiency in the basal leaf segments ( Fig 5A , bottom; Fig 5C , right ) , whereas another subset increases in translational efficiency right out to the leaf tip ( Fig 5A and 5C ) . The former group is strongly enriched for “biogenesis” genes ( RNA polymerase , ribosomes , assembly factors ) , and the latter for photosynthesis genes . Possible mechanisms underlying these distinct “translational regulons” are discussed below . A study in Chlamydomonas showed that changes in chloroplast mRNA abundance are not reflected by corresponding changes in rates of protein synthesis , leading to the conclusion that translation is the primary rate-limiting step [76] . The data presented here suggest that this is not the case in maize chloroplasts . The developmental shifts in mRNA abundance were largely mirrored by shifts in translational output ( Fig 5A and 5B ) , implying that mRNA abundance has considerable impact on the output of most chloroplast genes in maize . Likewise , chloroplast DNA copy number limits gene expression in developing maize chloroplasts [77] but does not limit gene expression in Chlamydomonas chloroplasts [76] . It is perhaps unsurprising that mechanisms of gene regulation have diverged in the chloroplasts of vascular plants and single-celled algae , given their very different developmental and ecological contexts . Our data revealed a strong correlation between gene function and the developmental dynamics of mRNA abundance ( Fig 5C middle ) : mRNAs encoding proteins involved in gene expression generally peak in abundance earlier in development than do those encoding components of the photosynthetic apparatus . This finding was foreshadowed by analyses of several chloroplast mRNAs during leaf development in barley and Arabidopsis [78–80] . Land plant chloroplasts harbor two types of RNA polymerase , a single-subunit nucleus-encoded polymerase ( NEP ) and a bacterial-type plastid-encoded polymerase ( PEP ) [5] . The ratio of NEP to PEP drops precipitously during chloroplast development , and this likely makes a large contribution to the changes in chloroplast mRNA pools [5 , 78 , 80 , 81] . There is evidence that NEP plays an especially important role in the transcription of “house keeping” genes , and PEP in the transcription of photosynthesis genes [5]; however , most chloroplast genes can be transcribed by both NEP and PEP [82] , and the degree to which each polymerase contributes to the transcription of each gene during the course of chloroplast development remains unknown . Chloroplasts harbor several nucleus-encoded sigma factors that target PEP to distinct promoters [83] , and these provide an additional means to tune transcription rates in a developmental context . Changes in RNA stability combine with changes in transcription to modulate mRNA pools during chloroplast development [78 , 80 , 81 , 84 , 85] . Determinants of chloroplast mRNA stability include various ribonucleases , RNA structure , ribosome occupancy , and proteins that protect RNAs from nuclease attack [8] . Most mRNA termini in chloroplasts are protected by helical repeat RNA binding proteins that provide a steric blockade to exoribonucleases [8 , 61 , 62 , 86] . The majority of such proteins belong to the pentatricopeptide repeat ( PPR ) family , a large family of sequence-specific RNA binding proteins that influence virtually every post-transcriptional step in gene expression in mitochondria and chloroplasts [87] . In addition , chloroplasts harbor abundant hnRNP-like proteins , and these have been shown to impact the stability of several chloroplast mRNAs [88 , 89] . Programmed changes in the abundance and/or activities of PPR and hnRNP-like proteins might contribute to the shifting mRNA pools during the proplastid to chloroplast transition . Changes in translational efficiency superimpose on changes in mRNA abundance to modulate the output of plastid genes during the transformation of proplastids into chloroplasts . ORFs encoding proteins involved in photosynthesis generally exhibit maximal translational efficiency in young or mature chloroplasts ( segments 9 and 14 ) , whereas those that function in gene expression generally peak in translational efficiency earlier in development ( Fig 5C right ) . Furthermore , our data suggest that mRNAs encoding PSII reaction center proteins are translated with higher efficiency in mesophyll chloroplasts than in bundle sheath chloroplasts ( Fig 6C ) . It will be interesting to explore the mechanisms that underlie these differential effects on translational efficiency . Some possibilities include shifts in stromal pH , Mg++ , or the polymerase generating the mRNA ( NEP versus PEP ) , which might impact the formation of RNA structures at specific ribosome binding sites . Programmed changes in the activities of nucleus-encoded gene-specific translational activators could modulate translational efficiencies in a developmental context . Most such proteins in land plant chloroplasts are PPR ( or PPR-like ) proteins , and several of these also stabilize processed mRNAs with a 5’ end at the 5’ boundary of their binding site [15 , 30 , 60–62 , 64–67 , 90–93] . Indeed , many polycistronic transcripts in chloroplasts are processed to smaller isoforms whose ends are defined and stabilized by PPR-like proteins [7 , 8 , 86] . The impact of this type of RNA processing on translational efficiencies in vivo remains unclear . The removal of upstream ORFs is not required for the translation of several ORFs that are found on processed RNAs with a proximal 5’-terminus [35 , 36] . Some proteins have dual translation activation and RNA processing/stabilization functions , implying that the two activities are coupled [15 , 30 , 60–62 , 65 , 66 , 91–93]; however , the translation activation and RNA processing/ stabilization effects of such proteins could be independent consequences of their binding upstream of an ORF [62 , 86] . We showed here that there is a correlation between the accumulation of processed RNA isoforms and changes in relative translational efficiencies in two polycistronic transcription units ( S8 Fig ) . Deciphering the cause and effect relationships underling these correlations presents a challenge for the future . The data presented here lead to numerous new questions for future exploration . Is the synthesis of nucleus-encoded subunits of photosynthetic complexes tuned to that of their chloroplast-encoded partners ? What is the mechanistic basis for the preferential translation of some mRNAs in developing chloroplasts and others in photosynthetic chloroplasts ? To what extent do environmental inputs such as light and temperature modify the developmental dynamics of chloroplast mRNA abundance and translation ? The use of ribosome profiling can be anticipated to accelerate progress in addressing these and many other long-standing questions relating to the biology of organelles .
For the developmental analysis , Zea mays ( inbred line B73 ) was grown under diurnal cycles for 9 days and harvested as described [12] . Leaf sections from twelve plants were pooled for each of three replicates; each pool contained between 0 . 15 g and 0 . 3 g tissue . Plants used to prepare mesophyll and bundle sheath fractions were grown similarly , except the light was set at 300 μmol·m-2·s-1 and the tissues were harvested 13 days after planting , 2 hours into the light cycle . The apical one-third of leaf two and three were pooled from fifteen seedlings for each replicate , and the bundle sheath and mesophyll-enriched fractions were obtained with a rapid mechanical procedure . The tissue was cut into ~ 1 cm-sections , placed in a pre-chilled mortar and pestle , and lightly ground for 2 min in 5 ml of ice-cold modified polysome extraction buffer lacking detergents ( 0 . 2 M sucrose , 0 . 2 M KCl , 50 mM Tris-acetate , pH 8 . 0 , 15 mM MgCl2 , 20 mM 2-mercaptoethanol , 2 μg/ml pepstatin A , 2 μg/ml leupeptin , 2 mM phenylmethanesulfonyl fluoride , 100 μg/ml chloramphenicol , 100 μg/ml cycloheximide ) . The material that was released into solution constituted the mesophyll cell-enriched fraction . A portion of this was frozen in liquid N2 for RNA isolation , and the remainder was stored on ice while bundle sheath strands were purified from the tissue remaining in the mortar . The tissue was subject to four additional rounds of light grinding ( 2 min ) , each time in a fresh aliquot of 5-ml modified polysome extraction buffer . The light green fibers remaining constituted the bundle sheath enriched fraction; these cells were broken by hard grinding in 5 ml of modified polysome extraction buffer . A portion of this material was flash frozen for future RNA isolation and the remainder was used immediately for ribosome footprint isolation . Polyoxyethylene ( 10 ) tridecyl ether and Triton X-100 were added to the mesophyll and bundle sheath fractions retained for ribosome profiling ( final concentrations of 2% and 1% , respectively ) , and the material was filtered through glass wool . The isolation of ribosome footprints and total RNA were performed as described below . Ribosome footprints were prepared using a protocol similar to that described in [15] , but with two key modifications: ( i ) RNAse I rather than micrococcal nuclease was used to generate monosomes , and ( ii ) the centrifugation time used to pellet ribosomes through the sucrose cushion was shortened to reduce contamination by other RNPs . Tissues were pulverized in liquid N2 with a mortar and pestle , and thawed in 5 ml of polysome extraction buffer ( 0 . 2 M sucrose , 0 . 2 M KCl , 50 mM Tris-acetate , pH 8 . 0 , 15 mM MgCl2 , 20 mM 2-mercaptoethanol , 2% polyoxyethylene ( 10 ) tridecyl ether , 1% Triton X-100 , 100 μg/ml chloramphenicol , 100 μg/ml cycloheximide ) . A 2 . 4-ml aliquot was removed and frozen in liquid N2 for total RNA isolation . The remaining suspension was filtered through glass wool and centrifuged at 15 , 000xg for 10 min . The supernatant was digested with 3 , 500 units of RNAse I ( Ambion ) at 23°C for 30 min . 2 . 5 ml lysate was layered on a 2 ml sucrose cushion ( 1 M sucrose , 0 . 1 M KCl , 40 mM Tris-acetate , pH 8 . 0 , 15 mM MgCl2 , 10 mM 2-mercaptoethanol , 100 μg/ml chloramphenicol , and 100 μg/ml cycloheximide ) in a 16 x 76 mm tube and centrifuged in a Type 80 Ti rotor for 1 . 5 h at 55 , 000 rpm . The pellet was dissolved in 0 . 7 mL of ribosome dissociation buffer ( 10 mM Tris-Cl , pH 8 . 0 , 10 mM EDTA , 5 mM EGTA , 100 mM NaCl , 1% SDS ) . RNA was isolated with Tri reagent ( Molecular Research Center ) . RNAs between ~20 and ~35 nt were purified on a denaturing polyacrylamide gel , eluted , extracted with phenol/chloroform , precipitated with ethanol , and suspended in water . We have subsequently modified our protocol to purify RNAs between 20 and 40 nt; this results in a small shift in the size distribution of the reads ( S1B Fig ) . The ribosome footprint preparation was treated with T4 polynucleotide kinase . Twenty ng of the kinased RNA was converted to a sequencing library using the NEXTflex Small RNA Sequencing Kit v2 ( Bioo Scientific ) , which minimizes ligation bias by introducing four randomized bases at the 3’ ends of the adapters [17] . rRNA fragments were depleted by subtractive hybridization after first-strand cDNA synthesis , using 54 biotinylated DNA oligonucleotides corresponding to the most abundant rRNA fragments detected in pilot experiments ( see S1 Table ) . 10 μl of the oligonucleotide mixture ( concentrations as in S1 Table ) was added to 40-μl of the first-strand synthesis reaction and heated to 95°C for 2 min . A 50-μl aliquot of pre-warmed 2X hybridization buffer ( 10 mM Tris-Cl pH 7 . 5 , 1 mM EDTA , 2 M NaCl ) was added and incubated at 55°C for 30 min . The solution was transferred to a new tube containing 1 mg of prewashed Dynabeads M-270 Streptavidin ( Invitrogen ) and incubated at room temperature for 15 min with frequent agitation . The tube was placed on a magnet for 5 min and the supernatant was collected and desalted using Sephadex G-25 Fine ( GE Healthcare ) . The sample was concentrated to 18 μl and used as input for the PCR amplification step in the library construction protocol . After 14 cycles , PCR products were separated by electrophoresis through a 5% polyacrylamide gel and a gel slice corresponding to DNA fragments between markers at 147 and 180 bp ( representing insert sizes of 20–53 bp ) was excised . The DNA was eluted overnight , phenol/chloroform extracted , precipitated with ethanol , suspended in water , and stored at -20°C . For RNA-seq , rRNA was depleted from the RNA samples using the Ribo-Zero rRNA Removal Kit ( Plant Leaf ) ( Epicentre ) . One hundred ng of the rRNA-depleted RNA was used for library construction using the NEXTflex Rapid Directional qRNA-Seq Kit ( Bioo Scientific ) according the manufacturer’s instructions . The adapters provided with the kit include 8-nt molecular labels that were used during data processing to remove PCR bias . The libraries were combined and sequenced using a HiSeq 2500 or NextSeq 500 instrument ( Illumina ) . The read lengths were 50 or 75 nt for Ribo-seq and 75 nt for mRNA-seq . Adapter sequences were trimmed using cutadapt [94] . Ribo-seq reads between 18 and 40 nt were used as input for alignments . Alignments were performed using Bowtie 2 with default parameters [95] , which permits up to 2 mismatches , thereby allowing edited sequences to align . Reads were aligned to the following gene sets , with unaligned reads from each step used as input for the next round of alignment: ( i ) chloroplast tRNA and rRNA; ( ii ) chloroplast genome; ( iii ) mitochondrial tRNA and rRNA; ( iv ) mitochondrial genome ( B73 AGP v3 ) ; ( v ) nuclear tRNA and rRNA; nuclear genome ( B73 AGP v3 ) . Maize genome annotation 6a ( phytozome . jgi . doe . gov ) was reduced to the gene set annotated in 5b+ ( 60 , 211 transcripts ) ( gramene . org ) . For metagene analysis , all coding sequence ( CDS ) coordinates from all transcript variants were combined to make a union CDS coordinate . Custom Perl scripts extracted mapping information using SAMtools [96] and analyzed mapped reads as follows . The distribution of ribosome footprint lengths and the RPKM for both the Ribo-seq and RNA-seq data were calculated based only on reads mapping to CDS regions . For RPKM calculations , we defined the total number of mapped reads as the number of reads mapping to nuclear CDSs . Translation efficiency was calculated from the division of ribosome footprint RPKM by RNA-seq RPKM . Because unspliced RNAs constitute a substantial fraction of the RNA pool from intron-containing genes in chloroplasts , these genes require special treatment to infer the abundance of spliced ( functional ) mRNA . The fraction spliced at each intron was calculated in several ways . ( i ) RNA-seq reads were aligned to the chloroplast genome with splicing-aware software TopHat 2 . 0 . 11 [97] . The number of reads spanning each exon-exon junction ( spliced ) was divided by the sum of spliced ( exon-exon ) and unspliced ( exon 1-intron or intron-exon 2 ) reads; ( ii ) RNA-seq reads were aligned with Bowtie2 to a reference gene set that included both spliced and unspliced forms ( 100-nt on each side of each junction ) . The fraction of spliced RNA was calculated as for method ( i ) ; ( iii ) RNA-seq reads were aligned with TopHat to the genome and the spliced fraction was calculated from ( exon RPKM—intron RPKM ) /exon RPKM . Values calculated by each method are provided in S3 Table . Summary plots report mRNA abundance and translational efficiencies only for genes for which all of these methods gave similar results . We cannot confidently infer the amount of fully spliced RNAs from genes with two introns ( ycf3 and rps12 ) , so these are also excluded . Hierarchical clustering was performed using the Bioinformatic Toolbox of MATLAB software ( Mathworks ) using standardized values as input: values from the four leaf segments for each gene were standardized to have a mean of 0 and a standard deviation of 1 such that developmental shifts can be compared among genes despite differences in signal magnitude . Hierarchical clustering was performed using Pearson correlation coefficient values and unweighted average distance . Antibodies to AtpB , D1 and PetD were raised by our group and have been described previously [30] . Antibodies to Atp6 , NdhH , PPDK , and Rpl2 were generously provided by Christine Chase ( University of Florida ) , Tsuyoshi Endo ( Kyoto University ) , Kazuko Aoyagi ( UC Berkeley ) , and Alap Subramanian ( University of Arizona ) , respectively . Antibodies to PEPC , malic enzyme , and RbcL were generous gifts of William Taylor ( University of California , Berkeley ) . Illumina read sequences were deposited at the NCBI Sequence Read Archive with accession number SRP070787 . Alignments of reads to the maize chloroplast genome used Genbank accession X86563 .
|
Chloroplasts are subcellular organelles in plants and algae that carry out the core reactions of photosynthesis . Chloroplasts originated as cyanobacterial endosymbionts . Subsequent coevolution with their eukaryotic host resulted in a massive transfer of genes to the nuclear genome , the acquisition of new gene expression mechanisms , and the integration of chloroplast functions into host programs . Chloroplasts in multicellular plants develop from non-photosynthetic proplastids , a process that involves a prodigious increase in the expression of chloroplast genes encoding components of the photosynthetic apparatus . We used RNA sequencing and ribosome profiling to generate a comprehensive description of the dynamics of chloroplast gene expression during the transformation of proplastids into the distinct chloroplast types found in bundle sheath and mesophyll cells in maize . Genes encoding proteins that make up the chloroplast gene expression machinery peak in protein output earlier in development than do those encoding proteins that function in photosynthesis . Programmed changes in translational efficiencies superimpose on changes in mRNA abundance to shift the balance of protein output as chloroplast development proceeds . We also mined the data to gain insight into general features of chloroplast gene expression , such as relative translational efficiencies , the impact of RNA editing on translation , and the identification of rate limiting steps in gene expression . The findings clarify the parameters that dictate the abundance of chloroplast gene products and revealed unanticipated phenomena to be addressed in future studies .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"plant",
"cell",
"biology",
"messenger",
"rna",
"chloroplasts",
"dna",
"transcription",
"plant",
"science",
"molecular",
"biology",
"techniques",
"photosynthesis",
"cellular",
"structures",
"and",
"organelles",
"genetic",
"footprinting",
"research",
"and",
"analysis",
"methods",
"gene",
"expression",
"molecular",
"biology",
"ribosomes",
"biochemistry",
"rna",
"genetic",
"fingerprinting",
"and",
"footprinting",
"plant",
"biochemistry",
"plant",
"cells",
"cell",
"biology",
"nucleic",
"acids",
"protein",
"translation",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types"
] |
2016
|
Dynamics of Chloroplast Translation during Chloroplast Differentiation in Maize
|
Prophylactic vaccination is a powerful tool for reducing the burden of infectious diseases , due to a combination of direct protection of vaccinees and indirect protection of others via herd immunity . Computational models play an important role in devising strategies for vaccination by making projections of its impacts on public health . Such projections are subject to uncertainty about numerous factors , however . For example , many vaccine efficacy trials focus on measuring protection against disease rather than protection against infection , leaving the extent of breakthrough infections ( i . e . , disease ameliorated but infection unimpeded ) among vaccinees unknown . Our goal in this study was to quantify the extent to which uncertainty about breakthrough infections results in uncertainty about vaccination impact , with a focus on vaccines for dengue . To realistically account for the many forms of heterogeneity in dengue virus ( DENV ) transmission , which could have implications for the dynamics of indirect protection , we used a stochastic , agent-based model for DENV transmission informed by more than a decade of empirical studies in the city of Iquitos , Peru . Following 20 years of routine vaccination of nine-year-old children at 80% coverage , projections of the proportion of disease episodes averted varied by a factor of 1 . 76 ( 95% CI: 1 . 54–2 . 06 ) across the range of uncertainty about breakthrough infections . This was equivalent to the range of vaccination impact projected across a range of uncertainty about vaccine efficacy of 0 . 268 ( 95% CI: 0 . 210–0 . 329 ) . Until uncertainty about breakthrough infections can be addressed empirically , our results demonstrate the importance of accounting for it in models of vaccination impact .
Computational models have much to contribute to the advancement of vaccines as tools for public health benefit . These contributions range from aiding the design and interpretation of vaccine trials [1] to projecting the impact of vaccination policies on public health [2] . Projecting impact has been a major focus of modeling efforts over several decades [3] , with applications to a wide range of vaccine-preventable diseases . A challenge common to all of these projections is accounting for the many forms of uncertainty that are relevant to vaccination impact . These can include alternative scenarios for how vaccination could be targeted [4 , 5] , unknown aspects of the pathogen’s natural history [6 , 7] , and uncertainty about the vaccine’s profile [8 , 9] . Uncertainty about a vaccine’s profile is something that all projections of vaccination impact must confront . At a minimum , a vaccine’s profile is characterized by the relative risk , RR , of some outcome in vaccinees as compared to unvaccinated people . This quantity is related to vaccine efficacy , VE , as VE = 1-RR . Other aspects of a vaccine’s profile can include whether protection is “leaky” ( reduces per-event risk uniformly for all ) or “all or none” ( reduces risk fully , but only for some ) , or whether protection wanes over time or depends on an individual’s age or other characteristics . Uncertainty about these and other aspects of a vaccine’s profile can be addressed through sensitivity analysis [10 , 11] or by fitting a model to vaccine trial data [12 , 13] . By either approach , uncertainty about a vaccine’s profile can be propagated into uncertainty about vaccination impact . One vaccine with a complex profile for which impact projections [14] have played an important role in shaping recent policy decisions [15] is CYD-TDV ( brand name Dengvaxia , by Sanofi-Pasteur ) . This vaccine was developed to protect against dengue , a major viral disease of humans caused by any of four serotypes of dengue virus ( DENV ) and transmitted among humans by Aedes aegypti mosquitoes . Following a series of efficacy trials [16] , VE estimates for CYD-TDV were significantly lower for children under nine years of age ( 0 . 45 ) than for children nine years of age or older ( 0 . 66 ) . VE estimates were also greater for individuals with prior , natural exposure to DENV ( referred to as seropositive ) , especially for children under nine ( seropositive: 0 . 70 , seronegative: 0 . 14 ) . Additionally , VE estimates varied by serotype , ranging 0 . 34–0 . 62 in children under nine . Assumptions about vaccine profile used in projections of CYD-TDV vaccination impact to date have focused primarily on reconciling these age and serotype differences in VE [13 , 17] . One aspect of CYD-TDV profile that has not been explored in impact projections concerns the clinical nature of trial endpoints . Specifically , the primary endpoint for efficacy trials of CYD-TDV was virologically confirmed dengue among trial participants who experienced acute febrile illness; i . e . , a fever of ≥38°C for at least two consecutive days [16] . Among trial participants for whom acute febrile illness was averted due to vaccination , the vaccine could have either blocked DENV infection altogether or ameliorated symptoms but still allowed infection [18] . If we denote the relative risk of infection conditional on exposure as RRinf|exp and disease conditional on infection as RRdis|inf , it follows that relative risk of the disease endpoint is RRdis = RRdis|inf x RRinf|exp . Without measuring a secondary endpoint related to infection , RRinf|exp cannot be estimated and we are left knowing only the product RRdis . The distinction between RRdis|inf and RRinf|exp is an important one , because people with clinically inapparent DENV infections have been shown to transmit DENV to mosquitoes [19] and have been estimated to contribute appreciably to transmission [20] . This raises the possibility of breakthrough DENV infections among CYD-TDV vaccinees , particularly if RRinf|exp is large . Vaccines that prevent breakthrough infections can have a substantial impact on public health outcomes [21] , due to the fact that they provide both direct protection of vaccinees and indirect protection of others . Indirect protection derives from a population-level phenomenon known as herd immunity [22] , projections of which require assumptions about population-level transmission dynamics . To the extent that there is uncertainty about breakthrough infections among vaccinees , there will inevitably be uncertainty about the extent of indirect protection of those who go unvaccinated [23] . Our primary interest here was in assessing the extent of uncertainty in CYD-TDV impact projections attributable to uncertainty about breakthrough infections . The effect of breakthrough infections on vaccination impact is a function of the extent to which they erode indirect protection , which depends in part on the nature of contact between vaccinated and unvaccinated people and on the structure of transmission more generally [24 , 25] . To obtain a realistic portrayal of the structure of transmission in an endemic setting , we used an agent-based simulation model of DENV transmission developed and calibrated for the city of Iquitos , Peru , which has had ongoing studies of dengue epidemiology for more than a decade [26 , 27] . We simulated DENV transmission in the presence and absence of routine vaccination across a range of assumptions about breakthrough infections . To place these results into context , we compared them to results from simulations with varying values of VE under two different models of dengue vaccine profile .
Wherever possible , we parameterized the model directly based on empirical estimates from Iquitos or from studies conducted elsewhere and reported in the literature . This included human demography [31 , 32] , human mobility [34] , human-mosquito encounters [32] , mosquito abundance patterns in space and time [36 , 47 , 48] , mosquito movement [37] , mosquito mortality [35] , mosquito blood-feeding rates [38 , 39] , virus incubation in mosquitoes and humans [41] , infectiousness of humans to mosquitoes [44] , and naturally acquired immunity to DENV [45] . Two primary uncertainties that we were not able to quantify a priori were DENV importation into Iquitos and the scaling relationship between household mosquito surveys and true mosquito abundance . We calibrated those parameters such that simulated model behavior was consistent with empirical estimates of time-varying , serotype-specific patterns of incidence of DENV infection [49] . These empirical estimates to which our model was calibrated were based on interval-censored , serotype-specific seroconversions obtained through longitudinal cohort studies conducted over a period of 11 years [49] . The basis of our calibrations was maximization of the goodness of fit of simulated incidence Is , t of serotype s at time t to probabilistic estimates of Is , t by Reiner et al . [49] . For each month between January , 2000 and June , 2010 , we first performed maximum-likelihood fitting of a Dirichlet distribution to 1 , 000 draws of the serotype proportions of Is , t and a normal distribution to 1 , 000 draws of the total incidence It at time t from the posterior distribution estimated by Reiner et al [49] . We then used the product of the probability densities of those distributions evaluated at the simulated values of Is , t for each serotype as our measure of goodness of fit . Using this measure of goodness of fit , we obtained estimates of unknown parameters for DENV importation and scaling of mosquito abundance using a particle filtering algorithm . The premise of this algorithm is to make use of the fact that most of the unknown parameters pertain to only a portion of the time series—and thereby only a portion of the likelihood—to allow for calibrating different subsets of the unknown parameters sequentially rather than simultaneously . There are a wide range of particle filtering algorithms , but ours most closely resembles a sequential importance resampling algorithm [50] . The first step in our algorithm involved proposing a set of 1 , 000 initial particles spanning a range of parameter values , simulating the first year of the model for each particle , evaluating the goodness of fit measure described above on a monthly basis within the first year , and combining the monthly goodness of fit measures to obtain an annual goodness of fit measure for each particle for the first year . Next , we resampled the particles 1 , 000 times with replacement weighted by weighti=exp ( −cLLi ) ∑iexp ( −cLLi ) ( 1 ) for each particle i , where c is a scaling parameter that we tuned to a value of 0 . 1 to result in resampled particles containing at least 10% of the original particles . We then obtained maximum-likelihood estimates of the means and covariance matrix describing the distribution of the particles . Using that fitted multivariate normal distribution , we then drew 1 , 000 new particles and simulated both the first and second year of transmission . We then computed the goodness of fit measure for the first two years by aggregating monthly goodness of fit measures across the first two years . Additional steps in the algorithm were repeated in the same way in yearly increments through the last year for which empirical estimates of time-varying , serotype-specific incidence were available . Finally , we performed two additional rounds of resampling on the full time series following the last year of simulation and particle resampling . The resulting set of 1 , 000 particles constituted our distributional estimate of parameter values most consistent with available empirical estimates . For each of the two sets of assumptions about vaccine profile , we performed 1 , 000 pairs of simulations with model parameters drawn from the final set of parameter samples obtained through the calibration process . In both cases , we randomly sampled values of p between 0 and 1 for each simulation pair . For the CYD-TDV vaccine , we also sampled values of a parameter q that represents the quantile of the RRdis estimates . For the generic dengue vaccine , we sampled values of VEmean of 0 . 15–0 . 85 , VEserostatus of 0–0 . 15 , and VEserotype of 0–0 . 15 . Values of the latter three parameters were chosen to ensure that the maximum VE could not exceed 1 or fall below 0 and that a broad range of VEmean was covered . Parameter draws were performed with the sobol function in the pomp package [54] in R [52] to maximize the evenness of our coverage of parameter space . The two simulations in each pair exhibited identical dynamics for the first 11 years , because they were both driven by the same parameter particle and both shared common random number seeds for processes related to mosquito-human contact and DENV infection , respectively . Following that initial time period , we continued one simulation without vaccination but commenced the other with routine vaccination at age nine , both for an additional 20 years . Consistent with other CYD-TDV impact projections [14] , we assumed 80% coverage . For each simulation pair , we recorded the following in the population as a whole: proportion of cumulative infections averted and proportion of cumulative disease episodes averted , with both accruing over the period that followed the time period calibrated to Iquitos . Despite taking steps to reduce noise by controlling random number seeds , simulation results were still relatively noisy due to the highly stochastic nature of the model . To distinguish signal from noise when examining relationships between predictor and response variables in these simulation analyses , we fitted generalized additive models ( GAMs ) to these relationships using the mgcv package [55] in R . To assess the robustness of our conclusions about the effect of breakthrough infections on vaccination impact projections , we performed a series of sensitivity analyses . These analyses repeated the full process of model calibration and vaccination impact projection under alternative assumptions about several parameters: duration of cross-immunity , mosquito infectiousness , human infectiousness , mosquito movement probability , extrinsic incubation period , mosquito death rate , and mosquito biting rate ( Table 2 ) . Analyses identical to the primary analysis with default parameter values were performed on each of the sets of vaccination impact projections associated with each of these alternative parameter scenarios . Our primary interest in these sensitivity analyses was assessing the consistency of differences in infections and disease episodes averted across the range of values of parameters describing different aspects of vaccine profile .
Over the course of the calibration process , the marginal distributions of individual parameters narrowed relative to their starting ranges , particularly during three final iterations that made use of the full time series ( S1–S9 Figs ) . Correlations among parameters all started at zero but diversified over the course of successive iterations of the calibration process ( S10 Fig ) , suggesting that effects of some parameters on model performance interacted with effects of others . In general , the strongest correlations tended to be between the e0 and e1 parameters for scaling mosquito emergence and parameters describing DENV introduction patterns in years with large epidemics . To account for uncertainty associated with correlations among parameter estimates , replicate parameter draws sampling from the final set of parameters ( rightmost columns in S1–S9 Figs ) were used in all simulations subsequent to calibration . The behavior of the calibrated model ( Fig 3 , colored bands ) was largely in agreement with the estimates of time-varying , serotype-specific incidence of infection to which it was calibrated ( Fig 3 , gray bands ) . For all serotypes , the 95% prediction interval of simulations from our agent-based model and estimates from Reiner et al . [49] overlapped for the majority of the 2000–2010 timeframe . Both patterns reflected relatively low and seasonally variable patterns of DENV-1 and DENV-2 transmission , and both captured large seasonal peaks in 2002–2003 for DENV-3 and in 2009–2010 for DENV-4 , coinciding with the respective invasions of those serotypes . Note that y-axis ranges span a full order of magnitude ( Fig 3: 0 . 008–0 . 08 ) for panels corresponding to different serotypes . These results were largely similar under eleven alternative scenarios about model parameters described in Table 2 ( S11–S22 Figs ) , indicating that the model’s ability to reproduce dynamics from Iquitos was robust to these assumptions and that our algorithm for calibrating the model led to convergent estimates across multiple runs . Although the calibrated model was in relatively good agreement with estimates of incidence patterns by Reiner et al . [49] for large transmission seasons early in the occurrence of a given serotype , the model had a tendency to produce somewhat lower incidence patterns afterwards ( Fig 3 ) . These periods of lower incidence were associated with lower population susceptibility to a given serotype ( Fig 4 ) and tended to require a larger number of infections to seed transmission ( Fig 5 ) . The relatively low number of infections required to seed the larger epidemics was encouraging with regard to the model’s ability to reproduce large epidemics in Iquitos on the basis of the model’s characterization of local transmission processes . The relatively high number of infections required to seed inter-epidemic transmission is likely a result of the limited population size of 200 , 000 and the highly stochastic nature of transmission at times of low incidence . We obtained estimates of the parameters for relative risk in Eq ( 2 ) that best matched empirical estimates [16] of c1 = 0 . 47 , c2 = 0 . 148 , and c3 = 9 . 17 for seropositive vaccine recipients and c1 = 1 . 26 , c2 = 0 . 28 , and c3 = 9 . 27 for seronegative vaccine recipients . We obtained estimates of the parameters determining the standard error of the log of the risk ratio in Eq ( 3 ) of c4 = 100 and c5 = 0 . 5 . Under this model and with these parameters , relative risk decreased steeply with age until around age 20 , when it began to decrease more slowly towards almost no risk in older people ( Fig 6 ) . As in the CYD-TDV trial data , relative risk under our model was several fold lower in seropositive than seronegative children , and relative risk in excess of 1 was likely only at ages below nine years ( Fig 6 ) . Implementing either vaccine scenario in our simulations required an estimate of RR per event , rather than RR over the course of a trial [16] . For a leaky vaccine , these two different interpretations of RR may vary depending on how many times study participants are exposed [56] . In S2 Text , we showed that these values of RR are unlikely to differ for a dengue vaccine by more than 5% . Given that relatively small difference and in the absence of more detailed information about the number of exposures that participants experienced during CTD-TDV trials , we operated under the assumption that these two interpretations of RR were equal . Under our assumptions about how efficacy observed in trials derived from two different forms of protection , an assumption of equal parts protection against infection and protection against disease ( i . e . , p = 0 . 5 ) gave , on average , relative risks of 48% for either infection or disease in seropositive nine-year olds and 80% for either in seronegative nine-year olds ( Fig 6 ) . In the event that 90% of protection derived from protection against disease and only 10% from protection against infection ( i . e . , p = 0 . 1 ) , the relative risk for seropositive nine-year-olds was 27% for disease and 87% for infection and 68% for disease and 96% for infection for seronegative nine-year-olds ( Fig 6 ) . Under both models of vaccine profile that we considered , alternative parameter values yielded results that were largely similar to those under default parameter values . Across all scenarios with the CYD-TDV vaccine , differences in the proportion of infections and disease episodes averted were relatively large across the range of p but small across the range of q ( S24–S31 Figs ) . Across all scenarios with the generic dengue vaccine , differences in the proportion of infections and disease episodes averted were large across the ranges of p and VEmean and small across the ranges of VEserostatus and VEserotype ( S32–S39 Figs ) .
We developed an agent-based model for DENV transmission and used it to assess the extent of uncertainty in dengue vaccination impact projections attributable to uncertainty about breakthrough infections . Our analysis was not intended to represent a comprehensive assessment of the suitability of the CYD-TDV vaccine as a public health tool or to make a recommendation about its use . Instead , the value of this analysis is that it provides an assessment of the extent to which a potentially important source of uncertainty about vaccine profile might affect vaccination impact projections . Our results indicate that uncertainty about the extent to which a dengue vaccine prevents breakthrough infections makes a contribution to uncertainty about vaccination impact equivalent to not knowing whether VE is 0 . 70 or 0 . 43 . In the event that information about breakthrough infections remains lacking , limiting a model to a single assumption about this aspect of a vaccine’s profile could result in the communication of recommendations to decision makers that convey a false sense of confidence . Across two sets of assumptions about aspects of vaccine profile unrelated to breakthrough infections and thirteen different assumptions about model parameterization , our results were consistent in their suggestion that the projected impact of routine vaccination of nine-year olds is sensitive to the extent to which vaccination prevents breakthrough infections . These findings mirror conclusions from impact projections of vaccines for herpes simplex virus [10] , malaria [21] , and tuberculosis [57]; i . e . , that vaccines that protect against infection should result in greater indirect protection of unvaccinated people than vaccines that primarily protect against disease . Compared to numerical uncertainty about VE for CYD-TDV , uncertainty about breakthrough infections contributed much more to uncertainty about vaccination impact in our analysis . This does not mean that precise numerical estimates of VE are unimportant , but instead that uncertainty about VE following trials is typically low relative to uncertainty about the degree to which efficacy derives from protection against infection or protection against disease . Echoing this , our results using a generic dengue vaccine across a wide range of VE show that the numerical value of vaccine efficacy is indeed important . Efficacy trials substantially narrow that range , because that is precisely what they are designed to do . At present , however , they are generally not designed to narrow uncertainty about breakthrough infections . An important question following our analysis is whether , and how , information about protection against infection could be obtained empirically for a dengue vaccine . Disease is typically the preferred endpoint in efficacy trials , but arguments in favor of an infection endpoint can also be made in certain cases [58] . For example , the possibility of a secondary endpoint of infection has been proposed for efficacy trials for Zika vaccines [59] . Motivations for this include the infrequent occurrence of the disease outcome of greatest interest ( congenital Zika syndrome , CZS ) [60] and a clearer relationship between Zika virus infection and CZS than between more common disease outcomes and CZS [61] . At the same time , the short duration of Zika virus infection [62] makes reliable detection of active infection impractical , and issues such as cross-reactivity with other flaviviruses [63] make serological surrogates of infection problematic . Similar challenges would likely apply to estimation of efficacy against an infection endpoint for dengue vaccines . In the event that these challenges for endpoint measurement could be overcome , RRinf|exp could then be estimated directly ( e . g . , [64 , 65] ) . Until that happens , our work demonstrates that uncertainty in RRinf|exp can ( and should ) be accounted for in vaccination impact projections . The importance of understanding the extent of breakthrough infections owes to their critical role in determining the extent of indirect protection from vaccination . Still other factors are expected to further modulate indirect protection in a given context [22] . First , vaccination coverage—and , for that matter , naturally acquired immunity—clearly has an influence , with high coverage potentially compensating somewhat for imperfections in a vaccine’s profile [66] . Second , contact structure is thought to play an important role in determining the extent of indirect protection [24 , 25] . Recent modeling analyses of influenza [67] and onchocerciasis [68] showed that realistic assumptions about contact structure can lead to substantial differences in intervention impact projections . Although we view our own realistic portrayal of mosquito-human contact patterns as a strength of our model , the realism of a model’s assumptions does not necessarily translate into accuracy of a model’s predictions . To address that issue , uncertainty about model structure can be propagated into uncertainty about vaccination impact by combining projections from models with diverse structures and assumptions , as has been done recently for malaria [69] , dengue [14] , and pneumococcal disease [7] . Although our model of vaccine efficacy is consistent with several key findings from clinical trials of CYD-TDV [16] , such as serostatus- and age-dependent efficacy against disease , there are other findings that we did not account for . One notable feature of CYD-TDV that we have not considered pertains to protection against severe disease [70] . In particular , to the extent that vaccination serves as a “primary-like” infection in seronegative vaccine recipients [52] , the incidence of severe disease could increase as transmission is lowered by indirect effects of vaccination and the proportion of seronegative vaccine recipients increases [71] . Clinical trial data indicate , however , that whatever protection against severe disease the vaccine does afford may wane within a few years of vaccination [16] . In addition , direct protection against severe disease could affect indirect protection of unvaccinated individuals , given that severe disease could be associated with heightened [20] or attenuated [72] infectiousness . At the same time , associations between disease severity and infectiousness may trade off with other factors [73] , such as limited mobility [74] . In the end though , the relatively infrequent occurrence of severe disease may limit the overall impact of such effects on population-level transmission [20] . Although we expect that our overall conclusions have implications beyond the context of dengue in Iquitos , we note that our quantitative projections of cumulative proportions of infections and disease episodes averted are not directly applicable outside this setting . These numerical results could vary as a function of vaccination coverage , overall transmission intensity , and many other factors , similar to how estimates of vaccine efficacy can be context-dependent [75 , 76] . That said , our projections of CYD-TDV vaccination impact are in rough agreement with a study based on an earlier version of our model and seven others [14] . We hope that our model’s detailed representation of a well-studied , dengue-endemic population provides a tool for future studies to explore additional questions about vaccination impact in ways that acknowledge realistic variability in transmission patterns of the four DENV serotypes .
|
Vaccines are vital tools for safeguarding public health from a variety of infectious disease threats . When decisions are being made about investments in vaccination , computational models provide decision makers with projections of the benefits of vaccination . There are many types of uncertainty that can affect these projections , such as statistical uncertainty about the extent to which vaccination reduces one’s risk of experiencing disease . While this type of uncertainty is well accounted for by vaccine trials , a different type of uncertainty often is not: whether the vaccine blocks infection altogether or simply reduces the severity of disease symptoms . In the case of the latter , “breakthrough infections” occur , meaning that those who are vaccinated are protected but those who are not receive little or no indirect benefit from herd immunity . Focusing on a newly licensed vaccine for dengue , we developed and applied a new simulation model of dengue virus transmission to assess the extent to which uncertainty about breakthrough infections contributes to uncertainty about vaccination impact . We found that a vaccine that prevents breakthrough infections is capable of nearly doubling the impact of vaccination as compared to a vaccine that confers protection solely by reducing the severity of disease symptoms .
|
[
"Abstract",
"Introduction",
"Model",
"overview",
"Results",
"Discussion"
] |
[
"death",
"rates",
"invertebrates",
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"animals",
"simulation",
"and",
"modeling",
"vaccines",
"preventive",
"medicine",
"viruses",
"systems",
"science",
"mathematics",
"rna",
"viruses",
"population",
"biology",
"infectious",
"disease",
"control",
"vaccination",
"and",
"immunization",
"insect",
"vectors",
"research",
"and",
"analysis",
"methods",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"computer",
"and",
"information",
"sciences",
"medical",
"microbiology",
"agent-based",
"modeling",
"microbial",
"pathogens",
"disease",
"vectors",
"insects",
"arthropoda",
"population",
"metrics",
"mosquitoes",
"eukaryota",
"flaviviruses",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"physical",
"sciences",
"organisms"
] |
2019
|
An agent-based model of dengue virus transmission shows how uncertainty about breakthrough infections influences vaccination impact projections
|
We have asked how the common S34F mutation in the splicing factor U2AF1 regulates alternative splicing in lung cancer , and why wild-type U2AF1 is retained in cancers with this mutation . A human lung epithelial cell line was genetically modified so that U2AF1S34F is expressed from one of the two endogenous U2AF1 loci . By altering levels of mutant or wild-type U2AF1 in this cell line and by analyzing published data on human lung adenocarcinomas , we show that S34F-associated changes in alternative splicing are proportional to the ratio of S34F:wild-type gene products and not to absolute levels of either the mutant or wild-type factor . Preferential recognition of specific 3′ splice sites in S34F-expressing cells is largely explained by differential in vitro RNA-binding affinities of mutant versus wild-type U2AF1 for those same 3′ splice sites . Finally , we show that lung adenocarcinoma cell lines bearing U2AF1 mutations do not require the mutant protein for growth in vitro or in vivo . In contrast , wild-type U2AF1 is required for survival , regardless of whether cells carry the U2AF1S34F allele . Our results provide mechanistic explanations of the magnitude of splicing changes observed in U2AF1-mutant cells and why tumors harboring U2AF1 mutations always retain an expressed copy of the wild-type allele .
Somatic mutations in genes encoding four splicing factors ( U2AF1 , SF3B1 , SRSF2 and ZRSR2 ) have recently been reported in up to 50% of myelodysplastic syndromes ( MDS ) and related neoplasms and at lower frequencies in a variety of solid tumors [1–9] . Among these factors , only U2AF1 is known to be recurrently mutated in lung adenocarcinomas ( LUADs ) [3 , 9] . The only recurrent missense mutation of U2AF1 in LUAD affects codon 34 and always changes the conserved serine in a zinc knuckle motif to phenylalanine ( p . Ser34Phe , or S34F ) . This striking mutational consistency suggests a critical , yet unknown , role for U2AF1S34F during lung carcinogenesis . In addition , the wild-type ( WT ) U2AF1 allele is always retained in cancers with common U2AF1 mutations , including U2AF1S34F [2] . However , the functional significance of the wild-type allele in cells with mutant U2AF1 is not known . U2AF1 is a component of the U2 small nuclear ribonucleoprotein auxiliary factor complex ( U2AF ) [10 , 11] . During early spliceosome assembly , U2AF recognizes sequences at the 3′ ends of introns to facilitate the recruitment of the U2 small nuclear ribonucleoprotein ( snRNP ) complex to the 3′ splice site; the recruitment occurs in conjunction with recognition of the intronic branch point by splicing factor 1 ( SF1 ) [12 , 13] . In vitro crosslinking assays showed that U2AF1 contacts the AG dinucleotide at the intron-exon boundary and flanking sequences [14–16] . Consistent with the critical role that U2AF1 plays in RNA splicing , U2AF1 mutations are known to cause specific alterations in RNA splicing , most notably affecting the inclusion of cassette exons in mRNA [17–20] . However , the precise molecular basis of these splicing alterations , as well as how they are quantitatively regulated , is unknown . One possibility is that U2AF1 mutations cause altered RNA-binding affinity , resulting in altered splice site recognition . A computational model of the structure of the U2AF1:RNA complex suggested that Ser34 is a critical residue that contacts RNA [17] . Another study reported that U2AF1S34F exhibited altered affinity relative to the wild-type protein for RNA oligonucleotides derived from a cassette exon whose recognition is repressed in S34F-expressing cells [18] . Finally , the S34F mutation reportedly prevented a minimal fission yeast U2AF heterodimer from binding to a particular 3′ splice site RNA sequence [21] . However , it is not known whether altered RNA binding accounts for most S34F-associated splicing alterations and whether mechanisms other than altered binding control S34F-associated splicing . Here , we combine genetic and biochemical approaches to show that wild-type U2AF1 antagonizes the S34F-associated splicing program in lung epithelial cells . Analyses of the transcriptomes of primary LUAD samples as well as isogenic lung cells in culture indicate that the ratio of mutant to wild-type U2AF1 gene products is a critical determinant of the magnitude of S34F-associated changes in alternative splicing . S34F-associated splicing alterations can be largely explained by differences in the relative affinities of U2AF-SF1 complexes containing mutant versus wild-type U2AF1 for RNA containing the relevant 3′ splice sites . Moreover , we show that proliferation of cancer cells with U2AF1S34F is critically dependent on expression of the wild-type , but not the mutant , allele of U2AF1 .
Before undertaking experiments to study the effects of mutant U2AF1 in cultured cells , we first examined 512 transcriptomes from primary human LUADs published by The Cancer Genome Atlas ( TCGA ) [9] . Thirteen of these tumors harbor the most common U2AF1 mutation , S34F . ( Two others carry rare mutations of unknown significance , S65L and G216R , and were not considered further . ) We identified cassette exons whose inclusion was increased or decreased by ten percent or more in each tumor with the S34F mutation relative to the median inclusion of each cassette exon across all tumors without a U2AF1 mutation . We then identified consensus sequence motifs for the 3′ splice sites lying immediately upstream of the cassette exons that were promoted or repressed in association with U2AF1S34F , represented by “sequence logos” as shown in Figs 1A and S1 . The same analysis was performed on 19 random tumors lacking a U2AF1 mutation as controls . As illustrated by data from patient 7903 in Fig 1A , over two hundred cassette exons were included more frequently and a similar number were included less frequently in this U2AF1-mutant tumor . Notably , as illustrated by the sequence logos , the nucleotide distribution at the -3 position ( boxed ) of promoted and repressed exons was different from that observed upstream of the much larger number of unaffected exons: A replaced T as the second most common nucleotide preceding the promoted exons , while T was more common than C in the sequence preceding the repressed exons . These patterns were observed in nine of the thirteen tumors with the U2AF1S34F allele ( S1A Fig ) . They have been observed previously in comparisons between transcriptomes carrying the U2AF1S34F allele with wild-type transcriptomes [17–20] , and are therefore henceforth referred to as the “typical S34F” consensus 3′ splice sites . In the other four tumors with the U2AF1S34F mutation , these “typical S34F” consensus 3′ splice sites were partially or completely absent ( Figs 1A and S1B ) . Those four tumors exhibited consensus 3′ splice sites that were similar to the consensus 3′ splice sites identified in tumors lacking a U2AF1 mutation , where variations in inclusion relative to the median wild-type sample are presumably stochastic ( S1C Fig ) . Thus , we henceforth refer to these sequence patterns , as shown for tumor from patient 7727 in Fig 1A , as “quasi-WT” . To explain why transcriptomes of some tumors with U2AF1 mutations showed a typical S34F-associated consensus 3′ splice sites , while others exhibited quasi-WT patterns , we estimated the levels of mutant and total U2AF1 mRNA based on available data from the tumors to determine the ratio of mutant to wild-type ( S34F:WT ) mRNA . Tumors with quasi-WT patterns had low S34F:WT mRNA ratios ( ranging from 0 . 27 to 0 . 31 ) , whereas all but one tumor with the S34F-associated pattern had higher ratios ( 0 . 43 or more ) ( Fig 1B ) . In contrast , absolute levels of U2AF1S34F mRNA or total U2AF1 mRNA levels were not different between these two groups of tumors ( S2 Fig , panels A and B ) . We next sought to understand the origin of the wide range of S34F:WT ratios that we observed . These ratios , ranging from 0 . 26 to 0 . 82 , could not be explained by contamination of tumor cells with non-tumor cells , since the proportion of tumor nuclei reported for these samples did not correlate with S34F:WT mRNA ratios ( Fig 1C ) or with the presence or absence of the expected S34F-associated consensus 3′ splice sites ( S2C Fig ) . Conversely , U2AF1 DNA copy number correlated with the estimated levels of total U2AF1 mRNA ( S2D Fig ) . Seven of the 13 U2AF1S34F mutant samples , including two of the four samples displaying quasi-WT consensus 3′ splice site sequences , showed either copy number loss or gain at the U2AF1 locus , suggesting that copy number variation ( CNV ) might account , at least in part , for the varying S34F:WT mRNA ratios in LUAD samples ( Fig 1B ) . Unbalanced allelic expression or proportions of tumor subclones might also contribute to the variable S34F:WT mRNA ratios , although these possibilities could not be readily tested using the available LUAD data . We next tested whether the U2AF1 S34F:WT ratio was associated with quantitative changes in splicing ( versus the qualitative differences in consensus 3′ splice sites described above ) . We correlated S34F:WT ratios with the quantitative inclusion of specific S34F-responsive cassette exons and 5' extensions of exons resulting from competing 3′ splice sites that were reported previously [17 , 19 , 20] . We studied three cassette exons that exhibited less ( ASUN and STRAP ) or more ( ATR ) inclusion , and two 5' extensions of exons ( FMR1 and CASP8 ) that were used less frequently in the presence of U2AF1S34F . Tumors with the highest S34F:WT mRNA ratios showed the lowest inclusion levels of the cassette exon in STRAP mRNA , whereas tumors without a U2AF1 mutation had the highest level of inclusion ( Fig 1D; Pearson correlation of -0 . 79 ) . Similar correlations were observed between inclusion levels of other representative exons and S34F:WT mRNA ratios ( S3 Fig , panels E , I , M and Q ) . As controls , we tested for correlations between the inclusion of these cassette exons and levels of U2AF1S34F mRNA , total U2AF1 mRNA , or percent tumor nuclei . None of these analyses , with the exception of the U2AF1S34F mRNA level versus the inclusion of the 5' extension of the FMR1 exon , showed a relationship as strong as that observed for the S34F:WT mRNA ratio ( S3 Fig panels B-D , F-H , J-L , N-P and R-T ) . These results indicate that the S34F:WT mRNA ratio predicts the magnitude of S34F-associated splicing in human LUAD . The results presented in the preceding section , based on analyses of LUAD tumors with the U2AF1S34F mutation , suggest that the magnitude of S34F-associated splicing is a function of the S34F:WT mRNA ratio . To directly test this hypothesis , we developed a cell line that allows manipulation of WT and mutant U2AF1 gene product levels and measurement of the corresponding effects on RNA splicing . The human bronchial epithelial cell line ( HBEC3kt ) was previously derived from normal human bronchial tissue and immortalized by introduction of expression vectors encoding human telomerase reverse transcriptase ( hTERT ) and cyclin-dependent kinase-4 ( CDK4 ) [22] . To knock in a U2AF1S34F allele at an endogenous locus in HBEC3kt cells , we adopted a published genomic DNA editing approach [23] , using the PiggyBac transposon that leaves no traces of exogenous DNA at the locus ( Fig 2A; see Supplemental Materials and Methods for details ) . We identified three cell clones at intermediate stages of gene editing after screening more than 50 primary transfectants ( S4 Fig , panels A and B ) . Sanger sequencing of these intermediate clones revealed that one of the three clones carried the desired S34F missense sequence , while two clones were wild-type ( S4C Fig ) . Wild-type intermediates are expected because a homologous sequence between the S34F point mutation and the drug cassette in the vector can serve as the 5' homology arm for recombination ( designated as 5' HA#2 in Fig 2A ) . From the final clones derived from these intermediates ( after transposition to remove the drug cassette flanked by the PiggyBac elements ) , we chose one subclone from each of the two wild-type intermediate clones ( referred to as WT1 and WT2 cells ) and two subclones from the sole mutant intermediate clone ( referred to as MUT1a and MUT1b cells ) for all subsequent experiments with isogenic HBEC3kt cells ( S4D Fig ) . The MUT and WT cells all expressed similar levels of U2AF1 mRNA and protein ( S4E Fig ) . Using high-throughput mRNA sequencing ( RNA-seq ) and allele-specific RT-qPCR , we observed similar levels of wild-type and mutant U2AF1 mRNAs in the MUT cells ( Figs 2B and S5C ) , consistent with heterozygosity at the U2AF1 locus . To determine how the engineered U2AF1 S34F allele affects mRNA splicing , we first assayed the inclusion levels of 20 cassette exons that were previously reported to be associated with mutant U2AF1 in both LUAD and AML ( acute myeloid leukemia ) [19] . We confirmed that all 20 of these cassette exons , which included the previously studied ASUN and STRAP cassette exons , were indeed S34F-dependent in our engineered cells using RT-qPCR with isoform-specific primers ( Figs 2C and S6 ) . We next evaluated the global difference in cassette exon recognition in MUT versus WT cells using RNA-seq ( S1 Table ) . MUT and WT cells were grouped separately in an unsupervised cluster analysis based on cassette exon inclusion ( Fig 2D ) . We observed the expected consensus 3′ splice sites of cassette exons that were promoted or repressed in MUT versus WT cells ( Fig 2E ) . Overall , these results indicate that we successfully created clonal HBEC3kt cells isogenic for U2AF1S34F and that the MUT cells exhibited similar alterations in splicing relative to their WT counterparts as do primary LUAD transcriptomes . We next used our isogenic cell model with the U2AF1S34F mutation to test the hypothesis that the S34F:WT ratio , rather than absolute levels of the mutant or wild-type gene products , controls S34F-associated splicing . We tested two specific predictions . First , changing the levels of U2AF1S34F while keeping the S34F:WT ratio constant should not affect S34F-associated splicing . Second , changing the level of wild-type U2AF1 while keeping the level of U2AF1S34F constant ( e . g . , allowing the S34F:WT ratio to change ) should alter the inclusion of S34F-dependent cassette exons . We tested these predictions in the isogenic HBEC3kt cells by manipulating levels of wild-type or mutant U2AF1 gene products and measuring the subsequent changes in S34F-associated splicing . We first reduced the amounts of both mutant and wild-type U2AF1 mRNA concordantly in MUT1a cells , keeping the S34F:WT mRNA ratio constant . This was achieved by transducing MUT1a cells with short hairpin RNAs ( shRNAs ) that target regions of the U2AF1 transcripts distant from the S34F missense mutation . The same shRNAs were also introduced in WT1 cells as a control . Allele-sensitive RT-qPCR confirmed that the mRNA ratio of the mutant and wild-type U2AF1 remained constant ( S7A Fig ) , while the overall U2AF1 mRNA and protein levels were reduced by more than 90% ( Figs 3A , bottom panel , and S7B ) . Knockdown of total U2AF1 in both MUT1a and WT1 cells did not cause significant changes in recognition of the ASUN or STRAP cassette exons , two splicing events that are strongly associated with U2AF1S34F , in either cell line ( Fig 3A , upper panels ) . Similar results were obtained for two additional S34F-associated cassette exons in USP25 and AXL that exhibit increased inclusion in cells expressing U2AF1S34F ( S7 Fig , panels C and D ) . We next confirmed that knockdown of U2AF1 was sufficient to alter splicing events known to be dependent on wild-type U2AF1 . We studied a competing 3′ splice site event in SLC35C2 , in which the use of an intron-proximal over an intron-distal 3′ splice site depends on the level of U2AF1 independent of a U2AF1 mutation [24] . Knockdown of total U2AF1 in either WT1 or MUT1a cell lines reduced the use of the U2AF1-dependent intron-proximal 3′ splice site ( S7E Fig ) . Thus , reduction of U2AF1 to levels sufficient to affect U2AF1-dependent alternative splicing did not affect S34F-associated splicing in MUT1a cells when the S34F:WT ratio was maintained . We next altered the S34F:WT ratio by separately overexpressing mutant or wild-type U2AF1 in WT1 and MUT1a cells and examining the subsequent changes in the recognition of the ASUN and STRAP cassette exons . These cassette exons are preferentially excluded in cells expressing U2AF1S34F . Increasing the amount of U2AF1S34F protein—hence increasing the S34F:WT ratio in either cell type—further enhanced skipping of these cassette exons ( Fig 3B ) . Conversely , decreasing the S34F:WT ratio in MUT1a cells by increasing the production of wild-type U2AF1 protein reduced the extent of exon skipping to approximately the same levels seen in WT1 cells ( Fig 3B ) . We also altered the production of wild-type U2AF1 mRNA and protein in MUT1a cells by disrupting the endogenous wild-type U2AF1 locus with the CRISPR-Cas9 system . Single-guide RNAs ( sgRNAs ) designed to match either the WT or mutant U2AF1 sequences were shown to disrupt either reading frame selectively , generating indels ( insertions and deletions ) at the U2AF1 locus and thereby changing the S34F:WT ratios ( S8 and S9 Figs ) . Since WT U2AF1 is required for the growth of these cells ( as shown below ) , we extracted RNA and protein from cells six days after transduction with Cas9 and sgRNA-WT , when depletion of wild-type U2AF1 was incomplete ( Fig 3C ) . Selective disruption of wild-type U2AF1 increased the S34F:WT mRNA ratio in MUT1a cells; as predicted , the extent of exon skipping was further increased in ASUN and STRAP mRNAs ( Fig 3C ) . Notably , the degree of exon skipping induced by mutant cDNA was similar to that caused by disrupting the wild-type U2AF1 allele ( compare Fig 3C with 3B ) , even though the absolute protein levels of U2AF1S34F were different in the two experiments . These results show that wild-type U2AF1 antagonizes the activity of U2AF1S34F by a competitive mechanism , such that the S34F:WT ratio controls the magnitude of S34F-associated splicing changes independent of levels of either mutant or wild-type protein . Results in the preceding sections are based on studies of a few well-documented S34F-responsive cassette exons that likely serve as surrogates for the global effects of U2AF1S34F on splice site recognition . To determine whether these results reflect general rules governing S34F-associated splicing , we used RNA-seq to measure the consequences of altering S34F:WT ratios on global recognition of cassette exons . When wild-type U2AF1 was diminished by CRISPR-Cas9-mediated disruption in MUT1a cells ( see Fig 3C ) , the S34F-associated changes in inclusion ( Fig 4A ) or exclusion ( Fig 4B ) of cassette exons were enhanced . These global effects are consistent with our measurements of individual cassette exons by RT-qPCR ( Fig 3C ) . An unsupervised cluster analysis suggests that disruption of wild-type U2AF1 in MUT1a cells primarily enhances the magnitude of changes for S34F-associated cassette exons ( Fig 4C ) . We also observed a modest increase in the preference for C versus T at the -3 position of the consensus 3′ splice sites of cassette exons that were promoted versus repressed in association with U2AF1S34F ( Fig 4D ) following reduction of wild-type U2AF1 levels , consistent with the observed association between the S34F:WT mRNA ratio and typical S34F-associated consensus 3′ splice sites identified in LUAD tumor transcriptomes ( Fig 1 ) . We observed similar results when we extended our analyses to include other types of alternative splicing beyond cassette exon recognition ( S10 Fig ) . We used RT-qPCR to validate five splicing alterations that are sensitive to ablation of wild-type U2AF1 in the presence of U2AF1S34F ( S11 Fig ) . Two of the five events involve incorporation of cassette exons ( in ATR and MED15 ) , two involve competition between different 3′ splice sites ( in CASP8 and SRP19 ) , and one involves a choice between two mutually exclusive exons ( in H2AFY ) . We conclude that the importance of the S34F:WT ratio for S34F-dependent splicing changes extends from cassette exon recognition to other types of alternative splicing . Based on the role of U2AF1 in 3′ splice site recognition , we hypothesized that differential RNA binding by mutant and WT U2AF1 could contribute to the observed dependence of S34F-associated splicing on the WT:S34F ratios . It has previously been shown that the S34F mutation reduces the binding affinity of the U2AF1-containing complex for a representative skipped splice site [18] . However , whether changes in RNA binding could account for exon inclusion , as well as the generality of this observation , were unknown . We therefore tested whether altered RNA-binding affinity could explain mutation-dependent increases ( ZFAND1 , FXR1 , ATR , MED15 ) and decrease ( CEP164 ) in exon inclusion , as well as competing 3′ splice site recognition ( FMR1 ) . These splicing events exhibited S34F-associated alterations in both our isogenic cell lines ( S1 Table ) and in LUAD transcriptomes ( S2 Table ) . We determined the RNA binding affinities of purified U2AF1-containing protein complexes using fluorescence anisotropy assays , in which the anisotropy increases of fluorescein-labeled RNA oligonucleotides following protein titration were fit to obtain the apparent equilibrium binding affinities ( S12 Fig; Supplemental Materials and Methods ) . The recombinant proteins comprised either WT or S34F-mutant U2AF1 as ternary complexes with the U2AF2 and SF1 subunits that recognize the adjoining 3′ splice site consensus sequences . The constructs were nearly full length and included the relevant domains for 3′ splice site recognition [15 , 25 , 26] ( Fig 5A ) . The six pairs of tested RNA oligonucleotides ( 33–35 nucleotides in length ) were derived from the proximal or distal 3′ splice sites of the six genes listed above ( Figs 5B and S12 ) . Combined with prior results for sequence variants derived from the S34F-skipped DEK cassette exon [18] , we have in total measured binding affinities for 16 RNA oligonucleotides , which consist of five sequences with a U at the -3 position of the 3′ splice site ( “UAG” splice sites ) , seven “CAG” splice sites , and four “AAG” splice sites . The trends among S34F-altered RNA binding affinities of U2AF1 complexes ( Fig 5C–5H ) for the tested splice site sequences generally agreed with the nucleotide distributions observed in consensus 3′ splice site that are promoted or repressed by U2AF1S34F ( Figs 1A and S1A ) . The S34F mutation reduced the affinities of U2AF1-containing complexes for four out of five tested “UAG” splice sites , consistent with T as the most common nucleotide at the -3 position of the 3′ splice site ( henceforth referred to as -3T ) for S34F-repressed exons . Likewise , the S34F mutation often increased the affinities of U2AF1 complexes for “CAG” splice sites ( for three out of seven tested sequences ) , consistent with -3C as the most common nucleotide preceding S34F-promoted exons . The remaining tested “UAG” or “CAG” splice sites showed no significant difference between S34F and WT protein binding . The “AAG” splice sites lacked a consistent relationship to the S34F-induced RNA affinity changes of U2AF1 complexes per se . However , the S34F mutation increased the binding of the U2AF1-containing complex for one “AAG” splice site for an S34F-promoted exon ( ZFAND1 ) , which is consistent with the greater prevalence of -3A than -3T preceding S34F-promoted exons . Overall , the altered binding affinities of U2AF1-containing complexes for the proximal 3′ splice site could account for four of the six S34F-associated alternative splicing events ( CEP164 , FMR1 , ZFAND1 , FXR1 ) ( Fig 5C–5F , Supplementary S3 Table ) . Similar to the previously-tested S34F-skipped splice site in DEK [18] , the S34F mutation decreased affinities of the U2AF1-containing complexes for the skipped 3′ splice sites of the CEP164 and FMR1 exons ( Fig 5C and 5D ) . Remarkably , the S34F mutation enhanced binding of the U2AF1-containing complexes to the proximal 3′ splice sites of ZFAND1 and FXR1 ( Fig 5E and 5F ) , which could explain the enhanced inclusion of these exons in cell lines and LUAD ( Supplementary S3 Table ) . In agreement with the observed splicing changes , the S34F mutation had no significant effect on the distal 3′ splice sites of these exons . The affinities of the mutant U2AF1 complexes for the proximal splice site oligonucleotides of the remaining two S34F-promoted exons ( ATR and MED15 ) were either similar to wild-type or decreased ( Fig 5G and 5H ) . These results differed from the S34F-dependent increase in U2AF1 binding to the proximal 3′ splice site that was observed for ZFAND1 and FXR1 ( Fig 5E and 5F ) , which could readily explain the enhanced exon inclusions . However , for the ATR pre-mRNA , the U2AF1 mutation reduced binding to the distal more than to the proximal 3′ splice site ( Fig 5G , third and fourth columns ) . Given co-transcriptional splicing in the 5'-to-3′ direction , the downstream ( as opposed to upstream ) 3′ splice sites could compete as splicing acceptors for a given 5' donor splice site when transcription is relatively rapid as compared to splicing . ( Such 3′ splice site competition likely occurs relatively frequently , as the ATR cassette exon is alternatively spliced even in wild-type cells ) . As such , a “net gain” in affinity for the proximal relative to distal 3′ splice site could explain the observed S34F-associated exon inclusion in ATR mRNAs . For the MED15 pre-mRNA , deviation of the S34F-associated splicing changes and a simple RNA affinity model suggested that additional mechanisms control selection of the MED15 3′ splice sites . In total , our measurements of in vitro RNA-binding affinities are sufficient to explain six of seven tested alterations in splice site recognition driven by U2AF1S34F ( Fig 5C–5H and in [18] ) . Other than its effect on RNA splicing , the consequences of the U2AF1S34F mutation on cell behavior are largely unknown . Recurrent mutations , such as U2AF1 S34F , are considered likely to confer a selective advantage to cells in which they occur when expressed at physiologically relevant levels . However , mutant HBEC3kt cells ( MUT1a and MUT1b ) do not exhibit obvious phenotypic properties of neoplastic transformation—such as a growth advantage over wild-type cells ( S13 Fig ) or an ability to grow in an anchorage-independent manner—that are frequently observed in cultured cells expressing well-known oncogenes , like mutant RAS genes . Another attribute of some well-known oncogenes , such as BCR-ABL fusion in chronic myeloid leukemia or mutant EGFR or KRAS in LUAD , is the dependence on sustained expression of those oncogenes for the maintenance of cell growth or viability . To determine whether LUAD cells harboring a pre-existing S34F mutation are dependent upon ( or “addicted to” ) the mutant allele , we searched the COSMIC database for LUAD cell lines with the U2AF1S34F allele [27] . Two cell lines ( H441 and HCC78 ) were found and both these cells exhibit copy number gains at the U2AF1 locus ( three copies for H441 cells; four copies for HCC78 cells ) and one copy of a variant allele . We confirmed that U2AF1S34F was the minor allele in these cells by Sanger sequencing and allele-specific RT-qPCR ( S14 Fig , panels A and B ) . We further used the CRISPR-Cas9 system to selectively disrupt the wild-type or mutant U2AF1 sequences and then assessed the impact of inactivating the U2AF1 alleles on the clonogenic growth of the two LUAD lines with the U2AF1 mutation . In addition , we performed similar experiments with the LUAD cell line A549 ( wild-type for U2AF1 ) and the HBEC3kt-derived MUT1a cell line . In all instances , loss of the mutant allele did not impair cell growth . Only one line ( H441 ) exhibited altered growth , in the form of a two-fold increase in clonogenicity ( Fig 6A ) . Successful disruption of the U2AF1S3F allele was confirmed by restoration of a normal RNA splicing profile in subclonal cells derived from the clonogenic assay colonies ( see S14–S17 Figs , S4 and S5 Tables , and text below ) . In contrast , loss of the wild-type allele completely inhibited clonogenic growth in all tested cell lines , regardless of whether the line carried the U2AF1S34F allele or not . A rescue experiment confirmed that loss of cell growth was due to loss of wild-type U2AF1 expression . The loss of clonogenic capacity after disrupting endogenous U2AF1 in A549 cells was prevented by first transducing them with a form of wild-type U2AF1 cDNA ( Fig 6B ) that is not predicted to be a target for sgRNA-WT ( S8 Fig ) . Overall , these findings indicate that wild-type U2AF1 is required for the clonogenic growth of cells , including lung cancer cell lines , that the S34F mutant is unable to compensate for loss of the wild-type allele , and that LUAD cells with the S34F mutation are not dependent on the mutant allele for growth in vitro . To examine the effect of U2AF1S34F on tumor growth in vivo , we derived H441 and HCC78 cells transduced with Cas9 and sgRNA-S34F or sgRNA-GFP as polyclonal pools or as clones ( S14 and S15 Figs ) . The cell lines were verified to either carry or not carry the U2AF1S34F allele , and we confirmed that the Cas9 and sgRNA-S34F-transduced cells lost the S34F-associated splicing program ( S16 and S17 Figs ) . We inoculated these subclonal cell lines subcutaneously in nude mice and monitored xenograft tumor growth . The H441-derived cell lines , in which the U2AF1S34F allele was disrupted , were able to establish tumors in vivo at rates similar to those observed for tumor cells carrying the mutant allele ( S6 Table ) . The HCC78-derived cell lines did not grow palpable tumors after xenografting within the observation period , so the requirement for U2AF1S34F in vivo could not be tested in that line . These experiments show that U2AF1S34F is dispensable for growth of these LUAD cell lines in vivo , a result consistent with the clonogenicity assays shown in Fig 6 . We conclude that U2AF1S34F appears to be neither sufficient nor necessary for lung cell transformation in these assays . In contrast , wild-type U2AF1 is required for cell viability , consistent with the retention of a wild-type allele in human cancers carrying common U2AF1 mutations .
U2AF1S34F is known to induce specific splicing alterations , but it is not known how these changes are regulated . We show that the ratio of S34F:WT U2AF1 gene products is a critical determinant of the magnitude of S34F-associated splicing . This conclusion was demonstrated in an isogenic lung epithelial cell line engineered to express U2AF1S34F from one of the two endogenous U2AF1 loci , and was further supported by analyses of human LUAD transcriptomes carrying the U2AF1S34F allele . These results suggest that wild-type U2AF1 antagonizes the splicing program associated with the S34F mutation . We find that a major functional difference between purified S34F mutant and wild-type U2AF1 proteins resides in altered binding affinities for a subset of 3′ splice sites . The trends in the RNA sequence preferences of S34F-mutant U2AF1 are consistent with the preferred 3′ splice sites of S34F-affected transcripts ( here and in [17–20 , 28] ) , which we use as the signature of S34F-associated differential splicing ( Fig 1A ) . For oligonucleotides that showed significant changes in binding affinities , the S34F mutation typically reduced or enhanced respective binding of the U2AF1 splicing factor complexes to 3′ splice sites preceded by a -3U or -3C ( Fig 5 and [18] ) . In support of our findings for the relevant ternary complex of human U2AF1 , U2AF2 and SF1 subunits , recent studies confirmed that the corresponding S34F mutation inhibited binding of the minimal fission yeast U2AF heterodimer to a “UAG” splice site RNA [21] . These S34F-altered RNA affinities are consistent with the location of the substituted amino acid in a zinc knuckle that may directly contact RNA [17] . Although the effects of the S34F mutation on binding 3′ splice sites preceded by -3A are variable , extrusion or alternative U2AF1 binding sites for disfavored nucleotides could occur in different sequence contexts by analogy with other RNA binding proteins [29 , 30] . Several lines of evidence support the idea that U2AF1S34F is capable of initiating pre-mRNA splicing once it binds to RNA . Nuclear extracts of cells overexpressing mutant U2AF1 can support in vitro splicing reactions more efficiently than nuclear extracts derived from cells overexpressing wild-type U2AF1 for a minigene with a specific 3′ splice site sequence [17] . In addition , mutant U2AF1 can compensate for loss of the wild-type factor for the inclusion of some U2AF1-dependent cassette exons [31] . Our observations that the altered RNA-binding affinities correlate well with S34F-associated splicing for the majority of splice sites that we tested further suggest that mutant and wild-type U2AF1 are functionally equivalent for downstream steps of splicing ( Fig 5 and [18] ) . In light of our findings and existing evidence from the literature , we propose a model wherein mutant U2AF1 drives differential splicing by favoring the recognition of one of two competing 3′ splice sites ( S18 Fig ) . This model is motivated by three key facts . First , alternative splicing , in contrast to constitutive splicing , necessarily results from implicit or explicit competition between splice sites . ( For example , cassette exon recognition can involve competition between the 3′ splice sites of the cassette exon itself and a downstream constitutive exon . ) Second , mutant and wild-type U2AF1 complexes have different binding specificities , largely due to their preferences for distinct nucleotides at the -3 position , that lead them to preferentially bind distinct 3′ splice sites . Third , mutant and wild-type U2AF1 are likely functionally equivalent once they bind to RNA . Therefore , altering the cellular ratio of mutant and wild-type U2AF1 changes the amount of total U2AF1 bound to a given 3′ splice site in a sequence-specific manner , thereby promoting or repressing that splice site relative to a competing 3′ splice site . Our proposed model is not exclusive of other possible contributing effects , such as competitive binding of mutant and wild-type U2AF1 to a factor with low stoichiometry ( S18 Fig ) , or effects of U2AF1S34F on the kinetics of co-transcriptional splicing as suggested recently [32] . Future studies are needed to resolve these possibilities . U2AF1S34F is recurrently found in LUAD , other solid tumors , and myeloid disorders , suggesting that the mutant allele confers a physiological property that provides a selective advantage during neoplasia . A gene ontology ( GO ) analysis for genes that show S34F-associated alterative splicing in HBEC3kt-derived isogenic cells shows significant alterations in many biological processes such as mRNA processing , RNA splicing , G2/M transition of mitotic cell cycle , double-strand break repair and organelle assembly ( FDR-adjusted p-values < 0 . 001 ) . However , we did not observe signs of neoplastic transformation or changes in cell proliferation after introducing U2AF1S34F into the endogenous U2AF1 locus in HBEC3kt cells ( S13 Fig ) . Moreover , targeted inactivation of U2AF1S34F in LUAD cell lines does not diminish , and in one case even increases , clonogenic growth in culture ( Fig 6 ) . The lack of a testable cellular phenotype has been a major hindrance to understanding the functional significance of mutant U2AF1 in carcinogenesis . Cell proliferation is only one of the many hallmarks of cancer , so careful examination of other cell properties in the isogenic cells may be needed to establish the presumptive role of U2AF1S34F in carcinogenesis . More recently , after completion of our study [33] , Park et al reported that tumorigenic cells emerge after U2AF1S34F is ectopically produced in Ba/F3 pro-B cells or in an immortalized line of small airway epithelial cells [34] . The authors attributed the transformation events by mutant U2AF1 to the consequences of altered 3′ processing of mRNA’s . In particular , they observed an increase in the length of the 3′ untranslated region of ATG7 mRNA and a reduction in the amount of ATG7 protein , proposing that the anticipated defect in autophagy predisposes cells to mutations , some of which are transforming . This “hit-and-run” type of mechanism is consistent with our observation and theirs that mutant U2AF1 is dispensable for maintenance of the transformed phenotype in LUAD cell lines ( Fig 6 ) and in their cell lines [34] . Their observations do not , moreover , exclude a role for S34F-associated splicing in the oncogenic mechanism . We have shown that the wild-type U2AF1 allele is absolutely required for the growth of lung epithelial and LUAD cells that carry a mutant U2AF1 allele ( Fig 6 ) . This result indicates that mutant U2AF1 cannot complement a deficiency of wild-type U2AF1; it may also explain why tumors homozygous for the U2AF1S34F mutation have not been observed , although the number of tumors found to have even a heterozygous mutation is still relatively small , so the analysis may not be adequately powered . Still , the frequent occurrence of a low ratio of S34F:WT U2AF1 mRNA , accompanied by increased copy number of wild-type U2AF1 alleles in lung adenocarcinoma cell lines and possibly LUADs , suggests that there may be selection for a lower ratio of S34F:WT in addition to the likely selection , perhaps at an earlier stage of tumorigenesis , for the S34F mutant . These results are consistent with a study of mutations affecting the splicing factor SF3B1 , which reported that cancer cells harboring recurrent SF3B1 mutations also depend on wild-type SF3B1 for growth [35] . Finally , a recent study similarly found that a wild-type copy of SRSF2 is required for leukemic cell survival , and that SRSF2 mutations generated a therapeutic index for treatment with a small molecule that inhibits 3′ splice site recognition [36] . Together , our results and these recent studies provide a genetic rationale for targeting wild-type splicing factors ( or the splicing machinery more generally ) in cancers harboring spliceosomal mutations .
The ethics statement is available in S1 Text . The HBEC3kt cells ( a gift from Dr . John Minna , UT Southwestern Medical Center ) , H441 , A549 ( ATCC ) and HCC78 cells ( DMSZ ) were cultured as previously described [22] or according to vendors’ instructions . The primary antibodies for immunoblots are: rabbit anti U2AF1 ( 1:5000 , # NBP1-19121 , Novus ) , rabbit anti GFP ( 1:5000 , #A-11122 , Invitrogen ) , mouse anti ACTB ( 1:5000 , Clone 8H10D10 , Cell Signaling ) . Lentiviruses were produced and titered in HEK293T cells as previously described [37] . An MOI ( multiplicity of infection ) of 1–5 were used for all assays . Total RNA was extracted and reverse transcribed as previously described [38] . Splicing alterations were measured by quantitative PCR on reverse-transcribed cDNA ( RT-qPCR ) , using isoform-specific primers ( S7 Table ) . These primers were designed following a previously described method [39] . The PCR efficiency and specificity of each primer set were determined before they were used for measuring splicing changes ( See Supplemental Materials and Methods for details ) . Clonogenic assay was performed by infecting cells with lentiviruses two days before seeding them into 100 mm dishes ( 1000 live cells per dish ) to grow colonies . Growth media were supplemented with puromycin ( 1 μlg/ml ) for selecting infected cells and were changed once a week for up to three weeks . Cell colonies were stained with 0 . 03% methylene blue ( in 20% methanol ) for 5 min . Clonogenicity is defined as colony numbers formed as a percent of those in control cells . Details of all DNA constructs used in the study and the genome editing approaches for creating the U2AF1S34F allele and allelic-specific disruption of U2AF1 are described in the Supplemental Materials and Methods . High throughput mRNA sequencing ( RNA-seq ) was conducted in the Sequencing Facility of the National Cancer Institute . RNA quality was assessed by 2100 Bioanalyzer ( Agilent ) . Total RNA with good integrity values ( RIN > 9 . 0 ) was used for poly A selection and library preparation using the Illumina TruSeq RNA library prep kit . Two or four samples were pooled per lane and ran on the HiSeq2000 or HiSeq2500 instrument using TruSeq V3 chemistry . All samples were sequenced to the depth of 100 million pair-end 101 bp reads per sample . Splicing analysis of RNA-seq data from the TCGA LUAD cohort as well as engineered HBEC3kt , H441 , and HCC78 cell lines was performed as previously described [17] . A brief description of the method was provided in the Supplemental Materials and Methods . Purification of U2AF1 complexes , as illustrated in Fig 5 , was explained in Supplemental Materials and Methods . Sequences of synthetic 5′-labeled fluorescein RNAs ( GE Healthcare Dharmacon ) and binding curves are given in the Supplementary S12 Fig . Apparent equilibrium affinity constant of the purified U2AF1 complexes with RNA was measured based on changes in fluorescence anisotropy as previously described [18] . A brief description of this method was provided in the Supplemental Materials and Methods . All experiments were independently performed at least three times unless otherwise stated . Statistical significance was determined by two-tailed Student's t test or otherwise stated . In all analyses , p values ≤ 0 . 05 are considered statistically significant .
|
Large-scale genomics studies have identified recurrent mutations in many genes that fall outside the conventional domain of proto-oncogenes . They include genes encoding factors that mediate RNA splicing; mutations affecting four of these genes are present in up to half of proliferative myeloid disorders and in a significant number of solid tumors , including lung adenocarcinoma . Here we have characterized several properties of a common mutant version of the U2AF1 splicing factor , a component of the U2 auxiliary factor complex , in lung cells . We have found that mutant-associated changes in splice site selection are primarily influenced by the ratio of mutant and wild-type U2AF1 gene products; thus increasing wild-type U2AF1 levels represses the mutant-induced splicing program . We show that the altered splice site preferences of mutant U2AF1 can be attributed to changes in its binding to relevant 3′ splice sites . We also show that mutant U2AF1 is different from some oncogenes: the growth properties of lung cancer cell lines carrying the mutant allele are unaffected by loss of the mutant gene , while the wild-type allele is absolutely required for survival . These results advance our understanding of the molecular determinants of the mutant-associated splicing program , and they highlight previously unappreciated roles of wild-type U2AF1 in the presence of the recurrent U2AF1S34F mutation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"sequencing",
"techniques",
"adenocarcinoma",
"of",
"the",
"lung",
"medicine",
"and",
"health",
"sciences",
"messenger",
"rna",
"carcinomas",
"cancers",
"and",
"neoplasms",
"cloning",
"alternative",
"splicing",
"oncology",
"mutation",
"genome",
"analysis",
"adenocarcinomas",
"sequence",
"motif",
"analysis",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"genomics",
"lung",
"and",
"intrathoracic",
"tumors",
"gene",
"expression",
"rna",
"splicing",
"molecular",
"biology",
"biochemistry",
"rna",
"rna",
"processing",
"nucleic",
"acids",
"genetics",
"transcriptome",
"analysis",
"biology",
"and",
"life",
"sciences",
"computational",
"biology"
] |
2016
|
Wild-Type U2AF1 Antagonizes the Splicing Program Characteristic of U2AF1-Mutant Tumors and Is Required for Cell Survival
|
Kaposi's sarcoma ( KS ) is the most common cancer among HIV-positive patients . Histogenetic origin of KS has long been elusive due to a mixed expression of both blood and lymphatic endothelial markers in KS tumor cells . However , we and others discovered that Kaposi's sarcoma herpes virus ( KSHV ) induces lymphatic reprogramming of blood vascular endothelial cells by upregulating PROX1 , which functions as the master regulator for lymphatic endothelial differentiation . Here , we demonstrate that the KSHV latent gene kaposin-B enhances the PROX1 mRNA stability and plays an important role in KSHV-mediated PROX1 upregulation . We found that PROX1 mRNA contains a canonical AU-rich element ( ARE ) in its 3′-untranslated region that promotes PROX1 mRNA turnover and that kaposin-B stimulates cytoplasmic accumulation of the ARE-binding protein HuR through activation of the p38/MK2 pathway . Moreover , HuR binds to and stabilizes PROX1 mRNA through its ARE and is necessary for KSHV-mediated PROX1 mRNA stabilization . Together , our study demonstrates that kaposin-B plays a key role in PROX1 upregulation during lymphatic reprogramming of blood vascular endothelial cells by KSHV .
Kaposi's sarcoma ( KS ) is causally associated with human herpes virus ( HHV ) -8 , also called KS-associated herpes virus ( KSHV ) [1] . KSHV develops various-sized KS tumors that are structurally accompanied by aberrant angiogenesis of slit-like vessels frequently containing red blood cells and inflammatory cells [2] , [3] . KS tumor cells characteristically appear spindle-shaped and are believed to be derived from endothelial cells . KS tumor cells were initially proposed to originate from blood vascular endothelial cell ( BEC ) because of their expression of BEC-specific antigens [4] , [5] , [6] , [7] , [8] , [9] , [10] . Later , however , KS tumor cells were also found to express lymphatic endothelial cell ( LEC ) -specific markers such as VEGF receptor-3 ( VEGFR-3/flt4 ) and podoplanin [11] , [12] , [13] , [14] , [15] , [16] , [17] , [18] , [19] , arguing for their lymphatic origin . Recently , we and others have demonstrated that KSHV reprograms the transcriptional profile of BECs to resemble LECs by upregulation of PROX1 , the master regulator for the LEC-differentiation [20] , [21] , [22] , [23] . PROX1 , the mammalian homolog of the Drosophila neuronal cell fate regulator Prospero , is a homeodomain transcription factor essential for development of a variety of organs , including the lymphatic system [24] , [25] , the liver [26] , the lens [27] , [28] , the brain [29] , [30] , [31] , [32] , the ear [33] , [34] , [35] , [36] and the heart [37] , [38] . During early lymphatic development , endothelial cells in the cardinal vein exhibit a mixed phenotype of both BECs and LECs . A subset of venous endothelial cells begins to express PROX1 and migrates out to form the initial lymphatic vessels [24] , [25] . This lymphatic differentiation process is found to be arrested in PROX1 knockout mice , which fail to develop the lymphatic system [24] , [25] . We and others found that ectopic expression of PROX1 induces lymphatic reprogramming of post-developmental BECs [39] , [40] . Therefore , PROX1 is thought to override the BEC phenotype by repressing BEC-specific markers and to induce lymphatic phenotypes by upregulating LEC-specific genes , functioning as the master control regulator for LEC differentiation . Controlling mRNA stability is an important post-transcriptional regulatory process , which allows a rapid adjustment of the copy number of mRNAs by involving a sequence element called AU-rich element ( ARE ) [41] , [42] , [43] , [44] . AREs are usually 50–150 nucleotide long and locate in the 3′-untranslated region ( UTR ) of mRNAs with a short half life , serving as an mRNA-destabilizing determinant by promoting degradation of mRNAs . Notably , ARE-containing mRNAs are found to represent as much as ∼8% of total mRNAs encoded in human cells and are involved in many essential biological processes such as signal transduction , cell growth and differentiation , immune responses , hematopoiesis and apoptosis [43] , [44] . AREs are grouped into three classes based on the number and distribution of the core AUUUA pentamers [43] , [45] , [46] . Class I ARE genes contain several dispersed copies of the AUUUA motif within the AU-rich region and include c-myc , c-fos , cyclins A , B1 and D1 and interferon-γ . Class II ARE genes have at least 2 overlapping UUAUUUA ( U/A ) ( U/A ) motifs and include tumor necrosis factor ( TNF ) -α , interleukin ( IL ) -1β , IL-2 , IL-3 , granulocyte/macrophage colony-stimulating factor ( GM-CSF ) , Cox-2 and VEGF . Finally , less characterized class III AREs do not contain the canonical AUUUA motif and are found in genes such as c-jun , GLUT , p53 and hsp70 . Interestingly , while many cytokine-encoding mRNAs harbor the class II AREs , mRNAs encoding cell cycle regulators and transcription factors contain the class I and occasionally class III AREs [43] . Several ARE-binding proteins have been reported to either destabilize or stabilize ARE-containing mRNAs [43] , [47] . Notably , HuR , embryonic lethal abnormal vision ( ELAV ) -like RNA-binding protein , is one of the best characterized ARE-binding proteins and stabilizes labile ARE-containing mRNAs such as c-fos , MyoD , p21 , cyclins A , B1 and D1 , TNF-α , GM-CSF and VEGF [42] , [43] , [48] , [49] , [50] , [51] , [52] , [53] , [54] . Predominantly present in the nuclei , HuR shuttles between the nucleus and cytoplasm in response to various internal and external stimuli , and its mRNA-stabilizing function has been attributed to its cytoplasmic localization [46] , [50] , [52] , [55] , [56] . Importantly , the KSHV latent gene kaposin-B has been shown to activate the p38/MK2 pathway and to stabilize various cytokine mRNA containing AREs [57] , [58] . Kaposin-B can directly bind to MK2 and promote its kinase activity through its DR2 repeats and , in response to lipopolysaccharide ( LPS ) , kaposin-B and MK2 were shown to be exported to cytoplasm [57] , [58] . Kaposin-B and MK2/p38 proteins have been shown to enhance the stability of ARE-containing mRNAs such as GM-CSF and IL-6 , leading to an enhanced production of cytokines and signaling proteins [57] , [58] . However , the molecular mechanism underlying the kaposin-B/MK2-mediated stabilization of the ARE-containing mRNA remains to be better defined . While study of endothelial cell fate reprogramming by KSHV has provided important insights into KS oncogenesis , the molecular mechanism underlying KSHV-mediated PROX1-upregulation has only begun to be elucidated . An interesting recent report has shown that Akt activation through gp130 receptor may play an important role in KSHV-induced lymphatic reprogramming [59] . Here , we found that PROX1 harbors an unusually long 3′-UTR that contains the canonical ARE , which functions as a PROX1 mRNA-destabilizing determinant . Moreover , we discovered that HuR protein physically binds and stabilizes PROX1 mRNA and that cytoplasmic localization of HuR protein is activated by kaposin-B . Together , our data demonstrate that kaposin-B plays a key role in KSHV-mediated PROX1 upregulation .
We and others have previously demonstrated that KSHV induces lymphatic reprogramming of vascular endothelial cells by upregulating PROX1 and that this PROX1-upregulation occurs in KSHV-infected cells in vitro [20] , [21] , [22] , [23] . In this study , we further investigated the correlation between KSHV infection and PROX1 upregulation both in vitro and in vivo . For the in vitro study , we infected cultured human dermal BECs with KSHV for 7 days and performed immunofluorescent studies for PROX1 and LANA/ORF73 , a KSHV viral protein that marks a latent KSHV-infection . We found that PROX1 was upregulated predominantly in LANA-positive , KSHV-infected BECs , but not in LANA-negative , uninfected neighboring BECs ( Figure 1A-D ) . Our study revealed that ∼78% of the cells ( n = 390 ) was double negative for PROX1 and LANA , and ∼18% double positive ( Figure 1E ) , strongly correlating PROX1 upregulation with de novo KSHV-infection . We next stained KS biopsy sections with anti-PROX1 and LANA antibodies to analyze co-expression of PROX1 and LANA in KS tumor cells ( Figure 1F-I ) . We found that the majority of cells in KS tumors were infected with KSHV based on the characteristic LANA-speckles in KSHV-infected nuclei in vivo ( Figure 1G ) and that most of the LANA-positive , KSHV-infected cells upregulated PROX1 . Out of 160 cells , ∼81% of the cells was double positive for PROX1 and LANA , ∼12% double negative , and only 3∼4% cells single-positive ( Figure 1J ) , a finding consistent with the in vitro data . Together , our studies demonstrate that PROX1 upregulation occurs only in KSHV-infected cells . We next investigated how KSHV induces PROX1-upregulation in endothelial cells . Since KSHV upregulates PROX1 in the latent stage , when only a few viral genes are known to be expressed , we hypothesized that one or more KSHV latent genes may be responsible for the activation of PROX1 expression and thus tested their ability to upregulate PROX1 in various endothelial cell backgrounds such as LECs , BECs and human umbilical venous endothelial cells ( HUVECs ) . Real time RT-PCR ( qRT-PCR ) analyses revealed that ectopic expression of the KSHV latent gene kaposin-B in LECs significantly upregulated PROX1 ( 6∼7-fold ) ( Figure 2A ) . In comparison , kaposin-B did not notably induce PROX1 expression in either BECs or HUVECs ( Figure 2A ) . From the same set of experiment , we investigated the effect of kaposin-B in the regulation of other lymphatic genes ( podoplanin , VEGFR-3 , LYVE-1 , FGFR-3 , SLC and p57 ) in LECs , BECs and HUVECs ( Figure S1 ) . Interestingly , we found that although kaposin-B alone did not seem to induce the lymphatic reprogramming as extensively as KSHV [20] , [21] , [22] , [23] , kaposin-B alone was able to partially modulate the expression of other LEC-signature genes in KSHV-infected endothelial cells . We next overexpressed kaposin-B in LECs and performed the semi-quantitative RT-PCR analyses against PROX1 and IL-6 , a known kaposin-B target gene [57] . Like IL-6 , PROX1 was also significantly upregulated by kaposin-B in LECs ( Figure 2B ) . In a separate experiment , a Flag-tagged kaposin-B was transfected into LECs and the steady-state level of PROX1 protein was determined by western blot analyses ( Figure 2C ) . Together , these data demonstrate that the KSHV viral gene kaposin-B can upregulate PROX1 expression in LECs , but not in BECs and HUVECs where the lymphatic-specific PROX1 is not expressed . We then set out to investigate the molecular mechanism underlying kaposin-B-induced PROX1 upregulation . Kaposin-B has been demonstrated to upregulate various cytokine genes by stabilizing their mRNAs through AREs located in their 3′-UTRs [57] . We thus examined the mRNA structure of the PROX1 gene . Although the open reading frame ( ORF ) of human or mouse PROX1 gene is about 2 . 2 kb long and encodes a 737-amino acid-long protein , reported northern blot analyses revealed that PROX1 transcript was as large as 8-kb in various tissues [60] , [61] , [62] , suggesting that the PROX1 transcript has a long UTR at the 5′- and/or 3′ ends . In fact , we found that an 8-kb PROX1 transcript harboring an extended 3′-UTR has been annotated in a public genome database ( ENST00000366958 ) . However , the corresponding Prox1 transcript from mouse has not been annotated in the same public database . To further confirm the presence of PROX1 transcript with an extended 3′-UTR in both human and mouse , we aligned 7-kb downstream genomic sequences of human or mouse PROX1 against human or mouse expressed sequence tag ( EST ) databases and found that numerous EST sequences were mapped to the downstream of both human PROX1 and mouse Prox1 genes ( Figure 3A&B ) , indicating that this region is indeed transcribed as a part of the fifth exon of PROX1 mRNA in both species . We then investigated sequence conservation of this 3′-UTR among Prox1 genes of other species . Analyses using a genome browser revealed a high DNA sequence homology in this extended 3′-UTR , especially in the second half , of the Prox1 genes from primates , placental mammals or vertebrates covering 48-speices [63] ( Figure 3C ) . Since PROX1 was expressed in multiple organs such as the brain , liver , muscle and heart [60] , [61] , [62] and most of EST fragments were derived from cDNA libraries of these organs , we asked whether LECs express PROX1 mRNA with a long 3′-UTR and thus performed northern blot analyses by using three different 3′-UTR probes and a PROX1 open reading frame ( ORF ) probe against RNAs isolated from human LECs . Indeed , both the 3′-UTR and the ORF probes detected a single ∼8-kb band ( Figure 3D ) , indicating that LECs express a ∼8-kb long PROX1 mRNA . In addition , we performed the 3′-rapid amplification of cDNA end ( RACE ) analyses and found that a majority of human LEC-PROX1 mRNA terminates at 5 , 414-bp downstream from the termination codon of PROX1 ( Figure 3E ) . We identified a classical poly-A signal sequence ( AATAAA ) at ∼20-bp upstream of the termination site . Together , our data demonstrate that PROX1 mRNA expressed in primary human LECs harbors a 5 . 4-kb-long 3′-UTR and that this unusually long 3′-UTR is conserved among the Prox1 genes of many vertebrate species . Because kaposin-B stabilizes cytokine mRNAs through AREs in their 3′-UTR [57] , we asked whether PROX1 mRNA contains a canonical ARE in its unusually long 3′-UTR . We performed bioinformatic analyses against the conserved vertebrate Prox1 3′-UTRs and found an AU-rich region approximately 400-bp upstream of the mRNA termination site ( Figure 3F ) . Notably , this region contains three copies of the AUUUA core pentamer ( its location marked in Figure 3A ) that can be classified as a class I ARE [43] , [44] , [45] , [46] . Moreover , these three tandem copies of core pentamer were also found to be conserved in Prox1 mRNA of dog , mouse , human , chicken and zebrafish ( Figure 3F ) . We next investigated whether the newly discovered PROX1-ARE can serve as an mRNA instability determinant by utilizing the classical β-globin mRNA stability reporter system ( pTet-BBB ) [64] , [65] . In this system , the tetracycline ( Tet ) -controlled promoter directs the inducible expression of the rabbit β-globin gene and a specific DNA sequence element such as a putative ARE can be cloned into the 3′-UTR of the β-globin gene in order to evaluate its ability to destabilize otherwise stable β-globin mRNA [64] , [65] . We cloned a 40-bp fragment containing the PROX1 AUUUA core pentamer in pTet-BBB ( Figure 3G ) and evaluated its effect on the stability of β-globin mRNA in NIH3T3/Tet-Off cells by northern blotting analyses . Indeed , the three AUUUA motifs from the PROX1 3′-UTR significantly destabilized the stable β-globin mRNA as potently as the c-fos ARE that was used as a positive control ( Figure 3H ) , demonstrating that the 40-bp PROX1-ARE is sufficient to function as an instability determinant for PROX1 mRNA . Quantification of the northern blot bands revealed that while the unmodified β-globin mRNA shows a long half-life ( >210-minutes ) , the PROX1-ARE , like the c-fos-ARE , shortened the half-life of β-globin mRNA to ∼60-minutes ( Figure 3I ) . Together , our data demonstrate that PROX1 mRNA contains an ARE functioning as an mRNA instability determinant in its unusually long 3′-UTR and that this newly identified PROX1-ARE may play an important role in the post-transcriptional regulation of PROX1 expression . We next investigated the molecular mechanism underlying the regulation of PROX1 mRNA stability through its ARE and searched for ARE-binding factors that may interact with the PROX1-ARE . The nuclear protein HuR , also known as ELAVL1 , has been shown to be one of the best characterized ARE-binding proteins that bind various cytokine/chemokine mRNAs to increase their stability [66] . Therefore , we evaluated the possibility of HuR binding to the PROX1-ARE and promoting the mRNA stability . Toward this aim , a HuR-expressing vector was transfected into primary LECs and PROX1 expression was determined . Indeed , our semi-quantitative RT-PCR analyses showed that HuR overexpression resulted in upregulation of PROX1 as well as a known HuR-target gene VEGF [49] , [67] ( Figure 4A ) . This HuR-mediated PROX1 upregulation was also confirmed by using quantitative qRT-PCR ( Figure 4B ) and western blot ( Figure 4C ) analyses . We next asked whether HuR protein can physically interact with the AU-rich region of the PROX1 3′-UTR and performed co-immunoprecipitation ( co-IP ) for a protein-RNA complex of endogenous HuR protein and PROX1 mRNA from primary LECs by using an anti-HuR antibody as previously described [68] . Precipitated protein-RNA complex was de-crosslinked , reverse-transcribed and PCR-amplified by using two neighboring sets of PROX1 PCR primers; Probe-1 and Probe-2 . While Probe-1 primer pair detects the PROX1-ARE region , Probe-2 primer pair , 42-bp away from Probe-1 , binds at 25-bp downstream from the end of PROX1 mRNA ( Figure 4D ) . Importantly , whereas RT-PCR using Probe-1 amplified a corresponding product , RT-PCR using Probe-2 did not yield any product , indicating that the endogenous HuR protein is physically associated with the PROX1-ARE region and also that the Probe-1 product was not due to possible genomic DNA contamination in our co-immunoprecipitation assays ( Figure 4D ) . As controls , we could detect the association of HuR protein with VEGF and UBE2N mRNAs [49] , [68] , but not with GAPDH mRNA ( Figure 4D ) . We next performed a RNA electrophoresis mobility-shift assay ( RNA EMSA ) to corroborate the molecular interaction between HuR protein and PROX1-ARE mRNA . We in vitro transcribed a RNA EMSA probe spanning the PROX1-ARE region and then investigated if this RNA probe can make a stable RNA-protein complex with recombinant HuR protein and/or another known ARE-binding protein tristetraprolin ( TTP ) [47] . Indeed , while GST alone or GST-TTP protein did not make any detectable RNA-protein complex , GST-HuR recombinant protein formed a stable complex with the PROX1-ARE RNA probe ( Figure 4 E , lanes 1–4 ) . It is interesting to find that HuR , but not TTP , binds to PROX1-ARE RNA probe , although both are known to have an affinity to the ARE motif [69] . Moreover , the complex formation between PROX1-ARE RNA probe and GST-HuR protein could be inhibited by an unlabeled specific competitor ( PROX1-ARE RNA probe ) , but not by an unlabeled non-specific competitor ( yeast total RNA ) , indicating a specific molecular interaction between HuR protein and PROX1-ARE region ( Figure 4E , lanes 5∼14 ) . Taken together , our data demonstrate that HuR protein physically interacts with PROX1 mRNA through the AU-rich region . Our findings of kaposin-B-induced PROX1 upregulation and HuR-binding to PROX1-ARE directed us to ask whether HuR and/or kaposin-B upregulate PROX1 by enhancing PROX1 mRNA stability . Toward this question , we overexpressed HuR or kaposin-B in primary LECs and quantified the steady-state level of PROX1 mRNA by qRT-PCR . Indeed , the ectopic expression of HuR or kaposin-B delayed the turnover of PROX1 mRNA in LECs and increased the half-life of PROX1 mRNA from ∼60 minutes in the control LECs to ∼180 minutes in LECs overexpressing HuR or kaposin-B ( Figure 5A ) . We then asked whether HuR is required for kaposin-B-mediated PROX1 upregulation by knockdown of HuR in kaposin-B-expressing LECs . We found that HuR-knockdown significantly inhibited kaposin-B-mediated upregulation of PROX1 mRNA and protein determined by qRT-PCR and western analyses , respectively ( Figure 5B&C ) . Moreover , we confirmed that this reduction in kaposin-B-mediated PROX1 upregulation is due to a decrease in PROX1 mRNA stability upon knockdown of HuR ( Figure 5D ) . Together , our data demonstrate that kaposin-B upregulates PROX1 by promoting its mRNA stability through HuR . While HuR protein mainly resides in the nucleus , various cell stress signals activate cytoplasmic accumulation of HuR [52] . We next asked if kaposin-B activates localization of HuR protein to the cytoplasm where mRNA stability is regulated . Indeed , our immunofluorescent analyses revealed that ectopic upregulation of kaposin-B stimulated cytoplasmic mobilization of HuR protein ( Figure 6A ) . Moreover , we harvested the cytoplasmic and nuclear fractions from control vs . kaposin-B-overexpressing LECs to quantify the amount of mobilized HuR by kaposin-B . Consistent with the immunostaining data , a significant amount of HuR protein was found to be exported to the cytoplasm ( Figure 6B ) . Therefore , our data demonstrate that cytoplasmic accumulation of HuR protein is activated by kaposin-B , which may play an important role in PROX1 upregulation . Kaposin-B has been shown to activate the p38/MK2 pathway and stabilize various cytokine mRNAs [57] , [58] . We further examined this previous observation in LECs and found that the ectopic expression of kaposin-B activated phosphorylation of p38 and MK2 proteins ( Figure 7A ) . We then investigated whether activation of the p38/MK2 pathway is required for kaposin-B-mediated PROX1 upregulation . Notably , previous studies have shown that p38 MAPK promotes cytoplasmic accumulation of HuR in different cell types [70] , [71] , [72] . Therefore , we asked if activated MK2 can promote cytoplasmic localization of HuR protein in LECs and transfected LECs with vectors expressing a wild type , constitutively active ( EE ) or dominant negative ( K76R ) form of MK2 protein [73] . We found that whereas the constitutively active ( EE ) MK2 protein stimulated cytoplasmic accumulation of HuR , wild type or dominant negative MK2 protein did not ( Figure 7B ) . Moreover , the expression of PROX1 was upregulated by constitutively active ( EE ) MK2 protein , but not by wild type or dominant negative MK2 protein , in LECs determined by western analyses ( Figure 7C , D ) . Importantly , this upregulation of PROX1 by MK2 ( EE ) protein was abrogated by siRNA-mediated knockdown of HuR ( Figure 7C , D ) , indicating that HuR is required for the kaposin-B/p38/MK2 pathway-mediated PROX1 upregulation . We also confirmed these findings by using quantitative qRT-PCR measuring PROX1 mRNA level ( Figure 7E ) . Interestingly , we found that HuR knockdown slightly decreases PROX1 expression compared to control siRNA ( Figure 7D , E ) . We think that this is because HuR protein is present at a basal level in the cytoplasm of LECs ( Figure 6B ) and may stabilize PROX1 mRNA under the normal condition and thus knockdown of HuR resulted in decrease of PROX1 expression . This speculation is supported by our endogenous HuR co-immunoprecipitation data ( Figure 4D ) demonstrating a stable complex formation between HuR protein and PROX1 mRNA in normal primary LECs . We next asked whether MK2 ( EE ) protein-mediated PROX1 upregulation is due to PROX1 mRNA stabilization and thus studied the regulation of the PROX1 mRNA half-life by MK2 ( EE ) protein . Indeed , MK2 ( EE ) promoted PROX1 mRNA stability by increasing mRNA half-life by more than 60 minutes in LECs ( Figure 7F ) . Moreover , we found that this increase in PROX1 stability by MK2 ( EE ) could be abrogated by knockdown of HuR ( Figure 7G ) . Taken together , our data demonstrate that activation of the p38/MK2 pathway results in cytoplasmic accumulation of HuR protein and PROX1 upregulation through stabilization of PROX1 mRNA . Our data above demonstrate the essential contribution of HuR in PROX1 upregulation by kaposin-B . We next asked whether HuR plays an important role in KSHV-mediated PROX1 upregulation in primary human BECs and HUVECs . We first confirmed the KSHV-mediated PROX1 mRNA upregulation in BECs and HUVECs ( Figure 8A ) and also PROX1 protein expression by KSHV in HUVECs ( Figure 8B ) . We next investigated the cytoplasmic accumulation of HuR protein in KSHV-infected cells and found that KSHV infection resulted in a significant cytoplasmic localization of HuR protein in HUVECs ( Figure 8C ) . Consistently , we also observed HuR cytoplasmic localization in KSHV-infected BECs ( data now shown ) . We then asked whether HuR plays a role in the KSHV-mediated PROX1 upregulation by knockdown of HuR by siRNA in KSHV-infected BECs and HUVECs . Importantly , knockdown of HuR significantly decreased the half-life of PROX1 mRNA in KSHV-infected BECs and HUVECs ( Figure 8D , E ) . In comparison , we were not able to measure the half-life of PROX1 mRNA in uninfected BECs or HUVECs due to their low/absent expression of PROX1 . Taken together , our data demonstrate that KSHV infection stimulates cytoplasmic localization of HuR protein and that HuR plays an important role in KSHV-mediated PROX1 upregulation . Based on our data presented here , we build a working model for the molecular mechanism underlying KSHV-mediated PROX1 upregulation ( Figure 9 ) . When KSHV infects vascular endothelial cells , the virus may employ two or more mechanisms for PROX1 upregulation: one may be a transcriptional activation of PROX1 possibly through Akt activation [59] and the other a post-transcriptional PROX1 mRNA stabilization by kaposin-B , which activates the p38/MK2 pathway . Activated MK2 by kaposin-B stimulates the nuclear export and cytoplasmic accumulation of HuR protein . Cytoplasmic HuR protein binds to the AU-rich region in the 3′-UTR of PROX1 mRNA and slows down PROX1 mRNA turnover , thus increasing the steady-state level of PROX1 mRNA in KSHV-infected cells .
KS tumor cells were reported to be derived from endothelium about 40 years ago [74] . However , the exact histogenetic origin of KS had remained uncertain because KS cells were found to express mixed cell-lineage markers of BECs and LECs [75] . Previously , we and others demonstrated that KSHV induces lymphatic reprogramming of vascular endothelial cells by upregulating the master control gene of lymphatic differentiation , PROX1 [20] , [21] , [22] , [23] . This finding of endothelial cell fate reprogramming by KSHV has provided an important insight into the pathology of KS and KSHV . Nonetheless , the question how the virus induces the host cell fate change remained to be answered . Our previous study revealed that the KSHV latent gene LANA only marginally induced expression of PROX1 ( 1 . 93-fold ) [20] . Interestingly , recent two exciting studies have established a molecular connection between the function of LANA and PROX1 gene regulation: Di Bartolo et al . showed that KSHV LANA inhibits TGF-β signaling through epigenetic silencing of TGF- β type II receptor [76] and Oka et al . demonstrated that inhibition of TGF-β signaling upregulates PROX1 by ∼2 fold [77] . In spite of this intriguing molecular association , the degree of KSHV-induced PROX1 upregulation ( >8 fold ) shown by us and others [20] , [21] , [22] , [23] prompted us to hypothesize that LANA could not be the major activator for PROX1 upregulation in KSHV-infected cells and that another mechanism should also be present for KSHV-mediated PROX1 upregulation . This rationale directed us to search for an additional activator ( s ) among KSHV latent genes . In this study , we investigated the role of kaposin-B , a latent gene of KSHV , in the KSHV-mediated PROX1 upregulation and found that kaposin-B promotes mRNA stability of PROX1 . We defined the structure of PROX1 mRNA and identified a class I-type ARE in its 3′-UTR , through which PROX1 expression can be post-transcriptionally regulated by physiological or pathological stimuli . At this point , it seems that kaposin-B targets both class I ( e . g . , PROX1 ) and class II ( e . g . , GM-CSF [57] ) mRNAs and the specificity determinant for kaposin-B targets needs to be further defined . Moreover , we discovered that the ARE-binding protein HuR is exported to the cytoplasm by kaposin-B and also by KSHV infection , and increase PROX1 mRNA stability . These findings are consistent with our observation from in vitro cell cultures and KS tumor samples that PROX1 upregulation occurs in only KSHV-infected cells , not in neighboring uninfected cells ( Figure 1 ) . Kaposin-B-induced mRNA stabilization appears an attractive model for PROX1 upregulation since kaposin-B has been shown to upregulate other cytokine mRNAs such as GM-CSF and IL-6 [57] . However , it needs to be highlighted that kaposin-B is not the sole component in the molecular mechanism underlying PROX1-upregulation by KSHV because mRNA stabilization inevitably requires pre-existing mRNA and PROX1 mRNA is not expressed in BECs [24] , [25] . Therefore , other factors/stimuli are needed for the initial transcriptional activation of the PROX1 mRNA synthesis . This hypothesis is further supported by two of our findings . First , PROX1-upregulation by kaposin-B was much prominent in LECs where PROX1 mRNA is already present , in comparison to BEC and HUVEC-backgrounds where PROX1 expression is fairly low , if any ( Figure 2A ) . Second , PROX1 expression in BECs and HUVECs was more strongly activated by the entire virus ( KSHV ) , compared to by kaposin-B alone ( Figures 2A & 8A ) , again suggesting that kaposin-B alone is unable to efficiently activate PROX1 expression in BECs and HUVECs . Importantly , Morris et al has recently shown that Akt activation through gp130 is required for KSHV-mediated PROX1 upregulation and lymphatic reprogramming [59] . Considering the fact that KS tumors have been associated with numerous cytokines , chemokines and diffusible factors in their microenvironments [78] , [79] , [80] , [81] , it is highly likely that multiple KS-associated viral and/or cellular factors may activate the gp130/Akt pathway to prime the initial activation of PROX1 transcription [59] . We hypothesize that this transcriptional activation may require a subsequent secondary post-transcriptional mechanism involving kaposin-B to achieve PROX1 upregulation . This two-step mechanism is also consistent with the fact that PROX1-upregulation is limited to KSHV-infected cells . Taken together , Figure 9 illustrates our hypothesis that both steps ( transcriptional activation and mRNA stabilization ) may be necessary to achieve PROX1-upregulation and lymphatic reprogramming of blood vascular endothelial cells by KSHV . On the other hand , kaposin is a unique KSHV latent gene considering distinct features in its transcription and translation . Kaposin transcripts are the most abundantly expressed viral mRNA throughout all stages of KS progression determined by in situ hybridization assays and a complex translational program directs production of multiple isoforms of kaposin gene product , termed kaposin A , B and C [82] , [83] . Kaposin-B uses a non-conventional CUG start codon and consists of a series of tandem repeats of hydrophobic 23-amino acids , named DR1 and DR2 [82] , [83] . DR2 can directly bind to MK2 and , when overexpressed , DR2 domain alone can dominantly inhibit the mRNA-stabilization function of the whole kaposin-B protein [57] , [58] . It was also found that the DR1/DR2 repeats are more abundantly expressed in lytic- or TPA-treated cells [84] and that DR2 is reiterated three to five times in different stains of KSHV [58] , suggesting a significant variation in the expression level and DR2 repeat number of kaposin-B . Accordingly , it will be interesting to investigate whether kaposin-B-induced PROX1 mRNA stabilization is more prominent during viral reactivation . Moreover , it is possible that both the expression level and structure of kaposin-B may affect its mRNA-stabilizing function through different ARE-binding proteins ( including HuR ) and target a different set of cellular mRNAs . It would be also exciting to study if kaposin-B may have other functions in addition to its role in mRNA stabilization . Our current study brings up numerous questions . To date , the pathological role of PROX1 in KS development and progression has not been defined . Is the lymphatic phonotype more favorable for KSHV infection and propagation ? Is lymphatic reprogramming a by-product or a goal of PROX1 upregulation by KSHV ? Some insights may be obtained from interesting findings that PROX1 was shown to increase the invasion of endothelial tumor cells [85] and that PROX1 promotes the transition from benign to highly dysplastic phenotype in colon cancer [86] . While these studies support the oncogenic roles of PROX1 , many other studies demonstrate the opposing role of PROX1 as a tumor suppressor [87] , [88] , [89] , [90] , [91] , [92] . Therefore , cell type and tissue microenvironment may be crucial for PROX1 to play the oncogenic versus tumor suppressive role and further studies will be necessary to better understand the role of PROX1 in KS tumor development . Moreover , PROX1 has been reported to be important for cell-fate specifications in a broad range of cells including lymphatics [24] , [25] , liver [26] , lens [27] , [28] , brain [29] , [30] , [31] , [32] , the ear [33] , [34] , [35] , [36] and the heart [37] , [38] during development and has been associated with post-developmental processes such as cell cycle regulation [28] , [39] , [93] and inflammation [94] . It will be very interesting to investigate whether PROX1 is post-transcriptionally regulated for any of its functions during and after development .
Human primary dermal blood vascular endothelial cells ( BECs ) and lymphatic endothelial cells ( LECs ) were isolated from anonymous neonatal human foreskins and cultured as previously described [95] with an approval of the University of Southern California Internal Review Board ( PI: YK Hong ) . Primary human umbilical venous endothelial cells ( HUVECs ) were purchased from Lonza ( Basel , Switzerland ) , and cultured in EGM-2 medium ( Lonza ) . NIH3T3 cells containing tTA ( Tet-Off ) , named B2A2 , were kindly provided by Dr . Ann-Bin Shyu ( University of Texas Houston Health Science Center ) [64] , [65] , [96] . NIH3T3 cells were transfected by using Lipofectamine 2000 ( Invitrogen ) and primary endothelial cells were transfected by electroporation ( Nucleofactor II , Amaxa Biosystems ) . The pTet-BBB and pTet-BBB-Fos-ARE vectors were kindly provided by Dr . Ann-Bin Shyu ( University of Texas-Houston Medical School ) [64] , [65] . pTet-BBB contains the Tetracycline ( Tet ) - responsive element that drives transcription of the rabbit β-globin reporter gene and pTet-BBB-Fos-ARE bears the Fos-ARE inserted in the 3′ UTR of the reporter . To make pTet-BBB-Prox-ARE that contains the 40-bp PROX1-ARE ( 4 , 994 ∼5 , 034 bp downstream from the stop codon of human PROX1 ) , sense and anti-sense primers harboring the ARE and BamHI-half site at the both ends ( gatccTGCATAATTTATTGGTTTAATTTATCCTAATTTATTTGATG , gatccATCAAATAAATTAGGATAAATTAAACCAATAAATTATGCAG ) were annealed and cloned at the unique BglII site of the 3′-UTR of pTet-BBB . To clone the human PROX1 3′-UTR , a 5 . 4-kb fragment covering the human PROX1 3′-UTR was amplified by two primers ( ATTAGCGGCCGCTTTGAATGTATGAAGAGTAGCAGTCC , AATCAAACGGCACTGAGCTT ) from a bacterial artificial chromosome ( RPCI11-71F10 , Invitrogen ) and was cloned in pCRII-Blunt ( Invitrogen ) . Expression vectors encoding MYC-tagged MK2 ( WT , EE , and K76R ) were kindly provided by Dr . Matthias Gaestel ( Hannover Medical School , Germany ) [73] . Expression vectors for myc-tagged HuR and FLAG-tagged kaposin-B were kind gifts by Drs . Dominique Morello ( Institute Pasteur , France ) [97] and Craig McCormick ( Dalhousie University , Canada ) [57] , respectively . SiRNA for HuR was purchased from Santa Cruz Biotechnology ( siHuR; cat . sc-35619 ) and the control siRNA ( CUUACGCUGAGUACUUCGATT , UCGAAGUACUCAGCGUAAGTT ) against the firefly luciferase was previously described [95] . 3′-RACE assay was performed by following the manufacturer's instruction ( First Choice RLM Race kit , Applied Biosystems ) . Total RNA was isolated from human primary LECs and subjected to 3′ RACE by using two sets of PCR primers ( GGATTGGTCTCAGCGCTACC , GCGAGCACAGAATTAATACGACT; AACTGAACTGATAAAGTCAATTTTTG , CGCGGATCCGAATTAATACGACTCACTATAGG ) . Amplified PCR products were cloned in the pGEM-T Easy vector ( Promega ) and sequenced to define the end of PROX1 mRNA . De-identified anonymous KS specimens were obtained from the AIDS and Cancer Specimen Resource ( ACSR ) with an approval of the University of Southern California Internal Review Board ( PI: YK Hong ) . KSHV was purified from BCBL-1 cells by following a previous description [23] with a minor modification . Briefly , BCBL-1 cells were cultured to the density of 5∼10 million cells/ml and then activated with TPA ( 20 ng/ml ) and sodium butyrate ( NaB , 3 mM ) . After 24∼48 hours , TPA/NaB-containing media was replaced with normal media and cells were incubated for additional 5 days . Culture media was then collected and filtered through 0 . 45-µm filters and centrifuged for 20 minutes at 4°C at 8 , 000 rpm to remove cell debris . Supernatant was centrifuged for 5 hours at 4°C at 11 , 000 rpm to pellet the virus , which was then resuspended in endothelial cell media . Infectivity was measured by immunohistochemistry for LANA after infection for 5 days . GST-HuR and GST-TTP fusion proteins were isolated as previously described [98] . Expression vectors for GST-HuR and GST-TTP were kindly provided by Drs . Henry Furneaux ( University of Connecticut ) [51] and Gilles Pages ( University of Nice-Sophia Antipolis , France ) [99] , respectively . For RNA EMSA experiments , radio-labeled RNA transcripts ( 200 Kcpm/reaction ) was mixed with 200 ng GST fusion proteins or GST alone in a previously described binding buffer [100] . The reaction mixture was incubated for 30 minutes at 30°C and treated for 15 minutes at room temperature with 100 U of Ribonuclease T1 ( Roche ) . For competition assays , specific ( unlabeled PROX1-ARE RNA probe ) or nonspecific ( yeast total RNA ) competitors were incubated for 15 minutes at 30°C with the proteins in the binding buffer before the addition of the labeled transcripts . The reaction mixtures were resolved on 8% native polyacrylamide gels in 0 . 5× Tris borate-EDTA ( TBE ) buffer . Gels were dried and exposed to X-ray film . RNA/protein immunoprecipitation assay was performed essentially as described [96] by using protein A/G-Sepharose beads pre-incubated with anti-HuR ( 3A2 , SC-5261 , Santa Cruz Biotechnology ) . RNA was isolated from supernatants and reverse-transcribed with Superscript II ( Invitrogen ) . The transcripts were amplified by PCR by using the following primers: PROX1 ( probe-1 ) , ATCCTAATTTATTTGATGAAGGTG , TGCACATACATTCAGTTTAAGAGG; PROX1 ( probe-2 ) , TCAGTGCCGTTTGATTTTCTTAAA , GGAACA TCTTTCCTTGTTCTTAGA; VEGF , TCCAATCTCTCTCTCCCTGAT , CGGATAAACAGTAGCACCAAT [67]; and UBE2N , TACCCAATGGCAGCCCCTAA , TTCCACTGCTCCGCTACATCA [68] . The resulting PCR products were analyzed by 2% agarose gels . Cells were cultured on 8-mm cover slips and infected with KSHV for 5∼7 days or transfected with FLAG-tagged kaposin-B for 16 hours . Cells were then fixed with 4% paraformaldehyde for 10-minutes , washed in phosphate-buffered saline solution ( PBS ) and treated with blocking solution ( 5% Bovine serum albumin ) overnight . Subsequent immunostaining was performed as previously described [95] . Antibodies and dilution factors used for immunofluorescent staining analyses are follows; PROX1 ( 1∶1000 , ReliaTech , Germany ) , LANA ( 1∶1000 , Advanced Biotechnologies Inc , Maryland ) , HuR ( 1∶1000 , Santa Cruz Biotechnology ) , FLAG tag ( 1∶1000 , Sigma-Aldrich ) . Antibodies used for western analyses were PROX1 ( 1∶1000 , Millipore Corporation , MA ) , β-actin and FLAG tag ( all in 1∶2000 , Sigma-Aldrich Corporation ) , MYC tag ( 1∶1000 , Covance ) , lamin A/C , p38 and phospho-p38 ( all in 1∶1000 , Cell Signaling Technology ) , tubulin , MK2 , phospho-MK2 , and HuR ( all in 1∶1000 , Santa Cruz Biotechnology ) . Total RNA was purified from primary human LECs , separated in an agarose gel , transferred to nylon membrane and then hybridized with 32P-labeled DNA probes . A 2 . 2-kb fragment was prepared as the PROX1 ORF probe by digesting pcDNA3-hPROX1 with NotI [98] . Three 3′-UTR probes , P1 ( ∼2 . 3-kb ) , P2 ( ∼0 . 5-kb ) and P3 ( ∼0 . 7-kb ) , were prepared by EcoRV , XbaI/XmnI , and AfeI/SpeI digestions from PROX1 5 . 4 kb 3′ UTR , respectively . For the functional test of PROX1-ARE , NIH3T3/Tet-Off cells were transfected with pTet-BBB , pTet-BBB-Prox-ARE or pTet-BBB-Fos-ARE [64] , [65] and were grown in doxycycline ( 40 ng/ml ) -containing DMEM/FBS medium for 48-hours . Transcription of β-globin was induced by removing doxycycline ( replacing media with doxycycline-lacking media ) for 110-minues . Subsequently , doxycycline ( 500 ng/ml ) was added to the media to shut down the transcription and total RNA was harvested after 0 , 30 , 60 , 90 , 120 , 180 and 240-minutes . Northern blot analysis was performed by using 32P-labeled β-globin and GAPDH probes . Real-time RT-PCR ( qRT-PCR ) was performed by using TaqMan EZ RT-PCR Core Reagent ( Applied Biosystems ) . For dual-labeled probe-based qRT-PCR reactions , each reaction was multiplexed for both target gene and the internal control β-actin for normalization . Conventional RT-PCR was performed by using Superscript II ( Invitrogen ) and Taq polymerase ( New England Biolabs ) . Primer sequences will be provided upon request .
|
Kaposi's sarcoma ( KS ) is the most common cancer in HIV-positive patients and KS-associated herpes virus ( KSHV ) was identified as its causing agent . We and others have discovered that when the virus infects endothelial cells of blood vessels , KSHV reprograms the cell type resembling endothelial cells in lymphatic vessels . Although endothelial cells of the blood vascular system and of the lymphatic system share functional similarities , the cell type-reprogramming does not occur under a normal physiological condition . Therefore , cell-fate reprogramming by the cancer-causing virus KSHV provides an important insight into the molecular mechanism for viral pathogenesis . Our current study investigates the molecular mechanism underlying the KSHV-mediated cell fate reprogramming . We identified that a KSHV latent gene kaposin-B plays an important role in KSHV-mediated regulation of PROX1 to promote PROX1 mRNA stability . This study will provide a better understanding on the tumorigenesis and pathogenesis of KS with a potential implication toward new KS therapy .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"virology/viral",
"replication",
"and",
"gene",
"regulation",
"infectious",
"diseases/hiv",
"infection",
"and",
"aids",
"virology/viruses",
"and",
"cancer",
"virology/effects",
"of",
"virus",
"infection",
"on",
"host",
"gene",
"expression",
"cell",
"biology/gene",
"expression"
] |
2010
|
Kaposin-B Enhances the PROX1 mRNA Stability during Lymphatic Reprogramming of Vascular Endothelial Cells by Kaposi's Sarcoma Herpes Virus
|
Repositioning existing drugs for new therapeutic uses is an efficient approach to drug discovery . We have developed a computational drug repositioning pipeline to perform large-scale molecular docking of small molecule drugs against protein drug targets , in order to map the drug-target interaction space and find novel interactions . Our method emphasizes removing false positive interaction predictions using criteria from known interaction docking , consensus scoring , and specificity . In all , our database contains 252 human protein drug targets that we classify as reliable-for-docking as well as 4621 approved and experimental small molecule drugs from DrugBank . These were cross-docked , then filtered through stringent scoring criteria to select top drug-target interactions . In particular , we used MAPK14 and the kinase inhibitor BIM-8 as examples where our stringent thresholds enriched the predicted drug-target interactions with known interactions up to 20 times compared to standard score thresholds . We validated nilotinib as a potent MAPK14 inhibitor in vitro ( IC50 40 nM ) , suggesting a potential use for this drug in treating inflammatory diseases . The published literature indicated experimental evidence for 31 of the top predicted interactions , highlighting the promising nature of our approach . Novel interactions discovered may lead to the drug being repositioned as a therapeutic treatment for its off-target's associated disease , added insight into the drug's mechanism of action , and added insight into the drug's side effects .
The continuing decline of drug discovery productivity has been documented by many studies . In 2006 , only 22 new molecular entities were approved by the Food and Drug Administration ( FDA ) despite research and development expenditures of $93 billion USD by biotech companies and large pharmaceutical companies , and this low productivity has not improved since [1] . From discovering , developing to bringing one new drug to market , clinical trials are the most expensive step , accounting for 63% of the overall cost [2] . To this end , drug repositioning - finding new therapeutic indications for existing drugs - represents an efficient parallel approach to drug discovery , as existing drugs already have extensive clinical history and toxicology information . Many of today's repositioned drugs were discovered through serendipitous observations , including high profile drugs sildenafil by Pfizer - first developed for angina but later approved for erectile dysfunction - and thalidomide by Celgene - first marketed for morning sickness , then approved for leprosy and recently for multiple myeloma [3] . Repositioned drugs have also been discovered through rational observations , including imatinib ( Gleevec ) , which was first approved for chronic myeloid leukemia by targeting the BCR-Abl fusion protein but was subsequently approved for gastrointestinal stromal tumor due to its ability to potently inhibit c-KIT [4] . Another example is the anti-depressant duloxetine ( Cymbalta ) that is also indicated for stress urinary incontinence based on a shared mechanism of action between the two diseases [3] . In order to rationally reposition drugs , novel target-disease or drug-target relationships must first be elucidated . By screening compounds against a panel of proteins , there is potential to discover novel drug-target interactions . Drug candidates are routinely screened against a small panel of similar proteins to determine their specificity to the intended target . Large panels with hundreds of kinase proteins have been developed to assess kinase inhibitor specificity [5] , especially since we now know that many kinase drugs are multi-targeting . However , the druggable proteome is much larger than just the kinome , so larger and more varied protein panels are needed to truly assess drug specificity . With the availability of massively parallel DNA sequencing technology , recurrently mutated proteins in diseases – such as EZH2 in certain lymphomas [6] and FOXL2 in certain ovarian cancers [7] - are now being rapidly determined and are also relevant drug targets . However , testing all drugs against all targets experimentally is extremely costly and technically infeasible . Recent computational endeavors to predict novel drug repositioning candidates have used methods incorporating protein structural similarity [8] , chemical similarity [9] , or side effect similarity [10] . One study also incorporated some molecular docking to help filter interactions predicted through protein binding site similarity [8] . Here we present a large-scale molecular docking analysis of known drugs against known protein targets for the prediction of novel drug-target interactions . Molecular docking is a computational method that predicts how two molecules interact with each other in 3-dimensional space . It is well established as a virtual screening method in drug discovery [11] , where typically many chemicals are docked against a specific protein binding site , in order to discover novel inhibitors of that target . Compared to similarity analyses , docking has the potential to find drugs that bind to proteins with novel scaffolds as well as off-targets that may be structurally dissimilar to the known targets . Large-scale docking of many targets to many drugs is now feasible when run on powerful computer clusters . However , limitations in scoring methods result in high false positive prediction rates [12] , and large-scale studies amplify these low prediction accuracies . Our method emphasizes removing false positive predictions using scoring and ranking thresholds , and retaining only the highest confidence interactions as drug repositioning candidates .
A computational pipeline was developed for large-scale molecular docking of drugs to protein targets ( Figure 1 ) . Briefly , we collected all 3D structures available for each drug target , determined binding pockets in the structures , and docked drugs to each pocket . Results were collected and thresholds were applied to select the top predicted interactions , which were then visually inspected . We first docked 3570 known protein-drug interactions annotated by DrugBank , between 678 unique human proteins and 1309 small molecule drugs . We used the docking software ICM developed by Molsoft [13] , which ranks ligands using a Monte-Carlo based docking procedure and an empirical , energetics-based docking score . Like most docking software , ICM recommends a standard score cut-off for virtual screening efforts: −32 [14] , where more negative scores represent more likely binding interactions . However , studies have used different cut-offs depending on the protein target [15] . Here we used a score of −30 as the threshold for ‘good’ dockings scores . Of the 3570 known interactions docked , 1116 ( 31% ) had a good ICM docking score . 252 proteins had at least one known interaction predicted by docking – these formed the ‘reliable’ set of proteins that we believe are more suited for docking purposes . A breakdown of protein classifications for this reliable set revealed that 67% of targets were enzymes , of which 12% were protein kinases . In contrast , there were few G-protein coupled receptors in our database due to lack of crystal structures , which reflects both the current state of solved protein crystal structure space as well as popular drug targets . In high-throughput molecular docking , it is common to hold protein structures rigid during the simulation . With this restriction , re-docking a PDB ligand back to its native PDB structure ( cognate docking ) is a simpler task than docking a different ligand to the structure ( non-cognate docking ) because in the former case the protein is already in a specific ligand-bound conformation . Cognate-docking situations occur frequently and previous studies show that they can be docked well in 60–80% of cases [16] . In contrast , the more useful non-cognate docking is only successful in 20–40% of cases [16] . We analyzed the 1116 known interactions to examine whether those that docked well were only due docking cognate ligands . For each interaction , we observed whether the drug bound 1 ) a holo ( unliganded ) protein structure , 2 ) an apo ( liganded ) structure with a same or similar ligand as the drug ( the cognate-docking scenario ) , or 3 ) an apo structure with a chemically different ligand from the drug . Chemical similarity was defined as having a Tanimoto coefficient less than 0 . 54 . Figure 2 shows that cognate docking occurred in 380 of the 1116 interactions . Of these , only 56 were drugs docked to an apo protein with the same ligand ( Tanimoto coefficient of 0 ) . The majority of drugs docked well to holo structures as well as apo structures with dissimilar ligands . In short , the ICM docking method was able to predict known interactions for both cognate and non-cognate docking scenarios . Aside from the docking score , it was also important to verify that the ligands were docked in correct binding conformations . We further examined the 380 cognate dockings and found that the docked drug conformation was close to the known drug conformation ( RMSD value ≤2 Å ) in 69% of cases . The other 31% fell into two categories: 1 ) partly symmetrical ligands like NAD and 2 ) ligands that bound to a small pocket . In the first case , the molecule was incorrectly determined to be flipped , causing a high RMSD; however , its central portion was docked correctly due to symmetry . In the second case , the region of ligand bound in the pocket was docked correctly , but the region free in solvent contributed to a poor RMSD value . Overall , this analysis showed that when a known interaction was docked with a good score , the binding conformation was also reasonably predicted . We gathered the known protein-drug interactions into a network ( Figure 3 ) with proteins as rectangular nodes , drugs as circular nodes , and interactions as edges . Interaction edges with good docking scores were highlighted in red . Proteins from the same family were often grouped close together and shared many drug interactions , such as the retinoid X and retinoic acid receptors and the matrix metalloproteinases . Proteins having the most known drug interactions in the network included the transport proteins serum albumin and the phosphatase PTPN1 . The most highly-connected chemicals in the network were metabolites: ATP , NAD , and NADP . For some proteins such as MAPK14 , 13 of 14 known inhibitors were well predicted by docking , whereas for others such as ACE , only one of its nine known inhibitors scored well . For 426 of the 678 protein targets not included in Figure 3 , none of their known interacting drugs could be docked well , reflecting the limitations of current molecular docking methods . To this end , we chose the subset of 252 protein targets for which at least one known drug docked well ( from the 1116 interactions that docked well ) , deemed as more ‘reliable-for-docking’ compared to the other proteins . We proceeded to dock the 252 reliable protein set against the database of 4621 drugs . Considering the multiple crystal structures per protein and the multiple binding pockets per structure , there were a total of 1514 crystal structures and 2923 binding pockets . Each drug was docked to all binding pockets of a protein and whichever pocket gave the best docking score for the drug determined the final protein-drug score . This method allowed multiple conformations of a protein to be accounted for during docking and provided a simple model of protein flexibility . In total , we docked 1 . 2 million protein-drug interactions . 104 , 625 ( 0 . 9% ) had ICM docking scores ( icm-score ) of −30 or better , encompassing all1116 known interactions in the reliable data set . Since the fraction of known interactions in the predicted set was so low , we assumed that the vast majority of predictions were false positives . Though we believed that novel drug-target interactions existed and were enriched within these 104 , 625 , there was clearly a need for more stringent score thresholds . We investigated various methods of selecting top drug-target interactions . The standard software-recommended icm-score is based on a weighted sum of various binding energy terms [13] . The pmf-score , or potential of mean force score , is a measure of the statistical probability for the drug and protein to interact with each other ( for example , it examines interatomic distances and atom types of the docked interaction and compares that to existing interactions in PDB ) [14] . A consensus score was developed that uses both icm- and pmf- scores and allows us to select the x% of top interactions for each protein; it is described in more detail in case studies below . We also ranked interactions in two ways . The drug-rank is the rank of this drug compared to all drugs docked to this protein ( from 1–4621 ) , and the protein-rank is the rank of this protein when the drug is docked to all proteins ( from 1–252 ) . Requiring high drug and protein ranks ( i . e . a low value when the two ranks are summed together ) enforces a mutual specificity criterion . We hypothesized that by choosing interactions with good scores and ranks , we would better filter out false positive predictions . To assess performance , we measured the positive predictive value ( PPV ) , defined as the proportion of predicted interactions that are known binding interactions . The premise is that a better threshold would yield a set of predictions more enriched with known interactions , and novel interactions that are more likely be true binding events . Figure 4a shows that as the stringency of a threshold increased ( i . e . icm-score of −40 versus −30 ) , fewer interactions are predicted; however , the PPV increased due to a higher proportion of known interactions in the predicted set . This behavior is consistent for all thresholds , and the highest PPVs are generally observed within the top 100 predicted interactions . It is important to note that each of the 4621 drugs will always have a top-ranked protein ( interactions with protein-rank of 1 ) , and each of the 252 proteins will always have a top-ranked drug ( interactions of drug-rank 1 ) . Thus , the protein-rank threshold particularly is not sensitive alone . The protein-rank and pmf-score thresholds appeared to be the worst based on both the PPV plot ( Figure 1 ) and on enrichment factors ( Table 1 ) . However , they showed better PPVs when combined with other thresholds . For example , the drug rank and protein rank measure performs much better than drug-rank alone , and the consensus score ( combining icm- and pmf-score ) also performs better than the icm-score alone . We measured the enrichment factor for each type of threshold , at its most stringent setting ( leftmost points of Figure 4a ) and found that the pmf-score and protein-rank were the least effective at predicting known drugs ( Table 1 ) . Instead , combinations of score and rank criteria provided a 100–500× enrichment of known interactions compared to a random algorithm , and a10–50× enrichment compared to a standard binding energy-based ICM score cut-off of −30 . Interestingly , the drug-rank 1 and protein-rank 1 ( basically the sum of ranks is 2 ) combination threshold performs surprisingly well; however , adding the consensus score clearly improves PPV for the top ∼300 interactions ( Figure 4b ) which are the most interesting to us for manual inspection . Another threshold method is to use the scores of known binders as the score cut-off for each protein . We investigated this using the best and worst icm- and pmf-scores of known drugs . Table 2 shows that this did not result in a higher enrichment , nor did it help narrow down the number of predicted interactions . Overall , the combination of consensus score with the two ranks gave the highest PPV and enrichment values: in the top 50 predicted interactions , 49% are known . This gave us confidence that many of the other 51% , all novel interactions , are real . Two examples are presented to illustrate the utility of combining rank and scoring criteria . The first is for the signaling protein MAPK14 ( also known as p38 alpha ) , an integral component in numerous cellular processes . It is a drug-target for inflammatory diseases [17] . MAPK14 is known to be a challenging docking target due to its structural flexibility [18] and its shallow binding pocket [19] . However , these docking studies used only one 3D structure . In our dataset , there are 35 crystal structures of MAPK14 in different conformations , providing a simple view of protein flexibility . The consensus score is based on the observation that when docking a large number of diverse compounds to any target , most compounds have poor icm- or pmf- scores , and few compounds have both good icm- and pmf- scores . Therefore , we chose a linear threshold that eliminated the densest area of points in the poor scoring region ( top-right ) of a score plot like Figure 5 , and selects the compounds in the best scoring region ( bottom-left ) as potential interaction hits . As seen in Figure 4a and Table 1 , the consensus score performed better for PPVs and enrichments compared to a simple icm- and pmf- score combination . Figure 5 plots the icm- versus pmf- scores of the 4621 drugs docked to MAPK14 . Each drug is a point on the graph , where the 5% of drugs passing a consensus threshold are shown in orange , and the 1% passing a consensus threshold are shown in purple . For 67 drugs , MAPK14 was one of the top 5 scoring targets; they are circled in green . Table 3 shows that a combination of the consensus and protein rank criteria resulted in the best enrichment ( 110× ) of known drugs . There were 15 annotated known binders of MAPK14 in DrugBank , but we disregarded 2-chlorophenyl due to it being a very small molecule with a very weak MAPK14-binding affinity ( >1 mM ) . 10 of 14 known drugs were predicted through our stringent thresholds . Though 4 true positive binders were lost , 99 . 99% of points were eliminated , presumably consisting mostly of non-binders . Through literature search , we found that imatinib and quercetin have been previously tested against MAPK14 and did not show any inhibition [20] . This suggested that the 5% consensus threshold was too lenient for MAPK14 , whereas the 1% was more appropriate . Within the other approved drugs predicted to bind MAPK14 , we found literature validation for sorafenib , a multi-kinase inhibitor approved for renal cell carcinoma [21] , and gefitinib , a EGFR inhibitor approved for late stage non-small cell lung cancer [22] . Previous high-throughput studies have shown varying results regarding nilotinib-MAPK14 inhibition . Some enzymatic assays to MAPK14 showed weak inhibition: 570 nM or 2 . 2 µM depending on the assay type [23] . Direct binding assays have shown 100 nM Kd [23] or no binding at all in peptide pulldown experiment [20] . Since nilotinib was one of our top approved drugs predicted to bind MAPK14 , we decided to further experimentally validate the interaction . We performed MAPK14 ATP-competitive binding assays for two inhibitors that were available for purchase: zafirlukast , and nilotinib . As seen in Figure 6 , both drugs exhibited inhibition of MAPK14 at therapeutically relevant concentrations ( <10 µM ) in a dose dependent manner . Zafirlukast ( AstraZeneca ) is an oral leukotriene inhibitor that reduces inflammation of breathing passage in asthma patients . We found that it does inhibit MAPK14 weakly , and this may contribute to its inflammation reducing effect . The chronic myeloid leukemia drug nilotinib was especially potent with an IC50 of 40 nM . Despite their appeal as an inflammatory disease target , MAPK14 drug candidates to date have failed due to drug toxicity issues [24] . Though it may seem underwhelming to use a cancer drug with potentially serious side effects to treat inflammation , nilotinib is noted to have a much milder adverse effects profile compared to its similar drug dasatinib [20] . Another similar drug imatinib has shown promise in treating rheumatoid arthritis in mouse models [25] and specific patients [26] , [27] , speculated due to its inhibition of mast cell c-Kit and PDGFRB . Nilotinib also inhibits these two proteins , and its extra inhibition of MAPK14 may render it a better choice for arthritis models . Recently , nilotinib was tested in a glucose-6-phosphate-isomerase-induced arthritis mouse model and found to significantly prevent paw inflammation – to a greater extent than imatinib [28] . This study also suggested that the two drugs acted through some distinct mechanisms . Overall , these findings seem to agree with our observations that nilotinib potently inhibits MAPK14 , unlike imatinib , and thus has added potential as an anti-inflammatory drug . A second example is the Protein Kinase C inhibitor BIM-8 . We docked BIM-8 to the set of 252 reliable targets , and the results are plotted in Figure 7 . Each point on the graph represents a protein target , and targets for which BIM-8 passes the 5% consensus threshold are shown in orange . We compared our results to three previous studies . Two studies performed protein kinase assays with radioactive ATP and substrate peptides , where inhibitor binding decreases the amount of radioactive peptide produced [29] , [30] . The third study performed thermal shift assays where inhibitor binding increased the kinase stability and thus the melting point [31] . BIM-8 targets discovered by these papers are shown in shades of red in Figure 7 , and non-binders in these papers are shown green . The only annotated target of BIM-8 in DrugBank is PDPK1 . GSK3B and PIM1 , which are in the top 5 protein rank and top 5% consensus threshold , were also validated as inhibitors . PDPK1 was not found to be an inhibitor by the first two studies but was confirmed as a binder by the third study with a kinase assay and crystal structure . Overall , if we count that there are 4 known binders ( PIM1 , PDPK1 , GSK3B , LCK , since CDK and MAPK14 are probably weak or nonbinders ) , we can see that applying a 5% consensus threshold and protein rank criteria gave us a 63-fold enrichment over random selection , and a 63/10 . 5 = 46 fold enrichment over using a standard ICM score threshold of −30 ( Table 4 ) . For a global and quantitative review of the predicted protein-drug interactions , we plotted the icm scores of drugs docked to established drug targets ( Figure 8 ) . Each protein is represented by a column , on which a black line denotes a known drug docked to the target , a red dot denotes an approved drug docked to the target , and a blue dot denotes an experimental drug docked to the target . Only protein-drug interactions that docked with a score passing the consensus threshold and had a protein-rank ≤5 are shown . Overall , the known drugs ( black crosses ) had better scores than other drugs for a given target . This was expected , as many of these known drugs were chemically optimized for their targets . For a number of targets , the known drug was the only predicted interaction . None of the approved and experimental drugs from DrugBank were able to dock well , despite a reliable protein structure , suggesting that virtually screening larger chemical databases may be the only way to discover novel inhibitors by docking . For most targets , at least one experimental drug showed a better score than the known drugs; however , experimental drugs are often unavailable for purchase or experimental testing . Instead , we were most interested in cases with approved drugs such as the MAPK14-sorafenib example which was verified by the literature , and the MAPK14-nilotinib example which we verified with an in vitro kinase assay . Through literature search , we found experimental support for many of our top drug-target predictions that scored better than known interactions ( Table 5 ) . These all pass the 1% consensus threshold and are observed to have high drug and protein ranks for the most part . It is important to note that the drug-rank depends on the number of known binders for the protein; thus , since ESR1 had 39 annotated drugs in DrugBank , a drug-rank of 16 is not as low . In contrast , a drug-rank of 16 would be low for MMP13 , which has only seven annotated drugs in DrugBank . One type of validated interaction includes drugs that are close analogs of known drugs for that target; for example , the estrogen analog ERA-923 is a known selective estrogen receptor modular ( SERM ) [32] . Genistein is known to bind both ESR1 and ESR2 [33] . Becocalcidol and ED-71 are vitamin-D analogs and bind the vitamin D receptor [34] , [35] . Drosiprenone is a synthetic progestin with anti-mineralocorticoid receptor ( MR , NR3C2 ) effects and has potential for reducing cardiovascular risk in women taking oral contraceptives or postmenopausal hormone treatment [36] . Due to the many in depth studies on kinase inhibitor specificity , we were able to find collaborating evidence for some of our kinase protein interaction predictions . For example , vatalanib is a known pan-VEGFR inhibitor [37] , nilotinib is a potent KIT inhibitor [38] , and other inhibitors of MAPK14 and targets of kinase inhibitor BIM-8 were discussed in previous sections . Docosahexanoic acid ( DHA , DB03756 ) is an endogenous ligand for brain fatty acid binding protein ( B-FABP ) that is essential for brain growth and function [39] . We predicted that it binds the transport protein human serum albumin; indeed , this interaction has been validated and found to confer neuroprotection in animal models of ischemia [40] . This finding suggested that DHA might have potential repositioning value for ischemic stroke . Overall , we were able to find literature support for 30 of our top predicted interactions , which validated our computational method as useful for finding novel drug-target interactions .
The binding of a small molecule drug to its target protein in a cell is much more complex than a single docking calculation . For example , an ATP-competitive kinase drug would have hundreds of ATP-binding sites to choose from due to the large size of the kinome . Cancer drugs such as sunitinib are now known to potently inhibit many more kinase targets than previously expected [41] . In addition , non-kinase targets of kinase drugs have also been found: NQO2 was the first non-kinase target discovered for imatinib [20] , [42] , and several cytotoxic LIM kinase inhibitors were found to be actually inhibiting tubulin [43] . Such studies imply that the target search space for any inhibitor should be the entire druggable proteome . Our strategy was to find novel drug targets of existing drugs by computationally screening the druggable proteome . For this purpose , we chose molecular docking due to its speed , low cost , and detailed three-dimensional simulation . Moreover , docking can evaluate any protein with a solved structure due to its virtual nature , without the need for tailoring enzymatic assays or collecting drugs in solutions . However , docking is known to have a high false positive prediction rate , due to limitations such as incomplete binding pocket prediction , inadequate ligand conformation sampling , inaccurate scoring functions , lack of protein flexibility , and lack of water and cofactor molecules during the simulation . As evidenced in this study , only 31% of the 3570 known interactions docked with a good score . One review states that 10–50% of a set of diverse compounds can be expected to be docked correctly for a given target [12] . We are well within this range , and believe our method performs quite well considering the large variety protein targets involved and the automated nature of the pipeline . However , the other 69% of known interactions were not predicted due to docking limitations . Our method attempted to address these limitations . First , we manually included binding pockets that were present in PDB structure complexes but not predicted by the binding pocket search . Second , we docked each interaction 10 times to better sample ligand conformations . Third , we applied consensus score and rank criteria to further narrow down top scoring docking hits . Fourth , we used all available structures of a protein ( versus choosing one representative structure ) , to allow a simple view of protein flexibility . We did not incorporate water and cofactor molecules in our docking simulations due to the computational complexity involved . However , by selecting proteins for which at least one known drug docked and scored well , we selected proteins for which the limitations of the docking software were not critical for a good prediction . In short , assuming the docked conformation of the known ligand was correct , we used only proteins for which the binding pocket was genuine , the scoring functions were adequate , the protein was in a conformation amenable for drug inhibition , and the lack of water or cofactor molecules didn't drastically affect the prediction . Virtual screening studies typically involve docking large chemical databases to one protein target , selecting compounds that score within the top 0 . 5–1% of the database and then further prioritizing them by visual examination . When experimentally validating these top candidates , a 5% hit rate can be considered a successful endeavor ( where a good hit is a predicted compound showing an experimental binding affinity in the µM or lower range ) [44] . Depending on the target , the crystal structure , the software used , post-docking criteria ( such as chemical clustering ) , and even the individual performing the visual examination , the hit rate can be improved to 10–40% ( Cavasotto et al . had 14% hit rate from 50 tested compounds [15]; Sabio et al . had a 36% hit rate from 56 tested compounds [45] ) . In our case , both the standard scoring threshold and the known-inhibitor score were not sufficient . With a normal score threshold of −30 , docking 4621 drugs against 252 proteins resulted in 104 , 625 predicted interactions . This is roughly 1% of the docked interactions , so even selecting the top 1% of the docking hits for validation becomes prohibitive for large-scale studies . It is important to note that each protein has different physiochemical properties: for some proteins , hundreds of compounds pass the −30 cut-off , while for other proteins none pass . Thus , using the known-inhibitor score as a cut-off allows for a threshold that is tailored to each protein . However , this method still predicted ∼8000 interactions at the most stringent . Our consensus threshold allowed us to pick the top 1% ( or any x% ) of docked compounds with the best icm- and pmf- scores for each protein and further filter from there . Through testing many combinations , we found that using the consensus score with rank information allowed us the highest PPV – nearly 50% - and enrichment factor – 50 times better than standard −30 score threshold and 490 times better than random selection . This high enrichment for known interactions suggests that many of the other predictions that have not yet been experimentally tested may be true binding interactions . There are limitations to this scoring scheme . Since the pmf-score is a statistical score comparing the docked interaction to known interactions in PDB , a chemical with a different scaffold or novel binding conformation may have a poor pmf-score and become predicted as a false negative . However , our foremost goal in this study was to eliminate as many false positive predictions as possible and obtain a high enrichment of true positives in our predicted interaction set . Thus , it was acceptable to miss some false negative predictions . In addition , the consensus score is quite simple with a linear separation method , and may not be as informative as a machine-learning algorithm that trains on known ligand docking scores . However , we desired an automated scoring method that did not depend upon the existence of known ligands . That is , if a protein structure had just one , or no known binders , our method would still be able to select the top 1% of docking hits . To date , cross-docking of proteins to compounds has generally been used for small datasets . As an example , Huang et al . docked 40 targets against 40 compounds to check whether their docking method could distinguish between a target's cognate ligands and the other targets' cognate ligands [19] . In this large-scale cross-docking study , our use of a 1000-processor cluster was essential to completing the tens of millions of docking simulations in a timely manner . In addition , the large number of crystal structures and binding pockets involved required much of the docking pipeline be automated . High-throughput computational screening of drug-target interactions represents a parallel approach to high-throughput experimental screening . Due to differences in experimental methods , assay settings , and protein panels , different studies may present differing results . For example , small molecule affinity purification methods that use whole cell lysates would give different results from in vitro kinase assays that use a specific panel of proteins . In the case of gefitinib , two such studies had distinct differences in their proposed cellular targets [22] , [41] . Differences in methods are also further compared in a study by Manley et al [23] . We presented an example for BIM-8 , which binds to PDPK1 differently in two similar in vitro experiments . For MAPK14 , the experimental results for nilotinib also varied . We experimentally tested two purchasable approved drugs against MAPK14 and found that nilotinib was a strong nanomolar inhibitor , and zafirlukast was also an inhibitor , though not as potent . Thus , interactions that are predicted to be very likely inhibitors computationally may merit extra study even if experimental tests are initially negative . In short , we have developed a computational pipeline that can run large-scale cross-docking of compounds to targets . We developed stringent criteria to filter a large proportion of false positive interactions . The two case studies presented were selected based on known experimental binding assay data , so as to demonstrate the notable enrichment of known interactions using our scoring and ranking criteria . We hypothesized that predicting a set of interactions with a higher PPV ( enrichment of known interactions ) would also lend confidence to the other novel interactions in the set . This appears to have worked , as we were able to find validation for 31 predicted drug-target interactions that were not previously annotated in DrugBank , as well as validate two other inhibitors of MAPK14 . Other drug-target interaction predictions are currently undergoing experimental validation; novel interactions discovered are potential drug repositioning candidates , but also provide insight into a drug's mechanism of action and adverse effects profile .
We downloaded the DrugBank 2 . 5 database [46] , containing drug information and comprehensive information of their targets . We extracted human protein drug targets from DrugBank and retrieved their sequences from SwissProt [47] . Protein Data Bank structures showing at least 95% sequence identity for proteins at least 20 amino acids in size were downloaded . They were required to be X-ray crystal structures with a minimum resolution of 2 . 8 Å . Multiple chains were grouped into a set of non-redundant sequences , based on PDB's chain redundancy analysis at the 95% sequence identity level . We prepared protein structures for docking using Molsoft's ICM software version 3 . 4-9c [13] , removing water molecules , solvent ions , and other ligands from the structures . We added hydrogen atoms to the structures then optimized their positions . These prepared protein structure files can be downloaded from http://www . bcgsc . ca/downloads/yli/ . To predict pockets , or potential binding sites , we used the PocketFinder [48] method in ICM , which calculates a transformation of the van der Waals energy for an aliphatic carbon probe on a grid map . For each protein , the three largest pockets are retained in the database . If metal ions were found near a pocket , we prepared two receptors for docking , one of the protein with the metal ion and one without . The receptor was defined as the box 3 . 5 Å surrounding the pocket . If the pocket overlapped well with the ligand but the ligand extended out of the protein structure , we defined the receptor be the box 3 . 5 Å around the pocket but also including 2 . 0 Å around the ligand . This ensured that known ligand binding sites not predicted by our automated method were also included in our pocket database . We docked drugs to target receptors using the ICM virtual library screening ( VLS ) module . This method performs rigid-receptor flexible-ligand docking using a two-step Monte Carlo minimization method and energy scoring function to sample ligand conformations and select the best docking hits . MMFF partial charges and ECEPP/3 force-field parameters are used . Docking one interaction required on average 30 seconds to 1 min per processor . A given protein may have several structures , each of which with more than one pocket; in such cases we dock all pockets to a drug , and the best scoring interaction is selected to be the representative protein-drug score . To ensure a sufficient coverage of the docking energy landscape , we docked each drug-target interaction 10 times in the known docking analysis and 5 times in the large-scale cross-docking analysis . Docking was performed on a Linux cluster with 1000 processors – this level of throughput allowed us to complete 1–3 million dockings per day . 8867 known interactions between human protein targets and drugs were culled from the DrugBank Drugcards database . Of these , 3570 interactions with protein target crystal structures present in our database were docked . Due to the Monte-Carlo nature of the ICM method , each interaction was docked 10 times to better cover the docking energy landscape . After 10 iterations , the best scoring prediction was retained . If the protein structure was solved in complex with a ligand , a Tanimoto coefficient was used to determine if the docked drug was similar to the complexed ligand . A coefficient less than 0 . 54 represented similar molecules [49] , and thus cognate dockings . Evaluation of static RMSD values of protein-drug interactions representing 380 cognate interaction dockings was performed on a case-by-case basis as the chemical numbering of PDB heteroatoms and docked structures often differed , which caused incorrect RMSD calculations . Each RMSD comparison was required to match at least 30% of the docked ligand atoms to the cognate crystal-structure ligand . 320 interactions pass this requirement , of which 221 ( 69% ) showed RMSDs under 2 Å . The other 99 ( 31% ) had RMSDs larger than 2 Å . Cytoscape [50] was used to generate the known drug-target interaction map . Networks were fitted to a force-directed layout and manually edited for improved visibility . Drugs and protein targets are nodes in the network , interconnected by interaction edges . The edge lengths were not weighted , and are adjusted for maximum visible understanding . We applied several methods of score thresholding: applying cut-offs of the ICM docking score ranging from [−25 to −100]; applying cut-offs of the ICM potential of mean force score ranging from [−80 to −200]; applying a drug rank cut-off ranging from [1 to 4500]; applying a protein rank cut-off ranging from [1 to 252]; applying a combined docking score and mean force score cut-offs . For the consensus score thresholds , all slopes ( from −1 to −40 ) and intercept ( from 0 to −400 ) combinations were tested . For each line , we calculated the density of the points eliminated in a trapezoidal area delineated by the consensus line , the best icm- score for this protein , and the best pmf-score for this protein , the midpoint between the worst icm-score and its mean , and the midpoint between the worst pmf-score and its mean . For two consensus thresholds that predicted the same number of interactions , we used the one that eliminated a denser cloud of points . While evaluating PPV for combination thresholds , it was often observed that two sets of thresholds resulted in the same number of predicted interactions but different PPVs . In such cases , we considered only the threshold combination that gave us the higher PPV . 1 , 164 , 492 interactions between 252 proteins and 4621 drugs were docked using ICM . Though there were actually 4854 drugs small molecules , some were excluded being too small or too large for docking ( molecular weight under 100 or over 1000 g/mol ) . Due to the multiple binding pockets per protein and multiple crystal structures per protein , there were a total of 2923 binding pockets . Each interaction was docked 5 times to better cover the docking energy landscape and the best scoring conformation was retained . Overall there were 2923×4621×5 dockings or 68 million docking calculations . The icm and pmf scores of each interaction were gathered into large matrices for further analysis . Protein inhibition assays were performed by SignalChem ( Richmond , BC , Canada ) . Kinases assays consisted of 33P-ATP at 25 µM , the protein kinase , peptide substrate , assay buffer , and the drug . Blank assays without substrate or drug , and assays without the drug , were used as controls . Staurosporine at 1 µM was used as the positive control drug .
|
Most drugs are designed to bind to and inhibit the function of a disease target protein . However , drugs are often able to bind to ‘off-target’ proteins due to similarities in the protein binding sites . If an off-target is known to be involved in another disease , then the drug has potential to treat the second disease . This repositioning strategy is an alternate and efficient approach to drug discovery , as the clinical and toxicity histories of existing drugs can greatly reduce drug development cost and time . We present here a large-scale computational approach that simulates three-dimensional binding between existing drugs and target proteins to predict novel drug-target interactions . Our method focuses on removing false predictions , using annotated ‘known’ interactions , scoring and ranking thresholds . 31 of our top novel drug-target predictions were validated through literature search , and demonstrated the utility of our method . We were also able to identify the cancer drug nilotinib as a potent inhibitor of MAPK14 , a target in inflammatory diseases , which suggests a potential use for the drug in treating rheumatoid arthritis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"molecular",
"mechanics",
"biochemistry",
"small",
"molecules",
"chemistry",
"biology",
"computational",
"chemistry",
"drug",
"discovery"
] |
2011
|
A Computational Approach to Finding Novel Targets for Existing Drugs
|
The influenza virus uses the hemagglutinin ( HA ) and neuraminidase ( NA ) glycoproteins to interact with and infect host cells . While biochemical and microscopic methods allow examination of the early steps in flu infection , the genesis of progeny virions has been more difficult to follow , mainly because of difficulties inherent in fluorescent labeling of flu proteins in a manner compatible with live cell imaging . We here apply sortagging as a chemoenzymatic approach to label genetically modified but infectious flu and track the flu glycoproteins during the course of infection . This method cleanly distinguishes influenza glycoproteins from host glycoproteins and so can be used to assess the behavior of HA or NA biochemically and to observe the flu glycoproteins directly by live cell imaging .
Enveloped viruses are composed of elements produced by and recruited from the infected cell . The formation of new virus particles occurs either on intracellular membranes or at the plasma membrane . The assembly of a nascent virion requires the coalescence of the envelope ( glyco ) proteins embedded in a proper lipid environment , and the recruitment of matrix and nucleocapsid proteins together with the viral genome [1]–[3] . How apposition of envelope components and viral genomes are controlled as a means of ensuring production of infectious progeny is not well understood . The influenza virus particle contains a segmented , negative stranded RNA genome encoding 11 proteins , two of which the virus uses to interact with the host cell membrane [4] . Hemagglutinin ( HA ) , a type I transmembrane protein , binds to sialoglycoconjugates on the surface of the host cell and mediates entry of the viral particle [5] , [6] . HA also mediates fusion of the viral and host cell membranes to effectuate genome delivery to the cell to be infected [7] . Neuraminidase ( NA ) , a type II membrane protein , is a sialidase that assists in release of virions from the infected cell [8] . The inability to label either flu HA or NA in a manner that allows continuous monitoring of surface disposition , surface distribution , and release has hampered the study of flu particle biogenesis . The use of antibodies , while feasible in principle , requires their introduction as fluorophore-conjugates that would crosslink viral proteins unless used as monovalent F ( ab ) fragments . Moreover , this labeling method is indirect . Studies that address particle biogenesis have also mostly used fixed cells and by design have not addressed virus release in real time . Visualization of the influenza glycoproteins in living cells demands a method for site-specifically modifying HA and NA , at the exclusion of all host proteins inserted into the very same membrane . We know of no successful attempts to achieve this by genetically tagging the flu glycoproteins with fluorescent proteins or with other methods that yield visible HA or NA by covalent modification in the context of an infectious virus . We and others have developed a site-specific labeling method that exploits sortase transpeptidases found in gram positive bacteria [9] , [10] . These enzymes cleave the five amino acid LPXTG recognition sequence between the threonine and glycine residues , forming an acyl-enzyme intermediate that is resolved by nucleophilic attack by the N-terminus of an oligoglycine peptide , forming a new amide bond . This reaction is portable: upon incubation with recombinant sortase A , proteins that carry an LPXTG motif are readily labeled with oligoglycine-based probes bearing a broad range of functionalities [11] , [12] . The incoming nucleophile may carry any desired substituent for attachment , including fluorophores , biotin , lipids , or may even consist of other polypeptides- see [13] , for review . Here we report the creation of two influenza A/WSN/33 strains bearing the sortase cleavage site in the HA and NA proteins respectively . Infection of host cells with such strains allows site-specific labeling of HA or NA and allows us to observe the products of influenza infection in real time . We can thus visualize and examine biochemically the events that immediately precede viral release from the host cell surface , as well as the release of newly formed virus particles . The ability to execute sequential labeling reactions employing distinct tags allowed us to observe preferred sites from which virus particles are released .
While purified virus particles can be labeled with lipophilic dyes , the dequenching of which reports on fusion of the incoming viral envelope with target endosomal membranes [14] , [15] , the production of new virions is more difficult to visualize . Neither flu neuraminidase nor hemagglutinin tolerate fusion to fluorescent proteins or other modules that allow site-specific covalent attachment of fluorophores—attempts to do so are incompatible with virus production and assembly . For N-terminal fusions to NA , this is likely the result of failure to insert into the ER during its biosynthesis . The bulky GFP moiety , when fused to the C-terminus of NA , likely compromises its functional activity and oligomerization . For HA , the only viable option would be to place GFP at the C-terminus , but fusions as small as a ( His ) 6 epitope already impair virus assembly and fusogenic activity of HA [16] . Fusions to the N-terminus of mature HA have been reported [17] , but with the caveat that such fusions undergo significant proteolysis and yield substantial amounts of wild-type HA protein . The inability to specifically label the flu glycoproteins for biochemical and visual observation has hampered an analysis of the virus budding process . We devised a sortase-based labeling method to overcome at least some of these limitations , and our findings are likely to be more generally applicable to other viruses with problematic labeling characteristics when relying on fusion with fluorescent proteins . The sortase labeling ( sortagging ) method is particularly well-suited for labeling of type II membrane proteins , as has been done for CD40L [11] , CD74 and Dectin-1 [18] at the extracellularly exposed C-terminus . We installed an LPETG motif at the C-terminus of the A/WSN/33 NA protein , followed by an HA epitope tag ( this epitope is absent from the A/WSN/33 hemagglutinin ) ( Figure 1A ) . Because the portion distal to the cleavage site is lost upon sortagging , the presence of the epitope tag allows monitoring of material not accessible to the enzyme in intact cells . We used the 12 plasmid reverse genetics system to generate recombinant A/WSN/33 flu particles bearing this sortaggable NA construct [19] , and found that the resulting virus was infectious and indistinguishable from wild type virus in its ability to replicate in vitro ( Figure 1B ) . We refer to this strain as NA-Srt . HA is a type I membrane protein , synthesized as an HA0 precursor , which requires proteolytic cleavage by a trypsin-like activity to generate the disulfide-bonded HA1 and HA2 subunits . This cleavage exposes a key glycine residue at the N-terminus of HA2 that is essential for HA2 to retain its fusogenic activity [5] . We created a version of HA that allows trypsin cleavage in the loop that connects HA1 and HA2 , with concomitant exposure of the sortase recognition sequence ( placed immediately upstream of the trypsin cleavage site ) . Cleavage is likely to improve accessibility of the LPXTG motif , a requirement for efficient labeling [11] , [20] . We therefore generated a recombinant A/WSN/33 flu strain bearing this sortaggable HA , and found that this strain , too , was not attenuated in vitro ( Figure 1C ) . We refer to this strain as HA-Srt . The behavior of the NA-Srt virus was indistinguishable from that of the WSN parental strain when assessing virulence by monitoring weight loss in mice . Mice infected with sublethal doses of the HA-Srt virus also showed weight loss , albeit somewhat reduced compared to mice infected with wild-type virus . ( Figure 1D ) . We conclude that the installation of a sortase tag on either NA or on HA does not seriously impair virus assembly , virus release and infectivity in vitro and in vivo . We obtained NA-Srt virions by sedimentation from the supernatants of infected MDCK cells , and subjected this material to sortagging with a biotinylated probe , followed by detection of biotinylated material by immunoblotting . Only in the presence of sortase and probe did we detect specific labeling , accompanied by the loss of the HA epitope tag , as expected . Based on the intensity of the HA-positive materials recovered , we estimate that the labeling efficiency of HA in the sedimented virus is approximately 70–80% for the conditions used ( Figure 2A ) . This value is not atypical for sortase-mediated labeling reactions , which usually proceed to near-completion [11] , [21] . Using sortase we similarly installed TAMRA or Alexa647 dyes on intact virions pelleted from tissue culture supernatant or further purified through a sucrose gradient ( Figure 2A and Figure 2B ) . Dimerization of the NA-Srt protein incorporated into gradient-purified virions is unaffected by the LPETG tag ( Figure 2C ) We infected MDCK cells with the NA-Srt virus and at different times post infection , we subjected cells to sortagging with a biotinylated triglycine-based probe ( Figure 2D ) . We detected biotinylated , surface accessible NA by immunoblotting using streptavidin-HRP . We assayed for the unlabeled and cell-internal pools of NA by reactivity with an anti-HA epitope antibody . NA was first detectable at ∼1 hr post-infection and its levels peaked at ∼4 hrs , after which we observed no further increase . Sortase-mediated surface biotinylation of NA was detectable at 4 hrs post-infection and steadily increased over the duration of the experiment ( 7 hrs ) . Specificity of labeling is excellent: we observed no host cell proteins modified with the biotinylated probe . Cells infected with the HA-Srt virus can be similarly labeled . Glycosidase digestions confirm that sortase labels only the cell surface pool of flu glycoproteins ( Figure 2E ) , as follows . We labeled intact cells infected with either the HA-Srt or NA-Srt viruses with sortase using a biotinylated probe , followed by lysis and digestion with either Endoglycosidase H or PNGaseF . Immunoblotting showed that all of the biotinylated HA-Srt protein is partially EndoH-resistant , indicating successful traversal of the secretory pathway . Because mature NA and HA carry both complex-type and high mannose-type oligosaccharides [22] , [23] , resistance to digestion with EndoH is always partial , as seen by comparison with the PNGaseF digestion product . The entire unlabeled , anti-HA reactive pool of NA-Srt protein was fully EndoH-sensitive , however , as evident from a comparison with the PNGaseF digestion products . The fraction of NA-Srt inaccessible to sortagging is thus indeed composed of cell-internal NA-Srt protein . We conclude that only the cell surface pool of influenza glycoproteins , poised for incorporation into nascent virions , is labeled upon incubation of infected cells with sortase and a suitable probe . Having established the specificity of labeling of the sortagging method , we examined the biogenesis of flu virions and their release from infected cells through biochemical analysis and by live cell imaging . We labeled flu HA-Srt protein on the surface of infected and metabolically labeled cells by exposure to trypsin , followed by sortase-mediated installation of a single biotin at the C-terminus of HA1 . We performed a pulse-chase experiment to examine the kinetics of arrival of HA-Srt protein at the cell surface , and its subsequent release from the infected cell as assembled virions . We did not ascertain the presence of all subgenomic RNA fragments in the material released from the infected cell , as we have no means of testing whether individual particles carry a full complement of subgenomic RNAs , or whether the released materials contain substantial amounts of defective particles with incomplete sets of RNAs . However , our data are consistent with the release of HA-Srt into the medium corresponding to assembly and release of progeny virions ( see below ) . We initiated metabolic labeling with 35S labeled Cysteine/Methionine at ∼5 hours post-infection , a time when robust viral protein synthesis is ongoing . For some experiments , we infected cells at a low multiplicity of infection ( MOI ) ( Figure 3A ) , and started metabolic labeling at 14 hrs post-infection , when most of the cells are infected by progeny HA-Srt virus produced by the cells infected initially . In this setting , we observe a similar if not greater amount of labeling than for cells infected at a high MOI ( Figure 3D ) , indicating that the LPETG-tag is neither lost nor interferes with virus replication ( compare Figure 3B and Figure 3E ) . During the chase , we performed sortase labeling for 30 minutes at the indicated time points . We lysed the 35S-labeled sortagged cells and subjected them to affinity purification on a neutravidin-agarose matrix to recover the biotin-modified HA1 and associated proteins ( Figure 3B ) . At the 0 min time point of this experiment ( it includes a 30 minute incubation with sortase , during which intracellular transport of glycoproteins continues ) , no labeled HA1 is recovered , indicating that the newly synthesized pool of HA requires at least 30 minutes to reach the cell surface . We observe a steady increase in surface exposed ( sortase accessible ) HA-Srt protein as well as a minor fraction of associated , uncleaved HA0 . We attribute the presence of this HA0 to incomplete cleavage of the HA trimer by trypsin added to the medium . In this manner we recover -as part of a trimer- some HA0 devoid of biotin , along with the sortase modified , biotinylated HA . As expected , we do not observe biotinylated HA0 by streptavidin blot ( data not shown ) . We do observe a small amount of HA0 at the cell surface at the 0 min timepoint . This we attribute to a minor portion of labeled HA1 not detectable via autoradiography . The HA0 recovered at early time points is composed of both the mature HA0 and the high mannose intermediate ( Figure 3C ) . As expected , HA1 and HA2 are recovered together because of their covalent association , which persists after cleavage of HA0 . We examined the behavior of HA-Srt protein at later chase times ( Figure 3E and Figure 3F ) and again observed an increase in labeling , after which the amount of labeled HA-Srt protein decreases . By 10 hours of chase , less than 20% of the material that successfully reached the cell surface and is labeled by sortase is retrieved from the cells , indicating that most , but not all HA labeled in the course of the pulse is released from the cell , presumably as intact virions . As infected cells show clear signs of cytopathic effects at late time points post-infection , cellular functions required for virion assembly are likely to be compromised , thus preventing complete release of all viral products , a situation that likely applies in vivo as well . To determine the fate of surface-labeled HA-Srt , we subjected infected cells to metabolic 35S –Met/Cys pulse labeling , followed by a chase period of 2 hours to allow radioactive HA-Srt protein to accumulate on the cell surface ( Figure 4A ) . We labeled intact infected cells with sortase A and a biotinylated probe and recovered biotinylated HA-Srt protein from cell lysates as well as from the media ( Figure 4B and Figure 4C ) . We observe a gradual increase in HA-Srt released into the media over time , corresponding to the rate of loss of biotinylated HA-Srt from the cell surface ( Figure 4D bottom panel ) . However , we detect more released HA-Srt protein than is accounted for by the loss from the cell surface ( Figure 4D , top panel ) . We attribute this difference to the fact that biotinylated HA-Srt is tightly associated with non-biotinylated HA-Srt in intact virus particles , which are retrieved by the neutravidin-agarose matrix along with the sortagged fraction . Metabolically labeled HA-Srt does not bind non-specifically to this matrix , as virus-infected cells exposed to the biotinylated probe in the absence of added sortase do not show any signal . Does the decrease in cell-associated , biotinylated HA1 , accounted for by the appearance of biotinylated HA-Srt protein in the supernatant , correspond to the release of virus particles ? To demonstrate this , we recovered from tissue culture supernatants radiolabeled viral proteins not modified by sortase through adsorption of released virions onto chicken erythrocytes ( Figure 4E ) , and visualized the adsorbed materials by SDS-PAGE and autoradiography . Because binding of the virions to erythrocytes occurs via HA , any other protein recovered by low speed sedimentation of erythrocytes must be part of an adsorbed virus particle . Indeed , we detect the other viral proteins upon suitable exposure of the autoradiograms . These include polypeptides with the assigned molecular masses of the RNA polymerase subunits , as well as NP and M1 , all of them in quantities proportional to their methionine/cysteine content and to the reported copy numbers in intact virions [24] . We next compared the kinetics of HA accumulation in the media for the fractions recovered via Neutravidin-agarose or on chicken erythrocytes . Levels of total HA at the 1 hr time point were quantified and all other time points were normalized to these values ( Figure 4F ) . We observe indistinguishable kinetics for HA-Srt accumulation in the media , underscoring our conclusion that biotinylation does not affect budding of HA-Srt . Having established the specificity of the labeling reaction and the ability of the labeled flu glycoproteins to be incorporated into virions and released into the culture supernatant , we next visualized virus budding and release by labeling biotinylated , surface exposed HA-Srt protein with streptavidin-modified quantum dots . To examine the behavior of surface disposed HA-Srt and its release from infected cells using a similarly modifiable control protein as a reference , we generated an MDCK cell line stably transduced with CD154/CD40L , equipped with a sortase tag [11] . Like the NA-Srt and HA-Srt proteins , this molecule is readily labeled with biotin using sortase , yet should not be actively incorporated into nascent virions and so allows for a direct comparison with flu HA-Srt . Incubation of MDCK cells that display biotinylated HA-Srt or CD154 were readily labeled with quantum dots . We examined the fluorescence intensity as a function of time after labeling by cytofluorimetry of infected , labeled cells ( Figure 5A ) . Whereas the levels of fluorescence recorded for labeled CD154/CD40L were constant , those for labeled HA-Srt decreased exponentially over the first few hours of incubation . When we infected labeled CD154/CD40L-expressing cells with wild type WSN virus , we also observed constant staining intensity but at a lower level , presumably because host protein synthesis was much reduced in the virus-infected cells . Labeled CD154/CD40L is obviously not incorporated into budding virions . We labeled HA-Srt infected MDCK cells grown on glass coverslips with sortase A and biotinylated oligoglycine probe , followed by staining with streptavidin-functionalized quantum dots . Cells prepared in this manner were directly imaged by spinning disc confocal microscopy to observe the behavior of the HA-Srt protein during the course of virus production and release . The influenza glycoproteins are known to be inserted into host cell membrane regions referred to as lipid rafts , operationally defined as insoluble in non-ionic detergents . One key component of lipid rafts is the ganglioside , GM1 , the levels of which oscillate with cell cycle status [25] . We co-stained HA-Srt labeled cells with fluorescently labeled cholera toxin ( CTx ) , a marker for GM1 , and observed variable CTx staining between MDCK cells , presumably because of asynchronous growth and rapid internalization of CTx [20] . Uninfected MDCK cells show similar heterogeneous CTx staining , indicating that virus infection is not the cause of this variability ( data not shown ) . We observe co-localization of CTx with patches of quantum dot staining in HA-Srt infected cells ( Figure 5B ) , but in CD40L/CD154 labeled cells , quantum dot staining shows only partial overlap with CTx staining , reflecting broad distribution of labeled CD154 in the cell membrane ( Figure 5B ) . Where there are patches of quantum dot-stained HA-Srt , we see evidence of CTx colocalization , while for quantum dot labeled CD154 , this is not always the case . Of note , cells that do not stain with CTx nonetheless label perfectly well with streptavidin-modified Qdots ( and hence correspond to flu-infected cells ) in a pattern that is indistinguishable from that seen in the adjacent , CTx positive cells . Although GM1 is a raft component , the organization of the plasma membrane apparently does not require its presence in a CTx-reactive form for the organization of HA-Srt . We conducted a pulse labeling experiment using sortase to determine the fate of HA-Srt protein labeled with biotin and streptavidin modified Qdots , followed by a second round of site-specific labeling with an oligoglycine based Alexa fluor 488 probe [18] to distinguish this second pool of labeled material from the first round of labeled HA-Srt , at a later stage of maturation and virus budding . Infected cells were first labeled with sortase and a biotinylated probe , followed by staining with streptavidin quantum dots ( HA-Qdot ) . After 30 minutes , this first pulse was followed by a second round of labeling using sortase A and an oligoglycine-Alexa 488 probe ( HA-488 ) at 4°C to inhibit endocytosis of free dye ( Figure 6A , Figure S1 ) . When labeling with the Alexa488 probe is initiated directly after quantum dot staining , we observe clear colocalization of the Qdot signal with the Alexa488 signal for most of the patches , and this is reflected as an increase in the overlap coefficient of the Qdot signal with the Alexa488 signal over time ( Figure 6A and Figure 6B , Supporting Figure S1 ) . Given the near quantitative labeling we observe for HA at the cell surface , the Alexa488-labeled pool of HA must therefore correspond to HA molecules inserted at sites where HA , labeled with Qdots in the first round of sortagging , has coalesced . These insertion sites appear as discrete dots . When Alexa 488 staining is initiated at later time points , HA-488 is located not only in previously established HA-Qdot patches , but we observe the presence of an increasing number of new Alexa488 spots of HA-Srt outside of the Qdot patches . This results in a decrease in the overlap coefficient of the HA-488 signal with the Qdot signal over time ( Figure 6B ) . The majority of HA-Qdot remains co-localized with HA-488 over the course of the experiment ( Figure 6B ) , suggesting that new HA-Srt is continuously being exported to the same membrane patches during budding . However , while colocalization with HA488 persists , both the number of HA-Qdot patches as well as their intensity decreases over time when compared to HA-488 ( Figure 6A , Figure S1 ) . Although it is a formal possibility that budding may not have been completed during the 2–3 hour interval , these observations may also suggest that patches of HA-Srt on the cell surface serve as sites of multiple budding events . At the 3 hr timepoint , little of the initial HA-Qdot signal remains , paralleling exactly the decrease observed by flow cytometry ( Figure 5A ) . A substantial HA-Qdot signal remains and is found inside cells instead of at the cell surface at later timepoints ( Figure 6C ) . The release of HA-Qdot tagged virus in the confined environment of this tissue culture experiment unavoidably leads to adsorption and internalization of labeled virus by adjacent , uninfected as well as onto already infected cells . Labeling with Alexa-488 probe increases at every time point , consistent with continued output of HA-Srt . We next studied the behavior of HA-Srt at the cell surface in real time by timelapse imaging ( Figure 7A ) . We labeled HA-Srt with biotin probe and streptavidin-Qdots at 4 hours post infection and acquired images over the following 60 minutes . We see a clear disappearance of HA-Qdot from the infected cells over this time course ( Figure 7A ) . We do not see this loss of Qdot signal in CD154/CD40L-expressing cells labeled in the same fashion ( Figure 7B ) . This loss of Qdot signal is apparent when the total sum of pixel intensities is plotted ( Figure 7C ) . The initial increase seen in CD154/CD40L control cells is likely due to a slight movement of the cells which cannot be restricted in our system and indicates that the decrease in HA-Qdot signal seen must be a minimum estimate of HA-Qdot release . We also observe an increase in clustering of the Qdot patches over time in the CD154/CD40L control cells . This may also account for the increase in signal intensity , as clustering of quantum dots is known to increase the on-time of blinking Qdots [26] Timelapse movies show a similar decrease in Qdot signal for the infected cells relative to the CD154/CD40L-expressing control cells ( data not shown ) . We detect both an apparent decrease in the total number of patches as well as a decrease in the intensity . This is in agreement with the observations made by flow cytometry ( Figure 5A ) and the timecourse of pulse labeling with sortase ( Figure 6A , Figure S1 ) .
For viruses such as vaccinia virus , which tolerates GFP extensions in many of the proteins it encodes , direct visualization of viral replication has provided major insights into the interactions between host cell and pathogen [27] . For viruses such as flu that are refractory to labeling with fluorescent proteins , other methods are urgently needed . We here devised a system to observe the behavior of influenza glycoproteins in cells infected with a fully functional virus . Our approach leverages the sortase labeling technique through the generation of influenza viruses that carry the short sortase recognition sequences in their glycoproteins , resulting in minimally modified viral gene products in the context of an infectious particle . Upon infection of tissue culture cells , the engineered viruses show behavior identical to the parental strain in terms of infectivity , replication kinetics , and viral protein synthesis . Observation of the sortase-labeled glycoproteins thus reflects very well the behavior of their wild-type counterparts . By combining this labeling approach with live-cell imaging , we can monitor the behavior of the influenza glycoproteins in real time . We observe extensive colocalization of surface-disposed flu glycoproteins with lipid rafts , as inferred from staining with CTx . In MDCK cells , flu HA and NA are enriched in lipid rafts based on their transient insolubility in cold TX100 and on colocalization with CTx [28] . Several aspects deserve emphasis . First , the specificity of the reaction and its reliance on an active enzyme limit all labeling to proteins that bear the sortase recognition sequence . We have not detected spurious incorporation of label into proteins not specifically designed to serve as sortase substrates . This applies not only to our earlier results with CD154/CD40L [11] , dectin-1 and CD74 [18] , but is extended here to flu NA and HA . The requirement for proteolytic conversion of HA0 into HA1 and HA2 suggested the possibility of installing a sortase motif upstream of the trypsin cleavage site without affecting the folding or fusogenic activity of the HA1-HA2 heterodimer , and our results validate this approach . Second , the glycosidase digestions performed on HA and NA unequivocally demonstrate the selectivity of the sortase reaction for fully mature surface-disposed proteins . In this regard , the method compares favorably in terms of ease and specificity with other chemical ( iodination , biotinylation ) or enzymatic ( lactoperoxidase-catalyzed iodination ) surface labeling methods , whose products require additional purification steps to enable analysis of the protein ( s ) of interest . Third , the size of the substituents introduced is minimal compared to that of a fluorescent protein such as eGFP or other enzymatically active modules used for site-specific covalent modification [29] . Given the failure to construct infectious flu when incorporating GFP into any of its structural proteins including NA and HA , the small size of our probes and the choices of label available have enabled , for the first time , the visualization of flu release from the surface of infected cells in real time . Fourth , it is possible to perform sequential labeling experiments and install labels that distinguish between each pool of labeled protein , and so independently monitor each pool of labeled product . This allowed us to generate a starting population of HA-positive cells labeled with quantum dots , from which we generated , at set intervals , a second population of cells with a distinct label installed on HA . Using this approach we demonstrate that HA-Srt , labeled in a first round of sortagging , identifies patches into which HA , tagged in a subsequent round of labeling , is inserted . Although it is possible that budding may not have been completed within this time interval ( 2–3 hours ) , an alternative interpretation is the existence of specialized sites that serve as a platform for coalescence of viral glycoproteins ( a typical flu virion requires some 400 HA and 30 NA molecules ) . Whether the identity of such patches corresponds to lipid rafts or specializations of lipid rafts is not clear , but the ability to visualize such sites opens the possibility of identifying host factors that control the construction of these sites as launching pads for new virions . We anticipate that this system will yield a robust method to visualize the kinetics of particle formation and , in combination with perturbations of host cells , will reveal host proteins that contribute to the process of influenza virion biogenesis .
All animal protocols were conducted in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . All animals were maintained according to the guidelines of the MIT Committee on Animal Care ( CAC ) . These studies were approved by the MIT CAC ( protocol #1011-123-14 ) . All infections were performed under avertin anesthesia , and all efforts were made to minimize suffering . Sortase was produced as described ( Popp et al . , 2007 ) . Peptide probes were produced as described [18] . Mutant viruses were generated by reverse genetics using plasmids as described [19] . The hemagglutinin and neuraminidase plasmids were modified by standard molecular biology techniques to carry the sortase cleavage site . All viruses , including the wild-type WSN virus used were rescued as described [19] . Nucleotide and protein sequences for modified portions of flu glycoproteins are below . Protein Sequence ( Trypsin cleavage site is in bold ) …CPKYVRSTKLRMVTGLRNIPSIQYRGLFGAIAGFIEGGWTGMIDGWYGYHHQNEQGSGYAA… Nucleotide Sequence ( Trypsin cleavage site is in bold ) …tgcccaaaatatgtcaggagtaccaaattgaggatggttacaggactaagaaacatcccatccattcaatacagaggtctatttggagccattgctggttttattgaggggggatggactggaatgatagatggatggtatggttatcatcatcagaatgaacagggatcaggctatgcagcg… Protein Sequence ( Sortase recognition site is in italics , trypsin cleavage site is in bold ) …CPKYVRSTKLRMVTGLRNIPSIQYLPETGGRGLFGAIAGFIEGGWTGMIDGWYGYHHQNEQGSGYAA… Nucleotide Sequence ( Sortase recognition site is in italics , trypsin cleavage site is in bold ) …tgcccaaaatatgtcaggagtaccaaattgaggatggttacaggactaagaaacatcccatccattcaatacctgcccgagaccggcggcagaggtctatttggagccattgctggttttattgaggggggatggactggaatgatagatggatggtatggttatcatcatcagaatgaacagggatcaggctatgcagcg… Protein Sequence …SGSIISFCGVNGDTVDWSWPDGAELPFTIDK- Nucleotide Sequence agtgggagcatcatttctttttgtggtgtgaatggtgatactgtagattggtcttggccagacggtgctgagttgccgttcaccattgacaagtag Protein Sequence ( Sortase recognition site is in italics , HA epitope is in bold ) …SGSIISFCGVNGDTVDWSWPDGAELPFTIDKGGGGSLPETGGYPYDVPDYA- Nucleotide Sequence ( Sortase recognition site is in italics , HA epitope is in bold ) agtgggagcatcatttctttttgtggtgtgaatggtgatactgtagattggtcttggccagacggtgctgagctcccgttcaccattgacaagggcgggggcggatcccttcctgaaactggtggatacccatacgatgttccagattacgcttag CD154/CD40L bearing an LPETG tag [11] was cloned into pLHCX and used to make retrovirus as described [30] . Retrovirus was used to infect MDCK cells as described [30] and cells were selected in 250 µg/ml Hygromycin B . Viral titer was assessed by plaque assay on MDCK cells as described [28] . For multi-step replication assays , MDBK cells were infected at an MOI of 0 . 001 and incubated for the indicated times in viral growth medium ( VGM , DME with 0 . 3% BSA ) supplemented with 0 . 5 µg/ml TPCK-treated trypsin ( NA-Srt ) or 1 µg/ml TPCK treated trypsin ( HA-Srt ) . For NA-Srt multi-step replication , 100 µl of media was plaqued at the indicated time points on MDCK cells grown with 0 . 5 µg/ml TPCK-trypsin . For HA-Srt , HA-Srt and wild-type WSN virus ( n = 3 ) were used to infected MDCK monolayers at an MOI = 0 . 001 and viral supernatant was analyzed via standard hemagglutination assays . Viral particles from tissue culture supernatant were concentrated by pelleting through a 20% sucrose cushion ( Sigma ) at 25000 rpm in an SW-28 rotor for 120 minutes . Where indicated , virus was further purified through a continuous 15%–60% sucrose gradient , centrifuged at 30 , 200 rpm in an SW-40 . 1 rotor for 3 hours . For biotin and TAMRA labeling , virions were pelleted from tissue culture supernatant without a sucrose cushion . The pellet was resuspended in 1× sortase buffer and labeled with 150 µM sortase A/5 mM probe for 1 hour at 37°C . For Alexa647 probe labeling , sucrose gradient purified virions were mixed with 200 µM sortase A and 500 µM probe at 37°C for 2 hours . MDCK cells were plated in a 24 well dish at 70% confluency the night before the experiment . Cells were infected at an MOI of 1 and 0 , 1 , 2 , 3 , 4 , 5 , 6 or 7 hours post infection , cells were incubated with 100 mM sortase A and 100 mM G5K-biotin probe [11] in VGM for 30 minutes at 37°C . Cells were washed extensively in PBS , collected , and lysed in 1% SDS with protease inhibitor cocktail ( Roche ) . A BCA assay was performed ( Pierce ) and 20 µg of lysate was loaded for western blotting . Mice ( n = 4 in each group ) were inoculated intranasally with 40000 pfu of the indicated virus and body weight was monitored at the indicated intervals . Balb/C mice and B6129SF2/J mice were purchased from the Jackson Laboratory ( stock# 000651 and 101045 resp ) . Mice were anesthetized with Avertin and infected intranasally with 40 . 000 PFU WT , HA-Srt or NA-Srt virus . Infection was followed by daily monitoring of weight loss and animals were euthanized with C02 when weight loss exceeded 20% of initial body weight . MDCK cells were infected at an MOI of 0 . 4 overnight and labeled for 1 hour at 37°C with 100 mM sortase and 500 mM biotin probe . Cells were then lysed in glycoprotein denaturing buffer ( New England Biolabs ) and total protein in lysates were quantiatated by BCA assay ( Pierce ) . Five micrograms of cell lysate was digested with either PNGase F or EndoH according to manufacturer's directions ( New England Biolabs ) , resolved by 12 . 5% SDS-PAGE , transferred to nitrocellulose , and used for western blotting with the indicated antibodies . MDCK cells were grown in 6-well tissue culture dishes and infected with HA-Srt at either an MOI 0 . 05 for 14 hrs or MOI 0 . 5 for 4 . 5 hrs as indicated in figure legends . Cells were starved with methionine- and cysteine-free DMEM for 45 minutes at 37°C followed by a 20 minute pulse labeling with [S35]Cysteine/Methionine ( perkin elmer ) at 0 . 77 mCi/ml in methionine- and cysteine-free DMEM . Chase was initiated by addition of VGM supplemented with 1 mM methionine , 0 . 2 mM cysteine and 1 µg/ml TPCK treated trypsin . At indicated timepoints during chase , cell surface HA molecules were labeled with 0 . 25 mM G3K-Biotin and 0 . 1 mM SrtAureus for 30 minutes at 37° . MDCK cells or CD154 control cells were cultured o/n in 24-well tissueculture dishes . Cells were infected with HA-Srt or WT virus at an MOI of 1 . At 4 hrs post-infection , cell surface molecules were labeled by addition of 0 . 1 mM Srtaureus and 0 . 25 mM G3K-Biotin for 30 minutes at 37°C . Cells were washed in PBS and incubated for 5 minutes with either 20 nM qdot655 ( MDCK cells and CD154 cells infected with WT-WSN ) or 10 nM qdot655 ( CD154 cells ) . Cells were incubated at 37°C and at indicated timepoints collected via addition of trypsin and kept on ice . Cells were analyzed immediately on a FACSCalibur flow cytometer ( BD Biosciences ) and FlowJo software . MDCK cells or CD154/CD40L expressing MDCK cells were cultured in 35 mm glass bottom dishes ( MatTek Corporation ) and infected with HA-Srt or WT virus at an MOI of 0 . 35–0 . 5 for 4 hours . Cell surface molecules bearing the LPETG motif were labeled by incubation with 0 . 1 mM SrtAaureus and 0 . 25 mM G3K-biotin for 30 minutes at 37°C . Cells were washed and incubated with 20 nM ( MDCK ) or 10 nM ( CD154 ) qdot655 ( Invitrogen ) for 5 minutes . After extensive washing , cells were incubated with VGM containing 1 µg/ml trypsin-TPCK . For surface labeling with ( G ) 3K-Alexafluor 488 probe , cells were incubated with 0 . 1 mM SrtAaureus and 20 nM probe for 1 hr at 4°C to inhibit non specific endocytosis of free dye . For staining of lipid rafts , cells were stained with 20 µg/ml Alexa fluor 594 conjugated choleratoxinB dye ( invitrogen ) for 5 minutes at room temperature immediately following Qdot labeling and imaging performed directly after . Images were acquired using an Andor Revolution spinning disk system with Yokogawa CSU-X1 spinning disk head , Andor iXon+ EM-CCD camera , 488 nm diode laser for excitation , emission discrimination with an emission filterwheel , Piezo Z100 z-stage on a Nikon Ti-E motorized microscope stand with a 100× 1 . 49NA Plan Apochromatic objective all controlled with the Andor iQ2 software ( version 2 . 0 ) . Temperature , CO2 and humidity was controlled with a LiveCell stage-top incubation system ( Pathology Devices ) . Analysis was performed using Imaris , Volocity and ImageJ software . For colocalization analysis , background correction was applied using ImageJ background correction with a rolling ball radius of 20 pixels . Images were further analyzed with Volocity colocalization analysis software . Background threshold was manually set using a background ROI to correct side effects of possibly remaining background pixels . Imaris software was used for analysis of timecourse Z-stack series . Brightness of images was adjusted using the linear stretch algorithm with the maximum set to 55 . Background was corrected for using a 17 µm filter width . Imaris software was used to create 3D images as well as the quantification of pixel intensities .
|
Enveloped viruses such as the influenza virus cause significant disease in humans . In order to cause a productive infection , the virus particle must interact with the host cell using the viral proteins encoded within its genome . For many such viruses , it is possible to directly observe the early steps in infection , yet for technical reasons it has been extremely difficult to study the genesis of daughter virions as they bud off of infected host cells . Here we devised a chemoenzymatic labeling strategy to site-specifically append probes to the influenza hemagglutinin ( HA ) and neuraminidase ( NA ) proteins using the bacterial sortase A enzyme . Because labeling is confined to surface exposed HA and NA in the context of live , infected cells , it is possible to study budding biochemically and microscopically in real-time . Using this system , we can observe budding of flu virions from discrete sites at the cell surface . Our work will enable detailed investigation into the birth of viruses from infected host cells and can likely be applied to viruses other than influenza that have been similarly resistant to real-time microscopic observation during budding .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biotechnology",
"biochemistry",
"chemical",
"biology",
"immunology",
"biology",
"microbiology",
"chemistry"
] |
2012
|
Chemoenzymatic Site-Specific Labeling of Influenza Glycoproteins as a Tool to Observe Virus Budding in Real Time
|
Previous studies have demonstrated the role of volatile organic compounds ( VOCs ) produced by skin microbiota in the attraction of mosquitoes to humans . Recently , behavioral experiments confirmed the importance of VOCs released by skin microbiota in the attraction of Rhodnius prolixus ( Hemiptera: Triatominae ) , a vector of Chagas disease . In this study , we screened for VOCs released in vitro by bacteria isolated from human facial skin that were able to elicit behavioral responses in R . prolixus . The VOCs released in vitro by eight bacterial species during two growth phases were tested with adult Rhodnius prolixus insects using a dual-choice “T”-shaped olfactometer . In addition , the VOCs released by the bacteria were analyzed with headspace solid-phase microextraction gas chromatography-mass spectrometry ( HS-SPME-GC-MS ) . The VOCs produced by Staphylococcus capitis 11C , Staphylococcus warneri and Staphylococcus epidermidis 1 were attractive to R . prolixus , while the VOCs released by Citrobacter koseri 6P , Brevibacterium epidermidis and Micrococcus luteus 23 were non-attractive . The results shown here indicate that VOCs released by bacteria isolated from human facial skin have a potential for biotechnological uses as a strategy to prevent the vectorial transmission of Chagas disease mediated by Rhodnius prolixus .
Human skin is colonized by a wide variety of beneficial microorganisms that inhibit the growth of pathogens and promote the processing of proteins and free fatty acids on the skin [1 , 2] . The genera most frequently reported to be present on human skin are Pseudomonas , Janthinobacterium ( phylum Proteobacteria ) , Corynebacterium , Kocuria , Microbacterium , Propionibacterium , Micrococcus ( phylum Actinobacteria ) , Staphylococcus , Clostridium ( phylum Firmicutes ) , and some species of the phyla Bacteroidetes , Cyanobacteria and Acidobacteria [1] . Woesearchaeota , Thaumarchaeota and Aenigmarchaeota are some Archaea also reported on human skin [3] . In addition , the human skin microbiota play an important role in the generation of human odors [2 , 4 , 5] . Human odor profiles include more than 350 identified compounds [6 , 7] , and it has been shown that bacteria on human skin are involved in the release of approximately 150 volatile organic compounds ( hereafter referred to as VOCs ) [8] . Some of these VOCs released by bacteria are typically found in human odor [2] . It has also been shown that VOCs released by bacteria are playing a role in attracting blood-sucking insects [5 , 9] . Differential attraction to different human body parts has been observed; for instance , Anopheles atroparvus are attracted to faces , and Anopheles gambiae are attracted to feet and ankles [10] . In addition , it has been shown that this differential attraction of Anopheles to humans is mediated by the VOCs produced by bacterial species and is dependent on the growth phase of the bacteria [10 , 11] . The combination of VOCs produced during the stationary growth phase of Corynebacterium minutissimum , Staphylococcus epidermidis , Brevibacterium epidermidis and Bacillus subtilis was attractive to An . gambiae in behavioral experiments [11] . Triatomines ( Hemiptera: Reduviidae ) are known as kissing bugs because of their preference to bite human faces [12]; triatomines have been widely studied because some species are involved in the vectorial transmission of Chagas disease , which is caused by the parasite Trypanosoma cruzi [13] . Chagas disease is a public health problem in 17 countries of Latin America [13] . In Colombia , the prevalence of Chagas disease is estimated to be approximately 5% [14] , and the main species of triatomines involved in the transmission of this disease are Rhodnius prolixus , Triatoma dimidiata , Triatoma maculata and Triatoma venosa [15] . The attraction of R . prolixus to vertebrates is due to factors such as CO2 , heat , humidity and chemical compounds released during sweating [16] . However , some years ago , it was found that R . prolixus is strongly attracted to odor extracts from human faces and feet [17] . In addition , the differential attraction to different body parts was apparently related to the presence of VOCs produced by bacteria [17 , 18] . Based on this preliminary evidence and keeping in mind the results obtained with Anopheles , we suggest that the attraction of R . prolixus to the human face could be a result of the VOCs produced by the bacterial microbiota present on facial skin . The aim of this study was to demonstrate the effect of VOCs produced in vitro by eight bacterial species ( in exponential and stationary phases of growth ) isolated from human skin on the attraction of Rhodnius prolixus using behavioral experiments . Finally , using gas chromatography coupled to mass spectrometry , we identified some of the VOCs produced in vitro by the bacterial strains tested behaviorally with R . prolixus .
Bacteria were isolated from the facial skin of ten volunteers living in Bogotá ( 5 men and 5 women between 20 and 35 years old ) following the method described in [8] . Briefly , an acrylic plate ( 1 . 0 cm thick , diameter 2 . 9 cm ) was placed on one of the cheeks of each volunteer . Then , 0 . 75 mL of full-strength buffer ( 75 mM sodium phosphate buffer [pH 7 . 9] + 0 . 1% ( v/v ) Triton X-100; Merck , The Netherlands ) was added , and the surface of the skin was scrubbed for 1 min with a glass stick . The fluid ( containing the skin bacteria ) was collected , and the procedure was repeated two more times on the same area of the face . Pooled samples were diluted 5X in half-strength buffer ( 75 mM sodium phosphate buffer [pH 7 . 9] + 0 . 05% ( v/v ) Triton X-100; Merck , The Netherlands ) . According with Colombian laws ( Resolución 8430–1993 ) , Ethics Committee from Universidad de los Andes approved the research project and categorized it as a non-risk project ( Acta 159–2012 ) . Oral consent to isolated bacteria from the human skin was obtained from volunteers after explaining the main objectives of the project and also that the process of isolation does not include neither reagents that are toxic or dangerous to health , nor invasive procedures . Finally the data were coded to ensure that they were analyzed anonymously . Serial dilutions of the pooled samples from each volunteer were cultured on Columbia blood agar ( Oxoid ) and incubated at 37°C for 24 h [8] . Additionally , the samples were cultured on a new medium for culturing Propionibacterium [19] and incubated at 37°C for 7 days in anaerobic jars with AnaeroGen Gas Pack 150 ( Oxoid ) . Several isolated bacterial morphotypes were transferred successively until pure cultures were obtained . The isolates were identified by amplification of the 16S gene using universal primers 27F ( AGAGTTTGATCCTGGCTCAG ) and 1492R ( GGTTACCTTGTTACGACTT ) , and the amplicons were sequenced by Macrogen , Korea . Sequences obtained were analyzed using BLASTn [20] and the Ribosomal Data Project ( RDP ) [21]; only sequences with a percent identity of 97% or higher were considered . Six bacterial strains were selected for use in behavioral experiments with Rhodnius prolixus ( S1 Table , S1 Fig ) : Staphylococcus epidermidis 1 , Staphylococcus caprae 7P , Staphylococcus capitis 11C , Citrobacter koseri 6P , Micrococcus luteus 23 and Dermacoccus nishinomiyaensis 9C . The following criteria were used to choose the strains to be tested: To have at least one strain of all species isolated , to avoid testing two strains isolated from the same volunteer , to include strains with the highest identity ( S1 Table ) , and finally , to use only those strains with the best in vitro growing . The bacterial strains Staphylococcus warneri and Brevibacterium epidermidis donated by Niels Verhulst were also included in the behavioral experiments . To identify the specific times at which each bacterium reaches the exponential or stationary phase , growth experiments were conducted as follows: For each bacterial strain , 12-h cultures in 4 mL of standard liquid medium ( Infusion from heart muscle 2 g/L; peptone 13 g/L; yeast extract 5 g/L; sodium chloride 5 g/L; agar bacteriological 15 g/L , all reactives from Oxoid and distilled water 1 , 000 mL ) [11] were incubated at 37°C , with shaking at 200 rpm . Then , 250 μL of each culture was transferred to a 125-mL Erlenmeyer flask with 25 mL of standard liquid medium . One-milliliter samples were extracted , and the absorbance of the samples at 620 nm was measured using a spectrophotometer ( ThermoSpectronic BIOMATE 3 ) ; serial dilutions were plated on standard solid medium and incubated at 37°C for 24 h . The absorbance measurements were performed until each of the bacterial strains reached the stationary growth phase . Absorbance and log ( CFU/mL ) were plotted as a function of time ( S2 Fig ) . The results of these growth curves allowed for the identification in each bacterial strain of the time at which half of the growth phase was reached . That time allowed us to carry out the behavioral experiments with 20 adults of Rhodnius prolixus on the olfactometer without to reach the next growing phase ( see below ) . Male and female Rhodnius prolixus were obtained from a breeding colony maintained at Universidad de los Andes since 1979 . Rearing conditions were as follows: 27 ± 2°C , 75 ± 10% relative humidity and an inverted circadian rhythm ( 6:00 h/18:00 h , dark/light; artificial illumination ) . The insects were fed monthly in vivo with hen's blood . Males and females with body weights of 0 . 0381 ± 0 . 002 g ( approximately 15 days of starvation ) were used to conduct the behavioral experiments [16] . Behavioral experiments were conducted with a dual-choice “T”-shaped olfactometer modified from Ortiz & Molina ( 2010 ) [17] and Verhulst et al . ( 2009 ) [8] . The modified olfactometer consisted of an air pump with an airflow <10 cm/s measured with a thermo-anemometer ( 45118 , EXTECH Instruments ) between the humidifier and a charcoal filter ( Fig 1 ) . As shown before , the airflow is constant in both sides of the olfactometer and under this conditions , airflow can be excluded as a confounded variable affecting our results [17] . The humidified air was introduced through plastic tubes in an acrylic box ( 8 . 0 x 8 . 0 x 5 . 5 cm ) that had two separated compartments . The base of the acrylic box was a galvanized metal plate with a regulated temperature of 34 ± 1°C , controlled with a pyrometer ( 0–400°C ) . Each of the two compartments of the acrylic box was connected with plastic tubes to one of the arms of the “T”-shaped olfactometer ( Fig 1 ) . All behavioral experiments were carried out as follows: The insect was placed at the entrance of the “T”-shaped olfactometer , and a maximum of five minutes was given to the animal to choose one compartment of the olfactometer ( Fig 1 ) . Each insect was tested just once in a dimly lit room ( light intensities averaged 0 . 3 μW/cm2 as measured with a radiometer/photometer ILT1400A; International Light Technologies Inc . ) between 7:00 h and 10:00 h , which corresponds to the early scotophase of R . prolixus with an inverted circadian rhythm . Then , a trial starts by testing individually each insect , and finish when the insect is removed from the olfactometer . After testing 60–68 insects we finished an experiment . Each experiment consisted then of three replicates with 20 insects , and all individual components of the olfactometer were washed at least three times with water and gel soap and rinsed under a stream of water to eliminate the soap completely after complete each replicate . The olfactometer components were allowed to dry overnight [17] . For each of the different experiments , the glass vial in the compartment of the acrylic box contained 1 mL of standard liquid medium with a bacterial strain in either exponential or stationary phase of growth . The time necessary to test 20 insects ( ≤ 100 min ) , grant in our behavioral experiments that each bacterial strain continued its growth but without reaching the next growing phase ( S2 Fig ) . For the next set of 20 insects , a new medium with bacteria of each species and growing phase was used . The glass vial in the second compartment of the acrylic box contained 1 mL of standard liquid medium without bacteria . After 20 trials , the compartments of the vials were switched to avoid any bias . We expected one of the following three results in our behavioral experiments with R . prolixus: Attraction ( when insects choose non-randomly the side with the medium + bacteria ) , non-attraction ( when insects choose non-randomly the side in the olfactometer with medium but without bacteria ) and none ( when insects choose randomly both sides of the olfactometer ) . As a control , we tested the effect of VOCs emitted by the standard liquid medium used to culture the bacteria on the attraction of R . prolixus . For this control , one compartment of the acrylic box contained 1 mL of the standard liquid medium in a sterile glass vial , and the other compartment of the acrylic box contained a sterile glass vial . The glass vial with the standard liquid medium was switched between the two compartments after each replicate . The control standard liquid medium was tested with 63 insects , divided in three replicates . For each replicate , the control with standard liquid medium was changed to avoid any bacterial growth due to contamination from the environment . In control experiments also one of the following three results is expected: Attraction ( when insects choose non-randomly the side with the medium ) , non-attraction ( when insects choose non-randomly the side in the olfactometer without medium ) and none ( when insects choose randomly both sides of the olfactometer ) . VOCs from the standard liquid medium were analyzed for each of the eight bacterial strains in exponential and stationary phases of growth . Three replicate experiments were conducted for each bacterial species at each growth phase . Each bacterial strain was cultured to initiate growth [11] and incubated at 37°C and 200 rpm for 18 h . Then , 250 μL of each culture was transferred to a 125-mL Erlenmeyer flask with 25 mL of standard liquid medium . One milliliter of the culture was extracted and its absorbance measured at 620 nm . When the bacteria reached the exponential ( 6 hours ) or stationary ( 12 hours ) phases of growth the VOCs produced were sampled with headspace SPME and subsequently analyzed by GC-MS . The fiber ( Supelco , Fiber SPME 85 μm CAR/PDMS , light blue ) was exposed for 15 min to the headspace of the Erlenmeyer flask maintained at 37°C ± 2°C in order to capture the VOCs released by the bacteria . Then , the fiber was transferred to the GC-MS ( Thermo , Trace 1300 gas chromatograph , ISQLT single quadrupole mass spectrometer ) . The GC-MS analysis was performed with a ZB-5 MSi GC column ( HP-Zebron 30 m x 0 . 25 mm ) . The separation of the VOCs was achieved by using the following temperature profile: the temperature was initially maintained at 40°C for 1 min; then , the temperature was increased at a rate of 6°C/min until it reached 60°C , and then , the temperature was increased at a rate of 3°C/min until it reached 200°C; finally , the temperature was increased at a rate of 10°C/min until it reached 250°C for 12 min . Splitless injection was performed using helium as the carrier gas at a constant flow rate of 1 . 0 mL/min . The mass spectrometer was operated in full-scan mode with a scan range of 35–300 m/z at a rate of 0 . 2 scans/s . The ion-source temperature was 250°C with an ionizing energy of 70 eV and a mass transfer line temperature of 280°C . The total number of insects that reached each one of the two compartments of the olfactometer was established at the end of each behavioral experiment and compared using binomial tests and a confidence interval of p < 0 . 05 [22] . All analyses were conducted using R software ( version 2 . 11 . 1 ) . To analyze the GC-MS data , the profiles obtained in three replicates for VOCs from eight bacterial species were analyzed using the Qual Browser of Xcalibur 3 . 0 software ( Thermo Scientific ) , and each peak with an S/N ratio over 3 was manually integrated . Only those peaks that were absent in the control analysis were integrated and considered to be volatile compounds released by bacteria . Compounds were annotated using NIST MS search 2 . 0 with the NIST 14 database . The GC-MS data were processed to estimate the percentage of the area of each compound in every sample . Using the percentages of each replicate , a matrix was constructed , which reported annotated volatile compounds as columns/variables and estimated percentages in each chromatogram as rows/observations . As grouping variables , we included the eight bacterial species in both growth phases and the corresponding R . prolixus behavioral response ( attraction/none/non-attraction ) . VOC profiles , including CO2 , were analyzed using non-metric multi-dimensional scaling ( NMDS ) based on the Bray-Curtis distance matrix in order to reduce the complex data of the area percentages into a two-dimensional space . The closer the samples are in the NMDS , the more similar their VOC profiles . The reliability of the NMDS representation was estimated based on stress values , usually considered acceptable at a value less than 0 . 2 . We used ANOSIM with 999 permutations in order to determine which grouping variable ( species , growth phase , or both ) better explained the differences in VOC profiles and to determine if this pattern could be related to the results obtained in the behavioral experiments with R . prolixus . We performed all statistical tests using R studio software ( http://www . rstudio . org/ ) and the community ecology package vegan ( http://CRAN . R-project . org/package=vegan ) . As the dataset contains many zeros , PCA analysis or NMDS using Euclidean distances were not appropriate because they can result in artifacts due to the lack of normal data distribution [23] .
Fifty-five bacterial morphotypes were isolated from the facial skin of ten volunteers , but only 36 were identified by 16S sequencing ( S1 Table ) . From these sequences , we identified seven species ( S1 Fig ) . The only species isolated from the facial skin of all volunteers was Staphylococcus epidermidis . Interestingly , it was only possible to isolate four different species from two of the ten volunteers ( S1 Fig ) . From the species identified , we selected six species for growth analysis ( S1 Fig ) . In addition , the donated strains Staphylococcus warneri and Brevibacterium epidermidis were included in the experiments ( see Materials and methods ) . As a general result , we can highlight that all the species evaluated showed similar growth curves and reached the stationary phase at 12–14 hours ( S2 Fig ) . The control test showed that the R . prolixus were not attracted to the standard liquid medium ( Fig 2 , p = 0 . 093 ) . In contrast , R . prolixus showed a clear attraction for the standard liquid medium containing three of the four Staphylococcus species evaluated ( S . capitis 11C and S . warneri in the exponential phase and S . epidermidis 1 in the stationary phase ) ( Fig 2 ) . On the other hand , Citrobacter koseri 6P , Microccocus luteus 23 and Brevibacterium epidermidis were non-attractive to R . prolixus in one or both growth phases ( Fig 2 ) . Finally , in terms of behavioral response , R . prolixus was indifferent to the VOCs produced by the following bacteria: S . caprae 7P ( exponential and stationary ) , S . epidermidis 1 ( exponential ) , S . capitis 11C ( stationary ) , S . warneri ( stationary ) , D . nishinomiyaensis 9C ( exponential and stationary ) , B . epidermidis ( stationary ) and M . luteus 23 ( exponential ) ( Fig 2 ) . Forty-six different VOCs were used to compare similarities between the percentage distributions among eight species of bacteria in two growth phases . We identified tentatively 34 VOCs by comparing experimental mass spectra with NIST14 Mass Spectral Library and analyzing the fragmentation patterns in EI . The annotation of functional groups was achieved for ten VOCs based on analysis of fragmentation patterns , but mass spectra similarity was insufficient to tentative identification . Annotation of each compound was carried out assigning retention times and spectral features based on mass spectra . Average area percentages for each species-phase combination are summarized in Table 1 organizing VOCs by retention times . C . koseri 6P is the species with the highest number of VOCs , followed by M . luteus 23 , while S . caprae 7P has the lowest number . The only two VOCs shared by all eight species were CO2 and indole , which showed differences in the average percentage area between species in exponential and stationary phases . An NMDS plot was obtained for the VOCs only released by bacteria ( Fig 3 ) . To ensure the outcome we established the following two criteria: ( 1 ) the VOC was absent in the control analysis; and ( 2 ) the mass spectrum of the annotated compound corresponded to a bacterial VOC . NMDS was accurate in spatially representing the similarity/dissimilarity indices; the stress value obtained was 0 . 1776 . To determine whether the bacterial species , the growth phase or both were the variables affecting the VOC profiles produced by skin bacteria , we considered all three variables in the NMDS analysis . Taking into consideration the bacterial community structure , the differential VOC profiles are better explained through the species-phase interaction ( ANOSIM , R = 0 . 7635 , p = 0 . 001 ) . Furthermore , when considered separately as independent variables , i . e . , species ( ANOSIM , R = 0 . 5111 , p = 0 . 001 ) and growth phase ( ANOSIM , R = 0 . 05446 , p = 0 . 035 ) , both variables were also statistically supported as being factors involved in the differential VOC profiles . Despite the overlapping showed between ellipses representing the CI 95% using t distribution for each species ( Fig 3 ) ; we found that this grouping variable explains at least 51% of the variance observed in the chemical profiles of the evaluated bacteria according to R2 estimations . Ellipses for the interaction species-phase and behavioral responses were not represented because at least four replicates ( instead or three ) per grouping variable are needed to perform this task . The VOC profiles of all Staphylococcus bacteria were found to be similar to each other with the exception of S . epidermidis , which , in the exponential phase , has a VOC profile more similar to bacteria of another genus . Although B . epidermidis belongs to a different genus of bacteria , it had a VOC profile similar to most of the Staphylococcus strains in the exponential phase . Having collected evidence of a differentiation in the VOC profiles between bacterial species and growth phases , we went on to determine if this differentiation also correlates with the results of the behavioral experiments conducted with R . prolixus . Using the same NMDS but plotting ellipses for the “attractive” , “none” and “non-attractive” categories based on the results obtained in the behavioral experiments ( Fig 4 ) , we found a statistically supported differentiation ( ANOSIM , R = 0 . 2407 , p = 0 . 002 ) despite overlaps in the VOC profiles of bacteria in the “none” category with those of bacteria in the other categories . The ellipses are representing CI 95% using t distribution for each behavioral response , and they explain 24% of the variance observed in the chemical profiles of the tested bacteria according to R2 estimations . The larger overlapping in the ellipses is highlighting the differences in the explaining power of each grouping variable: bacterial species vs . behavioral response; and for that reason , they are not strictly assigning behavioral responses to the bacterial species ( Fig 4 ) .
We show here that bacteria isolated from human facial skin produce in vitro VOCs that could mediate the interaction between Rhodnius prolixus and humans depending on the VOCs released by individual bacterial species and its growth phase . These results support the tripartite interaction R . prolixus-bacteria-humans that we have shown previously in our group , as playing potentially a role in the vectorial transmission of Chagas disease [17 , 18] . Molecular data have shown the presence of 205 identified genera of bacteria on human skin [24] . However , the diversity of bacteria revealed by culture-based methods is smaller compared to molecular methods [24] . We isolated 36 different morphotypes of bacteria from the skin of human volunteers , and seven species were identified ( See S1 Table and S1 Fig ) . From the species isolated , we discarded the isolate Propionibacterium acnes from our behavioral experiments due to the difficulty in obtaining a growth curve for this bacterial species and the difficulty in using these bacteria in further behavioral experiments with R . prolixus under low oxygen conditions . The other six bacteria isolated from the facial skin of our human volunteers are well-known as parts of the human skin microbiota [24 , 25] with the exception of Citrobacter koseri ( Proteobacteria: Enterobacteriaceae ) , which has been sporadically isolated from human skin and has also been reported as the cause of neonatal meningitis [26] . S . epidermidis was the only species isolated from all our volunteers; the identity and number of other species isolated for each volunteer exhibits high variation , as expected based on studies of skin microbiota among individuals [24 , 25] . Skin bacteria isolated from human volunteers produce odors involved in the interaction between vectors of tropical diseases and humans [11] . Here , we reported for the first time the effect of VOCs released in vitro by bacteria isolated from human facial skin on the behavior of Rhodnius prolixus . In addition , our multivariate analysis confirm the differentiation in profiles of the VOCs released in vitro by bacterial species-growth phase ( see Fig 3 ) as has been previously reported by Thorn et al . ( 2011 ) by using Selected Ion Flow Tube Mass Spectrometry ( SIFT-MS ) [27] . Our behavioral experiments showed Rhodnius prolixus to be attracted to VOCs released by S . epidermidis 1 , S . warneri and S . capitis 11C ( see Figs 2 and 4 ) , while non-attractive behaviors were observed with B . epidermidis , M . luteus 23 and C . koseri 6P ( see Figs 2 and 4 ) . According with our eight isolated bacteria , most of the Staphylococcus species are attractive to R . prolixus , while the other genera of bacteria tested here are non-attractive except for Dermacoccus nishinomiyaensis 9C ( see Fig 2 ) . Our results confirm that 24% of the variation found in the profiles of the VOCs can be explained by the behavioral responses according to ANOSIM analysis . Even though , non-attractive ellipse representing CI of 95% overlap with S . epidermidis 1-exponential and B . epidermidis-stationary , and attractant ellipse CI overlap with S . capitis 11C-stationary , D . nishinomiyaensis-both phases and S . caprae-stationary , which are classified as without effect on the behavioral analysis ( See Fig 4 ) . Contrary to our results , behavioral experiments carried out with An . gambiae showed that higher attractiveness can be obtained to VOCs released by genera of bacteria other than Staphylococcus [11]; in fact , no bacteria showed any degree of non-attraction when tested with An . gambiae [11] . No statistically significant effect was observed on the behavior of R . prolixus to VOCs released in both growing phases by D . nishinomiyaensis 9C and S . caprae 7P ( see Fig 2 ) . To date , behavioral experiments conducted on Anopheles gambiae , Aedes aegypti [9 , 11] and R . prolixus ( see Fig 2 ) highlight the role of the VOCs released by S . epidermidis in attracting blood-sucking insects . Additionally , it seems that VOCs released by the bacteria during the stationary phase of growth are affecting the attraction of An . gambiae [11] and R . prolixus ( see Fig 2 ) . From the other bacterial species isolated from human facial skin , only the VOCs released by B . epidermidis have been tested with An . gambiae [11] and R . prolixus ( see Fig 2 ) ; the results obtained highlight the differences between these two blood-sucking insects . B . epidermidis , in stationary phase , was highly attractive to An . gambiae [11] , while for R . prolixus , the exponential phase of the bacteria was non-attractive , and the stationary phase showed no behavioral response ( see Fig 2 ) . Bacillus spp . not tested in R . prolixus , have showed also in vitro attraction to Anopheles and Aedes mosquitoes [11 , 28] . Among the VOCs released in vitro by the bacteria isolated , indole and CO2 were the only two compounds produced by all the species tested here ( see S1 Table and S1 Fig ) . CO2 plays an important role in the attraction of blood-sucking insects [29] , including Triatominae [17 , 30] . However , Verhulst et al . ( 2009 ) [8] showed that the amount of CO2 released in vitro due to bacterial metabolism is not in the range required ( 1500–2300 ppm ) to induce behavioral responses in R . prolixus [17] . According to the literature in Anopheles , indole is a volatile compound released by the microbiota in the oviposition sites , and indole is a major component of sweat and breath in humans [31 , 32 , 33] . The importance of this volatile compound is highlighted by the fact that detection of this compound by the antennae of An . gambiae is mediated by the odorant binding protein 1 ( AgamOBP1 ) [34] . The role of indole in the behavior of R . prolixus is not known , and extrapolation of behavioral information obtained in experiments with mosquitoes should be considered with caution . Bernier et al . ( 2000 ) identified 303 VOCs from human skin [7] , from these , we observed that our bacteria released in vitro nine out of the 34 identified ( see Table 1 ) . In addition , Verhulst et al . ( 2009 ) identified 11 VOCs released in vitro by bacteria isolated from human skin [8] . In our bacteria we identified also three of these VOCs; however , none of them have been reported as VOCs released by human skin ( see Table 1 ) . Both results confirm that some of the VOCs released by our bacteria in vitro are normal constituents of the human skin emanations , and some others are product only of the in vitro metabolism . The results shown in Fig 4 suggest that behavioral responses in R . prolixus are highly dependent on the complex mixture of VOCs released by bacteria and not on single compounds . Something similar was shown with behavioral experiments for Anopheles [8] . In the case of R . prolixus we found that benzeneacetaldehyde , which is the only compound different from CO2 and indole shared by the three attractive phase-bacteria ( S . epidermidis 1-stationary , S . warneri-exponential and S . capitis 11C-exponential ) , can be found also in other Staphylococcus species but without behavioral responses ( see Table 1 ) . Something similar occurs also with dimethyl disulfide and phenol , which are shared by the four non-attractive bacteria-phase ( B . epidermidis-exponential , M . luteus 23-stationary and C . koseri 6P-both phases ) , but they are also part of the chemical profile of other bacteria-phase which in the behavioral analysis showed none , or even attraction ( see Table 1 ) . In addition , the concentration at which the bacteria produce a VOC in a mixture should have an effect on the behavioral response of R . prolixus . These findings warrant further analysis of specific VOCs to analyze their ecological function in R . prolixus , because the behavioral response to bacteria seems to be very complex and seems to be dependent on the simultaneous interaction of several VOCs . Possible synergistic effects [35] , acting agonistically or antagonistically , could explain these findings . The results presented here support the hypothesis that VOCs produced by bacteria from the human facial skin microbiota differentially influence the behavioral response of R . prolixus [17] . These findings highlight the potential of bacterial VOCs for biotechnological use . Future behavioral experiments to test the attraction and/or repellency of R . prolixus to mixtures of VOCs are recommended; identified mixtures can be used as baits or repellents to avoid Rhodnius prolixus-humans contact and reduce the risk of vectorial transmission of Chagas disease mediated by Rhodnius prolixus .
|
Volatile organic compounds released by bacteria growing on human skin are mediating human-blood-sucking insects interactions . In this study we were interested in those volatile organic compounds produced in vitro by skin bacterial metabolism that are involved in the Rhodnius prolixus interaction with humans . Rhodnius prolixus is the main vector of Chagas disease in Northern South America . Using behavioral experiments , chemical analysis and microbiological techniques we determined the volatile organic compounds released in vitro by bacteria isolated from human faces and its effects on Rhodnius prolixus . We found that most of the Staphylococcus species tested here are mediating attraction , while non-Staphylococcus species are non-attractive . As a conclusion , our results showed that volatile organic compounds released by bacteria isolated from human faces have a potential for biotechnological uses as a strategy to control the vectorial transmission of Chagas disease mediated by Rhodnius prolixus .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"skin",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"integumentary",
"system",
"chemical",
"compounds",
"pathology",
"and",
"laboratory",
"medicine",
"face",
"pathogens",
"microbiology",
"organic",
"compounds",
"animals",
"animal",
"models",
"animal",
"behavior",
"experimental",
"organism",
"systems",
"volatile",
"organic",
"compounds",
"insect",
"vectors",
"zoology",
"bacteria",
"bacterial",
"pathogens",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"staphylococcus",
"medical",
"microbiology",
"behavior",
"microbial",
"pathogens",
"chemistry",
"head",
"insects",
"disease",
"vectors",
"arthropoda",
"eukaryota",
"organic",
"chemistry",
"anatomy",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"species",
"interactions",
"rhodnius",
"organisms",
"staphylococcus",
"epidermidis"
] |
2018
|
Behavioral responses of Rhodnius prolixus to volatile organic compounds released in vitro by bacteria isolated from human facial skin
|
Plant microRNAs ( miRNA ) guide cleavage of target mRNAs by DICER-like proteins , thereby reducing mRNA abundance . Native precursor miRNAs can be redesigned to target RNAs of interest , and one application of such artificial microRNA ( amiRNA ) technology is to generate plants resistant to pathogenic viruses . Transgenic Arabidopsis plants expressing amiRNAs designed to target the genome of two unrelated viruses were resistant , in a highly specific manner , to the appropriate virus . Here , we pursued two different goals . First , we confirmed that the 21-nt target site of viral RNAs is both necessary and sufficient for resistance . Second , we studied the evolutionary stability of amiRNA-mediated resistance against a genetically plastic RNA virus , TuMV . To dissociate selective pressures acting upon protein function from those acting at the RNA level , we constructed a chimeric TuMV harboring a 21-nt , amiRNA target site in a non-essential region . In the first set of experiments designed to assess the likelihood of resistance breakdown , we explored the effect of single nucleotide mutation within the target 21-nt on the ability of mutant viruses to successfully infect amiRNA-expressing plants . We found non-equivalency of the target nucleotides , which can be divided into three categories depending on their impact in virus pathogenicity . In the second set of experiments , we investigated the evolution of the virus mutants in amiRNA-expressing plants . The most common outcome was the deletion of the target . However , when the 21-nt target was retained , viruses accumulated additional substitutions on it , further reducing the binding/cleavage ability of the amiRNA . The pattern of substitutions within the viral target was largely dominated by G to A and C to U transitions .
Plant miRNAs regulate the abundance of target mRNAs by guiding their cleavage at the sequence complementary region . Previous reports have shown that changes of several nucleotides within a miRNA 21-nt sequence do not affect its biogenesis and maturation [1] , [2] . This finding raises the possibility to redesign the miRNA sequence to target specific transcripts , originally not under miRNA control . Such artificial miRNAs have been produced in dicotyledonous [3]–[5] and monocotyledonous plants [6] using different pre-miRNAs as backbones . We have successfully demonstrated that redesigned artificial miRNAs ( amiRNAs ) are biologically active and can be used to confer specific virus resistance in transgenic plants [4] . The pre-miR159a precursor was used to generate two amiRNA159s ( amiR159-P69 and amiR159-HC-Pro ) with sequence complementary to the RNA genome of two plant viruses , Turnip yellow mosaic virus ( TYMV ) and Turnip mosaic virus ( TuMV ) , respectively . The amiR-P69 was designed to target sequences encoding the P69 suppressor of TYMV whilst amiR159-HC-Pro would target sequences for the HC-Pro silencing suppressor of TuMV . Transgenic lines carrying both 35S-pre-amiR159-P69 and 35S-pre-amiR159-HC-Pro transgenes can express the appropriate amiRNA at high levels and showed specific resistance to either TYMV and TuMV , depending on the expression of the cognate amiRNA [4] . Specific resistance to TuMV was also seen with plants expressing amiR159-TuCP directed against the TuMV coat protein ( CP ) gene [4] . In animal systems , RNA interference ( RNAi ) , a gene-silencing mechanism similar to that of miRNA , has been used in clinical trials as antiviral therapeutics to inhibit replication of several human pathogenic viruses ( reviewed in [7] , [8] ) . As demonstrated for HIV-1 , a major problem of RNAi-mediated antiviral therapies is the emergence of resistant virus variants , which differ from the wild type virus by having fixed point mutations in the target sequence leading to imperfect matching; these mutant viruses are not properly processed by the enzymatic silencing machinery [9]–[13] . Some mismatches within the target sequence are tolerated by the RNAi machinery whereas other mismatches , such as those in the central region ( position 9 to 11 ) of the target sequence , compromise RNAi-guided antiviral therapies [14] , [15] . However , all these studies suffer from the drawback of having a superimposition of two different selective forces: on the one hand , purifying selection acts at the protein level ( i . e . , the necessity of maintaining a functional protein ) and , on the other hand , diversifying selection acts at the RNA sequence level favoring mutant genomes capable of evading RNA silencing . Mallory et al ( 2004 ) have used an in vitro wheat germ system to assay for critical positions within a miRNA target site needed for efficient plant mRNA cleavage [16] . Analysis of scanning mutants revealed that mismatches at the center and the 3′ end of the miRNA are more tolerated compared to mismatches at the 5′ region [16] . Recently , the molecular mechanism of RISC-mediated RNA cleavage has been investigated by in vitro reconstitution assays using human RISC [17]–[19] . It was found that the accessibility of RNA target site correlates directly with the RNA cleavage efficiency , indicating that RISC is unable to unfold structured RNA . In the course of target recognition , RISC transiently contacts single-stranded RNA nonspecifically and promotes siRNA-target RNA annealing . Furthermore , the 5′ portion of the siRNA within RISC creates a thermodynamic threshold that determines the stable association of RISC and the target RNA . Furthermore , in addition to this clear position-effect , overall desestabilization of the double strand structure has little effect on RNAi activity until an energy threshold is reached , beyond of which a negative correlation exist between stability and RNAi-mediated inhibition [15] . Here , we first investigated whether the 21-nt of an amiRNA target site is both necessary and sufficient for amiRNA-mediated specific resistance . Second , we were interested in identifying critical positions within the target site for this resistance . Third , we have explored the patterns of sequence polymorphism of viral sequences that evolve under the only selective pressure of amiRNA-mediated silencing . To address these issues , we established a heterologous-virus resistance system using a TuMV-GFP viral vector to carry a non-essential 21-nt sequence of the P69 gene targeted by amiR159-P69 . This heterologous-virus system allows us to modify any nucleotide within the 21-nt target site without altering virus coding sequences and thus without affecting replication and activity . In other words , this heterologous system allows separating the selective pressure imposed by protein functionality from the selective pressure imposed at the sequence level by RNA silencing . The 21 scanning mutant viruses were inoculated on amiR159-P69 plants and the proportion of transgenic plants that became infected was used to determine the importance of the mutated nucleotide position within the amiRNA target site .
We have previously demonstrated that a 21-nt amiRNA , with sequence complementary to a viral sequence , can mediate cleavage of target viral RNA and confer resistance on transgenic plants [4] . However , it was not known whether the 21-nt viral target site , complementary to the amiRNA sequence , was sufficient for specific resistance . To this end , we constructed a green fluorescence protein ( GFP ) gene carrying a 21-nt sequence ( 5′-CCACAAGACAAUCGAGACUUU-3′ ) of the TYMV P69 gene at its 3′-end and inserted the GFP-P6921nt fusion gene in between the NIb and CP genes to generate a TuMV-GP69 chimeric virus ( Fig . 1 ) . As a control , we mutated 4 nts ( position 9 to 12 from the 3′-end; underline ) of the target 21-nt sequence ( 5′-CCACAAGACCUGAGAGACUUU-3′ ) to give GFP-P6921ntm , which was inserted in the same position of the viral genome to generate the TuMV-GP69m chimeric virus ( Figure 1D ) . The presence of two TuMV NIa protease cleavage sites ( CVYHQA ) at both the N- and C-termini of the GFP-P6921nt fusion protein allows the release of GFP plus a 7 amino acid C-terminal extension ( PQDNRDF ) from the TuMV-GFP viral polyprotein ( Figure 1C and 1D ) . Virus infection can be easily confirmed and followed by monitoring GFP signals from infected tissues . We have previously shown that transgenic Arabidopsis thaliana plants expressing amiR-P69 can specifically target the P69 gene of TYMV and displayed specific resistance to TYMV [4] although these plants remained susceptible to heterologous virus ( TuMV-GFP ) infection ( Figure 2A , top second panel ) . Figure 2 shows that insertion of the 21-nt sequence of the TYMV P69 gene into TuMV-GFP , which was targeted by amiR159-P69 , rendered these amiR159-P69 plants resistant to TuMV-GP69 ( Figure 2A , top third panel ) . Control experiments showed that the amiR159-P69 plants remained sensitive to TuMV-GP69m ( Figure 2A , top fourth panel ) , which carried 4 mutations in the central region of the 21-nt site of the P69 gene . Systemic leaves of amiR159-P69 plants displayed GFP fluorescence when inoculated with TuMV-GFP or TuMV-GP69m ( Figure 2B , top second and fourth panels ) , but no GFP signal was detected upon TuMV-GP69 inoculations ( Figure 2B , top third panel ) . Plants expressing amiR159-HC-Pro were resistant to chimeric TuMV-GFP , TuMV-GP69 and TuMV-GP69m and no GFP was detected on systemic leaves of inoculated plants ( Figure 2B , bottom panels ) . These results were expected since all these 3 chimeric viruses contained the HC-Pro gene targeted by amiR159-HC-Pro [4] . Next , we established the heterologous virus resistance system in N . benthamiana and tested amiRNA-mediated resistance efficiency . Figure 3A shows amiR159-P69 expression levels in 4 independent transgenic N . benthamiana lines ( #1 , 2 , 3 , and 4 ) . Progeny plants of these lines were challenged with TuMV-GFP , TuMV-GP69 or TuMV-GP69m . The GFP signal produced by infection with TuMV-GFP or TuMV-GP69m can be detected at 4 dpi ( early stage of symptom development ) ( Figure 3B ) . As expected , at 7 dpi , transgenic N . benthamiana plants expressing amiR159-P69 were resistant to TuMV-GP69 but susceptible to TuMV-GFP and TuMV-GP69m ( Figure 3C ) . Plants that were sensitive to virus infection showed severe wilting symptoms ( Figure 3C ) . These results , which were very similar to those obtained with A . thaliana transgenic plants , provided further confirmation that the targeted 21-nt site is necessary and sufficient for specific amiRNA-mediated specific resistance . In addition , the results also suggested that pre-amiR159-P69 , which is a modified form of the Arabidopsis miR159 precursor , can be processed by N . benthamiana plants to produce functional amiR159-P69 to confer virus resistance . As the 4 nt mutation on the central positions of the target sequence ( position 8 to 12 ) compromised specific resistance , we decided to further investigate nucleotide positions within this 21-nt sequence that are critical for amiRNA-mediated resistance . Note that the sequence can be systematically altered without affecting essential viral gene functions because the amiR159-P69 target sequence is non-essential to the TuMV-GP69 chimeric virus . Accordingly , we generated a series of mutants by making all possible synonymous scanning substitutions within the 21-nt sequence of P69 in the background of TuMV-GP69 ( Figure 4A ) . Each A of the P69 viral sequence that pairs to a U of amiR159-P69 was changed to a C to create a C∶U mismatch; each C and G of the viral sequence was changed to an A to create A∶G and A∶C mismatches; and each U of the viral sequence was changed to a C to create C∶A mismatches . A total of 21 mutant viruses with single nt substitution from the 1st to the 21st position of the target site were used to challenge non-transgenic WT and amiR159-P69 N . benthamiana plants . The proportion of inoculated amiR159-P69 plants that showed visible symptoms after inoculation , i . e . pathogenicity , was used as a measure of the importance of the mutated nucleotide within the 21-nt target site in amiR159-P69–mediated specific resistance . We used TuMV-GFP , TuMV-GP69 , and TuMV-GP69m as controls . Whereas WT plants were susceptible to TuMV-GP69 no symptoms developed in amiR159-P69 plants even at 10 dpi ( Table 1 , and Figure 4B bottom third panel ) . By contrast , TuMV-GFP and TuMV-GP69m elicited 100% infection on WT as well as amiR159-P69 transgenic tobacco plants , and these infected plants displayed symptoms at 5 dpi ( Table 1 , and Figure 4B bottom second and fourth panel ) . WT plants were 100% susceptible to all 21 scanning mutant viruses and symptoms appeared at 5 dpi ( data not shown ) , indicating that the single nt substitutions on the target 21-nt sequence did not affect mutant virus replication nor movement . On the other hand , these mutant viruses showed variable pathogenicity after inoculation on amiR159-P69 plants ( Table 1 , and Fig 4B top panels & bottom first panel ) . Fifteen mutants showed pathogenicity values that were significantly greater than zero ( Table 1 ) . For these pathogenic mutants , the percentage of infected plants ranged from 8 . 33% ( m7 ) to 92 . 86% ( m9 ) . Mutants were classified according to their pathogenicity using a 2-step cluster analysis . The minimum number of clusters into which the mutants can be significantly partitioned was three ( Bayesian weight 98 . 44%; Kruskal-Wallis test: H = 17 . 739 , 2 d . f . , P<0 . 001 ) . Figure 4C assigns mutants to the different clusters . The first cluster ( green bars in Fig . 4C ) , is characterized by positions causing low pathogenicity , with an average value of 6 . 91±0 . 59% , suggesting that these sites are non critical for resistance . These low pathogenicity mutants are scatter along the entire 21-nt region . The second cluster contains mutants of intermediate pathogenicity ( yellow bars in Figure 4C ) , with an average value of 36 . 36±4 . 71% , suggesting that these positions are moderately important for amiRNA-mediated resistance . Most of these medium effect mutants are located between nucleotides 10 and 18 , with the exception of m2 , which is located at the 3′ end of the target sequence . Finally , the third cluster contains those mutants with a greater likelihood of resistance breakdown ( red bars in Figure 4C ) . On average , these large effect mutants have 81 . 85±4 . 14% pathogenicity , highlighting their importance in amiRNA-mediated resistance . These important sites mostly congregate on the 3′ third of the target sequence , plus m9 and m12 which are located in the center of the sequence . In good agreement with the pathogenicity data , symptoms elicited by these mutants were generally delayed in comparison with TuMV-GFP ( Table 1 ) . For the 9 small-effect mutants , the median delay in symptom development was two days , for the medium effect mutants , two days , and for the large-effect mutants , only one day . We sought to gain deeper insights into the question of why different substitutions within the 21-nt target sequence overcame the amiRNA-mediated resistance to different degrees . To address this issue , we recovered viral RNAs from symptomatic leaves of amiR159-P69 plants and analyzed the 21-nt target sequence on the viruses by RT-PCR and sequencing . Several possibilities could account for the resistance breakdown . ( 1 ) The scanning mutation could affect amiR159-P69-mediated cleavage to different degrees in amiR159-P69 plants . ( 2 ) The scanning mutant virus could undergo rapid evolution accumulating additional mutation ( s ) within the target site to further increase the number of mismatches and consequently the ability to replicate in presence of the amiR159-P69 . ( 3 ) Since the 21-nt sequence and the GFP gene are non-essential for virus survival , the surviving mutant virus could undergo in-frame deletions in this region that would render the virus unrecognizable by the amiR159-P69 . To discriminate amongst these possibilities , we designed two primers ( PTuNIb-8671 and MTuCP-8982 ) to amplify an 1136 bp DNA fragment including a partial NIb gene , GFP gene , the target 21-nt sequence , and a partial CP gene ( Fig . 1B ) . This primer set can be used to check for any possible deletion within the GFP-21nt sequence . The recovery of lower molecular mass PCR fragment ( s ) would indicate deletion of this region . In addition , we designed another primer set ( PXFP-532 and MTuCP-8982 ) to amplify a 482-bp fragment that included a partial GFP gene and a partial CP gene ( Fig . 1B ) . This 482-bp DNA fragment can be used to analyze sequences surrounding and within the 21-nt target site . Figure 5A top panel shows that several virus sequences recovered from symptomatic amiR159-P69 plants contained deletion of the NIb-CP gene , such as TuMV-GP69m2 ( lanes 3 , 4 and 7 ) and TuMV-GP69m3 ( lanes 10 and 13 ) . In addition , several viruses , such as TuMV-GP69m5-13 , -15 , and -19 contained partially-deleted 21-nt sequence ( Fig . 5B ) . Moreover , TuMV-GP69m5-15 also accumulated two additional mutations on positions 6 and 8 ( Fig . 5B ) . We selected viruses with no deletion on the NIb-CP gene region and sequenced the GFP-CP gene regions ( Figure 5A , bottom panel ) . Our results showed that virus sequences recovered from symptomatic amiRNA plants contained additional mutations within the 21-nt target site ( Table 2 ) . These scanning mutant viruses have 1–3 additional mutation ( s ) on the 21-nt target site and most of additional mutations introduced additional mismatches ( Table 2 ) . Only 11 out of the 21 positions showed additional mutations . These 11 positions are 3 , 4 , 6 , 8 , and 10–16 ( Figure 5C ) . Positions 8 , 14 and 16 were of little importance for amiRNA mediated resistance ( see above ) , whereas all other 8 positions had either a moderate or a large effect on the likelihood of escaping the amiRNA-mediated resistance . All together , our results indicate that up to 2 mutations on critical positions within the 21-nt sequence can overcome specific resistance . Interestingly , we found that 40 out of 55 observed additional mutations were transitions . Over 50% of the additional mutations in positions 3 , 4 , 6 , 8 , 11 , and 15 were transition mutations . For example , there were 100% U→C or C→U transitions in position 3 and 4 . In position 15 , the G→A transition represented 88 . 89% of all observed mutations at this particular site . Moreover , there were 50 to 66 . 67% of G→A transitions at positions 6 , 8 , and 11 . This result is not surprising , since it is well known that virus coding regions show an excess of transitions over transversions [20] , [21] . Three reasons can account for this bias: ( i ) the underlying mechanisms of mutation render transitions easier than transversions , ( ii ) the redundancy of the genetic code is expected to make the average effect of a transition smaller than the average effect of a transversion , and ( iii ) RNA editing by deaminase-like enzymes have been shown to induce transition mutations in single-stranded regions of certain viral genomes [22] , [23] . Our results show that transitions rather than transversions also mainly accumulate in viral sequences , such as that of the target of amiR159-P69 , which are not under the selective constrain imposed by being a coding sequence . Furthermore , not all transitions are equally represented in Table 2 , since G→A ( 17/40 ) and C→U ( 14/40 ) are significantly over represented ( χ2 = 12 . 600 , 3 d . f . , P = 0 . 006 ) . This bias amongst transitions is expected if the viral RNA was edited by cytidine deaminase enzymes .
Here , we have developed a heterologous-virus resistance system to study and identify critical positions of amiRNA target site for amiRNA-mediated resistance . The amiR159-P69 transgenic plant were resistant to TYMV , but not to TuMV ( a heterologous-virus ) , because there was no sequence homology with amiR159-P69 on the TuMV viral genome [4] . However , the chimeric heterologous-virus TuMV-GP69 carrying the 21-nt sequence of P69 gene cannot infect amiR159-P69 plants because of amiR159-P69-mediated cleavage . By contrast , the TuMV-GP69m virus with mutations on the central region within this sequence is sufficient to prevent amiRNA-mediated cleavage on the viral RNA and compromise specific virus resistance . These results indicated that the 21-nt target site is portable and is necessary and sufficient to confer virus resistance . Because of the genome organization and proteolytic processing strategy of potyvirus , TuMV can express GFP when a cDNA for this protein is inserted in-frame between the NIb and CP genes . The encoded GFP protein contains two NIa proteinase cleavage sites ( CVYHQ/A ) at the N- and the C-terminus such that GFP can be released from the viral polyprotein by proteolytic processing . In addition , the additional 21-nt target site that encodes seven amino acids is also nonessential for TuMV . Therefore , any modification on the GFP gene and the 21-nt target site would not affect the chimeric TuMV as evidenced by its ability to infect plants and stably replicate . Using an in vivo assay we identified critical positions on the 21-nt target sequence for RISC-amiRNA-mediated cleavage . Scanning mutations on the 21-nt target site of the challenging chimeric virus showed different degree of resistance breakdown on amiR159-P69 transgenic plants . Based on the proportion of amiR159-P69 plants that become susceptible , we defined critical , moderately critical and non-critical positions on the 21-nt sequence . Positions 3–6 , 9 , and 12 , are found to be critical for resistance because chimeric virus with mutations at these sites were pathogenic , on average , on ∼82% of amiR159-P69 plants . Positions 2 , 10 , 11 , 13 , 15 , and 18 are classified to be moderately critical mutations giving average pathogenicity of ∼36% in transgenic plants . The remaining positions are classified as non-critical for resistance since mutants at these sites were only pathogenic in less than 7% of inoculated plants . In summary , most critical positions are localized on sequences complementary to the 5′ portion of the amiRNA whereas the moderate critical positions are mainly localized on the central region of the target site . The exception being position 18 , which is complementary to the 18th nucleotide on 3′ portion of the amiRNA , and was also moderately important for amiRNA-mediated resistance . These results are consistent with those obtained with in vitro miRNA-mediated cleavage using a wheat germ system [16] . All together , results suggest that the 5′ portion of the miRNA is more important in governing the specificity of miR165/166 regulation [16] . Furthermore , the “two-state model” for RISC-mediated target recognition also proposes that the specific interaction between RISC and the substrate is initiated via the 5′ portion of siRNA , as the 3′ portion is less favorably structured to undergo base pairing before the initial recognition of a mRNA target [17] . Representative results showed that several virus sequences recovered from symptomatic amiRNA plants contained deletions and additional mutations within the 21-nt target site . This observation is consistent with the hypothesis that mutations in certain critical positions within the target site reduced amiRNA-mediated cleavage efficiency ( Figure 6A ) . The reduced RISC activity allowed certain mutant viruses to escape amiRNA-mediated cleavage , allowing them to replicate and complete an infectious cycle ( Figure 6B ) . During subsequent virus replication , additional mutations or deletions of the target sequence would be positively selected because they would escape from the amiRNA-mediated specific resistance more efficiently ( Figure 6B ) . Indeed , the effect of miRNA-mediated cleavage was drastically alleviated in transgenic plants expressing the silencing suppressor P1/HC-Pro . Chimeric Plum pox virus ( PPV ) carrying an endogenous miRNA target site can also overcome the resistance by deletion and mutation on the 21-nt target sequence [24] . Finally , deletions , in addition to point mutations , are also a very common pathway taken by HIV-1 to escape from RNAi-based therapy in cell culture experiments [10] , [13] . In general , these deletions have a major impact on the local RNA secondary structure , creating new hairpin structures not accessible to the siRNAs [13] . In some deletion mutant viruses recovered from breakdown plants , the entire GFP gene , along with small portions of the NIb gene C-terminus or the CP gene N-terminus , has been deleted from the viral genome ( data not shown ) . These results suggest that TuMV can tolerate small deletions in the NIb or CP gene . In addition , deletions in between the GFP gene and the target site ( Fig . 5B ) may be triggered by polymerase-jumps on repeat sequence ( ACAA ) . Widespread plant miRNA-directed translational repression as an important miRNA-mediated regulatory mechanism in plants has recently been reported [25] . Imperfect pairing with central mismatches in small RNA-target hybrids promotes translational repression because it excludes slicing [25] . This observation suggests the possibility that imperfect pairing between the amiRNA and mutant target sequences might lead to translational repression rather than viral RNA cleavage . In contrast to the catalytic effects of amiRNA-mediated viral RNA cleavage , translational repression requires stoichiometric amounts of amiRNAs and therefore is not as efficient . Inefficient translation inhibition might allow residual virus replication and progeny virus can still escape the repression by fixing changes in the target sequence . In this study , we have provided evidence that the 21-nt target site is necessary and sufficient for amiRNA specific resistance and we have also identified several positions on the target site that are critical for this resistance . These results are clearly important for future design of amiRNA-mediated virus resistance . Highly conserved regions on viral genomes should be selected as target sites to minimize the likelihood of fixation of mutations responsible for resistance breakdown , because these mutations might affect viral protein function and hence have a negative impact on virus fitness and survival . Furthermore , several amiRNAs targeting different conserved regions on a viral genome could be co-expressed in transgenic plants to minimize the chances of resistance breakdown . Finally , the heterologous-viral system described here also can be used for viral evolution studies in the future . As we have highlighted several times here , the amiR159-P69 target sequence inserted in the TuMV-GFP genome is functionally neutral . This has allowed us to separate selective pressures acting on the protein level from those acting on the RNA level . Consequently , the patterns of molecular evolution should be different . Not surprisingly , and in agreement with previous data obtained with other viruses , we have observed that most of the mutations fixed within the target were transition mutations [21] . We consider it striking that 77 . 50% of these fixed transitions were of the type G→A and C→U . These transitions are from the particular type induced by cellular cytidine deaminases ( CDAs ) involved in innate immune responses to viral infection [26] , a phenomenon particularly well described for HIV-1 and other retroviruses [23] , [27] but never before on an RNA virus . This observation prompted us to hypothesize that as an antiviral strategy plants may have an RNA-editing system that induces hypermutagenesis in viral genomes . A thaliana contains a family of nine paralogous genes that are annotated as CDAs owing to their homology to CDA1 [28] . These nine genes are good candidates to explore whether their gene products possess cytidine deaminase activity and whether they are indeed involved in plant antiviral defense .
Two amiRNA transgenic A . thaliana lines , amiR159-P69 and amiR159-HC-Pro , were used in this study [4] . Plants of N . benthamiana were transformed with Agrobacterium tumefaciens containing the pBA-amiR159-P69 plasmid by standard methods . T2 transgenic N . benthamiana ( a mixture of homozygotes and hemizygotes ) were analyzed for transgene and miRNA levels and 4 independent lines ( #1 , 2 , 3 , and 4 ) with high amiR159-P69 expression levels were selected for virus challenge experiments . Seeds were surface-sterilized and chilled at 4°C for 2 d before being placed on Murashige and Skoog ( MS ) medium with/without antibiotics or sowed directly on Florobella potting compost/sand mix ( 3∶1 ) . Plants were maintained in a growth room ( 16 h light/8 h darkness , 20 to 25°C ) . Ten µg of total RNA was resolved in a 15% polyacryamide/1× TBE ( 8 . 9 mM Tris , 8 . 9 mM boric acid , 20 mM EDTA ) /8M urea gel and blotted to a Hybond-N+ membrane ( Amersham ) . DNA oligonucleotides with the exact reverse-complementary sequence to miRNAs were end-labeled with 32P-γ-ATP and T4 polynucleotide kinase ( New England Biolabs ) to generate high specific activity probes . Hybridization was carried out using the ULTRAHyb-Oligo solution according to the manufacturer's directions ( Ambion ) and signals were detected by autoradiography . In each case , the probe contained the exact antisense sequence of the expected miRNA to be detected . The TuMV infectious clone ( p35STuMV-GFP ) comprises of a 35S promoter and the full-length cDNA of TuMV-GFP . The GFP gene was inserted between NIb and CP genes . This chimeric TuMV-GFP virus was used as a surrogate wild type virus and as a backbone for construction of various chimeric recombinant viruses described here . We fused the 21-nt sequence ( 5′-CCACAAGACAAUCGAGACUUU-3′ ) of the TYMV P69 gene targeted by amiR159-P69 to the 3′ end of the GFP gene . The GFP-P69 fusion sequence was then inserted in between the NIb and CP genes to generate the p35STuMV-GP69 infectious clone . As a control , the central 4 nts ( underlined ) of the 21-nt target sequence ( 5′-CCACAAGACCUGAGAGACUUU-3′ ) was mutated to give GFP-P69m which was also inserted in the same position of the virus to generate p35STuMV-GP69m . As the 21-nt target sequence is in a non-essential region of the TuMV-GP69 it can be altered without affecting essential viral gene function . We performed serial single nt mutagenesis from the 1st-nt to the 21st-nt of the target site on the TuMV-GP69 infectious clone by PCR mutagenesis and the resulting series of scanning mutants were confirmed by sequencing . A total of 21 single-nt substitution recombinant viruses were generated . Based on the mutation position , the recombinant viruses were named TuMV-GP69mX , in which X refers to the mutation position . For example , the mutant with substitution on the 1st-nt of the target site was named TuMV-GP69m1 . To evaluate the efficiency of amiRNA-mediated specific resistance toward wild type and mutant viruses , we have established a standard protocol for virus challenge inoculation and quantitative evaluation of pathogenicity . Our overall aim was to reduce the time for virus maintenance and propagation in host plants so as to minimize possible virus evolution . All recombinant viruses were propagated from DNA infectious clones . Aliquots of 20 µL , containing 1 µg of DNA in sterilized water , were mechanically applied onto carborundum-dusted leaves of Chenopodium quinoa Willd with a sterilized glass spatula . Seven days post-inoculation ( dpi ) , local lesions were obtained on inoculated leaves . Viruses were then isolated from single lesions and transferred to N . benthamiana for amplification . Four dpi leaves of N . benthamiana with viral infection symptoms were used as the source of inoculum to challenge WT and amiR159-P69 N . benthamiana plants for evaluation of virus pathogenicity ( i . e . , frequency of break-down ) . Twenty amiR159-P69 plants were used for each experiment , and the experiments were repeated 3 times . Resistance efficiency of amiR159-P69 plants challenged with recombinant viruses were compared with those obtained with control viruses , including TuMV-GFP , TuMV-GP69 and TuMV-GP69m . Pathogenicity was evaluated between two and four times for each one of the 21 TuMV-GP69mX recombinant viruses . Count data from experiments that were statistically homogeneous were pooled into a single set , whereas experiments that behaved as outliers were removed from the dataset . In plants displaying symptoms , it was important for us to verify the sequence of the 21-nt target site to ensure that no other mutations had occurred to confound our results . To this end , total RNA was extracted from infected leaf tissues using the Trizol reagent ( Invitrogen ) . One µg total RNA was used for reverse-transcriptional polymerase-chain reaction ( RT-PCR ) with PTuNIb-8671 ( 5′-GAACCAGCTCAAGAGGATCT-3′ ) and MTuCP-8982 ( 5′-GCCACTCTCTGCTCGTATCTTGGCACGCGC-3′ ) for amplification of the viral region between the partial NIb and the CP genes ( Fig . 1B ) . The PCR fragments then were analyzed by sequencing . The pathogenicity of different recombinant viruses was estimated as the frequency of infected plants out of the total number of inoculated plants . The LaPlace's point estimator for the Binomial frequency parameter was used instead of the commonly used maximum likelihood estimator [29] . The LaPlace method provides a more robust estimate of the Binomial parameter for small sample sizes [30] . Binomial 95% confidence intervals ( CI ) were also computed . TuMV-GP69mX recombinant viruses were classified into groups of similar pathogenicity using the two-step cluster analysis [31] . In brief , this method classifies data in groups that minimize the within-group variance whilst maximizing the among-groups variance . The method starts with the simplest model ( i . e . , all viruses are equally pathogenic ) and computes its likelihood; then , it classifies the mutants into two clusters and computes the likelihood of this model; finally , it does the same for three clusters , four clusters and up to 21 clusters ( i . e . , each site behaves in a different way and no classification is possible ) . For each model , Schwarz's Bayesian information criterion ( BIC ) was used as a measure of the goodness-of-fit and the model with the lowest BIC was considered to be the best one [32] .
|
RNA viruses are well-known for their tremendous capacity to evolve , a characteristic that threatens the development of effective antiviral strategies . A new antiviral strategy was recently proposed to control plant RNA viruses that relied on the expression in plants of artificial microRNAs ( amiRNAs ) targeting short sequences of 21-nt in the viral genome . Here , we have evaluated the likelihood that changes in the 21-nt target sequence would result in resistance breakdown . We found that changes at different sites in the target had different consequences on the ability of the virus to evade amiRNA surveillance . Then , we evolved viruses with a single substitution within the target under the selective pressure imposed by amiRNAs but without any selective pressure at the protein level . We found extra mutations accumulated in the target that further reduced base pairing with the amiRNA . These results showed that when allowed to replicate , RNA viruses would readily generate genetic variability that would facilitate evasion from the engineered miRNA-mediated virus resistance .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"virology/virus",
"evolution",
"and",
"symbiosis",
"virology/antivirals,",
"including",
"modes",
"of",
"action",
"and",
"resistance"
] |
2009
|
Molecular Evolution of a Viral Non-Coding Sequence under the Selective Pressure of amiRNA-Mediated Silencing
|
In prokaryotic genomes the number of transcriptional regulators is known to be proportional to the square of the total number of protein-coding genes . A toolbox model of evolution was recently proposed to explain this empirical scaling for metabolic enzymes and their regulators . According to its rules , the metabolic network of an organism evolves by horizontal transfer of pathways from other species . These pathways are part of a larger “universal” network formed by the union of all species-specific networks . It remained to be understood , however , how the topological properties of this universal network influence the scaling law of functional content of genomes in the toolbox model . Here we answer this question by first analyzing the scaling properties of the toolbox model on arbitrary tree-like universal networks . We prove that critical branching topology , in which the average number of upstream neighbors of a node is equal to one , is both necessary and sufficient for quadratic scaling . We further generalize the rules of the model to incorporate reactions with multiple substrates/products as well as branched and cyclic metabolic pathways . To achieve its metabolic tasks , the new model employs evolutionary optimized pathways with minimal number of reactions . Numerical simulations of this realistic model on the universal network of all reactions in the KEGG database produced approximately quadratic scaling between the number of regulated pathways and the size of the metabolic network . To quantify the geometrical structure of individual pathways , we investigated the relationship between their number of reactions , byproducts , intermediate , and feedback metabolites . Our results validate and explain the ubiquitous appearance of the quadratic scaling for a broad spectrum of topologies of underlying universal metabolic networks . They also demonstrate why , in spite of “small-world” topology , real-life metabolic networks are characterized by a broad distribution of pathway lengths and sizes of metabolic regulons in regulatory networks .
In prokaryotic genomes the number of transcriptional regulators is known to quadratically scale with the total number of protein-coding genes [1] . The toolbox model of co-evolution of metabolic and regulatory networks was recently proposed [2] to explain this scaling in parts of the genome responsible for metabolic functions . In this model prokaryotes acquire new metabolic capabilities by horizontal transfer of entire metabolic pathways from other organisms . One can conveniently think of these new pathways as coming from some “universal network” formed by the union of metabolic repertoires of all potential donor organisms . The essence of the toolbox argument [2] can be summarized as follows: as the non-regulatory part of the genome of an organism ( its “toolbox” ) grows , it typically needs to acquire fewer and fewer extra new genes ( “tools” ) in a pathway offering it some new metabolic capability ( e . g . the ability to utilize a new nutrient or synthesize a new metabolic product ) . As a consequence , the number of pathways and by extension the number of their transcriptional regulators grows faster than linearly with the number of non-regulatory genes in the genome . While this qualitative explanation is rather general and therefore does not depend on specific details such as topology of the universal network , the exact value of the exponent α connecting the number of transcription factors ( equal to - the number of pathways or leaves of the network ) to the number of metabolites in the metabolic network of an organism , as , is in general model-dependent . In [2] we mathematically derived the quadratic scaling ( ) for the toolbox model with linear pathways on a fully connected graph in which any pair of metabolites can in principle be converted to each other in just one step via a single metabolic reaction . While this situation is obviously unrealistic from biological standpoint , before present study it remained the only mathematically treatable variant of the toolbox model . The universality of the exponent was then corroborated [2] by numerical simulations of the toolbox model with linearized pathways on the universal network formed by the union of all metabolic reactions in the KEGG database . While the agreement between the values of the exponent in these two cases hinted at underlying general principles at work , the detailed understanding of these principles remained elusive . The question we address in this study is how the topology of the universal network determines this scaling exponent . To answer this question we first consider and solve a more realistic ( yet still mathematically treatable ) case in which the universal metabolic network is a directed tree of arbitrary topology . While being closer to reality than previously solved [2] case of fully connected network , the toolbox model on a tree-like universal network still retains a number of simplifications such as strictly linear pathways and one substrate → one product reactions . To make our approach even more realistic we propose and numerically study a completely new version of the toolbox model incorporating metabolic reactions with multiple substrates and products as well as branched and cyclic metabolic pathways . Furthermore , unlike random linear pathways on a universal network [2] that can be long and therefore suboptimal from an evolutionary standpoint , the new model uses evolutionarily optimized pathways with the smallest number of reactions from the KEGG database sufficient to achieve a given metabolic task .
We will first consider the case where the universal metabolic network is a directed tree . For simplicity in this section we will consider the case of catabolic pathways , while identical arguments ( albeit with opposite direction of all reactions ) apply to anabolic pathways . The root of the tree corresponds to the central metabolic core of the organism responsible for biomass production . Peripheral catabolic pathways ( branches of the tree ) convert external nutrients ( leaves ) to this core , while the internal nodes of the tree represent intermediate metabolites . Each of metabolites is characterized by its distance from the root of the network . The universal network has metabolites at distance from the root that included leaves ( nutrients used in the first step of catabolic pathways ) and branching points corresponding to intermediate metabolites generated by more than one metabolic reaction at the next level ( see Figure 1 ) . An organism-specific network ( filled circles and thick edges in Figure 1 ) at distance from the root contains metabolites composed of leaves , branching points , and metabolites inside linear branches ( “one reaction in-one reaction out” ) . For simplicity we assume that in the universal network ( and thus also in any of its organism-specific subnetworks ) no more than two reaction edges can combine at any node ( metabolite ) , while the most general case of an arbitrary distribution of branching numbers can be treated in a very similar fashion . The toolbox model specifies rules by which organism acquires new pathways in the course of its evolution . It consists of the following steps: 1 ) randomly pick a new nutrient metabolite ( a leaf node of the universal network that currently does not belong to the metabolic network of the organism ) 2 ) use the universal network to identify the unique linear pathway which connects the new nutrient to the root of the tree ( the metabolic core ) and finally 3 ) add the reactions and intermediate metabolites in the new pathway to the metabolic network of the organism ( filled circles and thick edges in Figure 1 ) . One needs to only add those enzymes that are not yet present in the “genome” of the organism . Graphically it means that the new branch of the universal network is extended until it first intersects the existing metabolic network of the organism . Consider an organism capable of utilizing nutrients represented by leaves in the universal network , where and . Since we assume that each nutrient utilization pathway is controlled by a dedicated transcriptional regulator sensing its presence or absence in the environment ( e . g . LacR for lactose ) , the corresponding regulatory network would also have transcription factors ( in the model we ignore transcription factors controlling non-metabolic functions ) . The non-regulatory part of the genome consists of enzymes catalyzing metabolic reactions ( strictly speaking is the number of metabolites/nodes so that the number of enzymes/edges is ) . Quadratic scaling plots [1] shows the number of transcriptional regulators vs . the total number of genes in the genome ( both regulatory and non-regulatory ) . However , since in all organism-specific networks NM ≫ NL , the quadratic scaling between and is equivalent to . We further assume that due to random selection nutrients are expected to be uniformly distributed among all d levels . Therefore , the expected number of leaves at a given level is given by where the fraction is the same at all levels . On the other hand the fraction varies from level to level . It usually tends to increase as one gets closer towards the root of the tree reaching 1 for d = 0 ( the root node itself ) . To derive the equation for , one first notices that each of metabolites at level is converted to another intermediate metabolite at level . Due to merging of pathways at branching points the number of unique intermediate metabolites at the level is actually smaller: . To calculate one uses the fact that each of the two nodes downstream of a branching point in the universal network is present in the organism-specific network with probability . The probability that they are both present is and thus the number of branching points at level of the organism-specific metabolic network is . The intermediate metabolites together with new nutrients entering at the level add up to the total number of metabolites at level : ( 1 ) This equation allows one to iteratively calculate for all d starting from . We will use this equation to derive the relationship between the number of leaves and the total number of nodes first for a critical branching tree and then for a supercritical one . The Galton-Watson branching process [3] is a simple stochastic process generating random trees , and we will consider its version where a node can have two , one , or zero neighbors ( parents ) at the previous level with probabilities p2 , p1 and p0 correspondingly . If the average number of parents equals one , then the process is referred to as critical , and if is greater than one then the process is supercritical . More generally critical and supercritical branching trees can be generated by a variety of random processes such as e . g . directed percolation [4] . While for simplicity we used the Galton-Watson branching process in our derivation below , it can be readily extended to this more general case . The principal geometric difference between supercritical and critical trees is that in the former case the number of nodes in a layer exponentially grows with [3] , while in a critical tree it grows at most algebraically ( for the Galton-Watson critical process [3] ) . The other difference is that while the critical branching process always stops on its own at a certain finite height , a supercritical process will go on forever so that to generate a tree one has to manually terminate it at a predefined layer . The most significant feature of a critical tree is that it has much longer branches than a supercritical one of the same size . Indeed , the diameter ( the maximal height ) of a random critical tree with nodes is while in a supercritical tree it is much shorter: . Thus supercritical trees ( unlike their critical counterparts ) have the small world property . A random critical network where each node has at most has two parents in the previous layer is defined by . Indeed , in this case . In such network and hence the Eq . ( 1 ) can be rewritten as ( 2 ) A critical branching process that has not terminated by level d satisfies or . More generally if algebraically increases with , asymptotically approaches 1 as ( 3 ) Here as , thus for remains approximately constant and according to Eq . ( 2 ) this constant ratio is defined by ( 4 ) This quadratic relation is exact in a critical branching tree where each node can branch out into at most two nodes at the next layer , and it is still correct to a leading order in for a critical branching tree with arbitrary branching ratios ( see “Quadratic relation between and for general critical branching processes” of Text S1 ) . Furthermore , one can show ( see “Calculation of the average in the toolbox model on a critical tree” of Text S1 ) that in large critical networks the overall fraction of metabolites present in organism-specific metabolic network is very close to this stationary limit of : . As was explained in the previous section the ratio between the total number of metabolic-related genes in the genome of an organism and its theoretical maximal value for a genome containing the entire universal network is also given by . Furthermore , in our model the number of leaves is equal to the number of nutrient-utilizing pathways or , alternatively , their transcriptional regulators . Thus like in a much simpler model of Ref . [2] the toolbox model on any critical tree-like universal network gives rise to quadratic scaling of the number of transcription factors with the total number of genes: ( 5 ) The geometrical properties of the universal network such as its total number of nodes/edges and number of leaves/branches determine the prefactor of this scaling law . Simulation of the toolbox model on the critical tree ( Figure 2 ) verified our mathematical predictions with the best fit to binned datapoints in Figure 2 giving the exponent α = 1 . 9±0 . 1 . For a supercritical branching process and according to Eq . ( 1 ) ( See SI for the derivation ) the steady state value of satisfies ( 6 ) Here and . Notice that for one has two solutions for : and . This indicates transition in which for exactly at zero one has , while for an arbitrary small yet positive the value of asymptotically converges to for . This transition resembles the first order phase transition , e . g . , liquid-gas transition , where right at the transition point very small variation of the external parameter such as temperature ( analogous to in this model ) results in a large jump of the order parameter such as density ( analogous to our ) . ( See [5] for details ) , The number of layers over which this conversion is taking place is itself a function of and for small it is large . For exponentially growing supercritical networks and for small , the network average value of defined as satisfies ( 7 ) Note that this equation connecting and ( see SI for detailed derivation ) is markedly different from Eq . ( 6 ) for steady state value in middle layers . In conclusion , while the toolbox model on a critical universal network is characterized by a quadratic scaling between and ( see Eq . ( 4 ) ) , the same model on a supercritical , exponentially expanding universal network gives rise to a linear scaling of vs . albeit with logarithmic corrections ( see Eq . ( 7 ) ) . This difference in exponent equally applies to the scaling of the number of regulators vs . the total number of genes in the toolbox model on critical and supercritical universal network . To test our mathematical results for a more realistic version of the universal tree we linearized pathways and reactions in the network formed by the union of all reactions in the KEGG database [6] . To this end we generated a random spanning tree on the KEGG network by the following algorithm: the metabolite pyruvate was selected as the root of the tree . We then randomly picked a metabolite located upstream of it and generated a linear pathway ( tree branch ) as a self-avoiding random walk on the KEGG network extended until it either merges with another pathway or reaches the root of the tree . This step was repeated until all upstream metabolites were covered . The resulting spanning tree was then used as the universal network on which we simulated the toolbox model by gradually increasing the number of pathways and recording the total number of metabolites in organism-specific metabolic networks . Our numerical simulations generated approximately quadratic scaling ( see Ref . [2] ) . To better understand the origins of this scaling we investigated the topology of the underlying universal tree . The criticality of a tree is defined by the asymptotic value of the ratio for large : for supercritical trees it reaches a limit , while for critical ones it converges to 1 as described in Eq . ( 3 ) . Figure 3 showing vs . in the linearized KEGG network convincingly demonstrates its criticality . Thus the quadratic scaling between the number of transcriptional regulators and the number of metabolites in the toolbox model simulated on the linearized KEGG network is explained by the mathematical formalism described in previous sections . In addition to using a random spanning tree to linearize the KEGG network we also tried a version using minimal paths . In this version the universal network is generated by randomly picking a metabolite and connecting it to the root of the tree ( pyruvate ) by the shortest path . At a first glance such “minimal path” selection appears to be reasonable from evolutionary standpoint . Indeed , evolution would favor simpler and shorter pathways in order to minimize the expenditure of resources to achieve a given metabolic goal [7] . However , the minimal paths version of linearization of the KEGG resulted in a supercritical universal network with logarithmically short branches . As predicted for supercritical trees ( Eq . ( 7 ) ) the toolbox model in this case had an approximately linear scaling of the number of transcriptional regulators ( leaves of branches on the network ) with the total number of metabolites: the measured best fit exponent was only . How do we reconcile the evolutionary pressure apparently selecting for minimal pathways with dramatically wrong scaling properties of this model ? We believe that most of the ultra-short “small world” pathways generated by minimal paths on the KEGG network are unrealistic from biochemical standpoint . Indeed , highly connected co-factors often position metabolites with very different chemical formulas in close proximity to each other . For example , the KEGG reaction R00134: would appear as a miraculous “one-step” conversion of carbon dioxide into formate , while the reaction R03546: would artificially link carbon dioxide and cyanate . The combination of these two reactions gives rise to equally impossible two-step path: formate → CO2 → cyanate . As a consequence of such artificial shortcuts branches of the universal network linearized by minimal paths are much shorter than they are in reality . . This problem is at least partially alleviated by 1 ) removing unusually high-degree nodes corresponding to common co-factors such as H2O , ATP , NAD in the metabolic network so that some unrealistic paths are eliminated , and also 2 ) using random spanning tree instead of the shortest paths . In Ref . [2] we followed both of these recipes to successfully reproduce the quadratic scaling in real-life data . Still no linearization procedure could completely avoid biochemically meaningless shortcuts . In the next section we introduce and study a new considerably more realistic version of the toolbox model operating on branched and interconnected universal networks . Pathways in this version of the toolbox model satisfy the evolutionary requirements for minimal size . Proper treatment of metabolic reactions with multiple substrates prevents biochemically meaningless shortcuts and as a consequence restores the quadratic scaling . Real metabolic reactions routinely include multiple inputs ( substrates ) and multiple outputs ( products ) ( see Table 1 and Table 2 for statistics in the KEGG database ) . Furthermore , metabolic networks often have two or more alternative pathways generating the same set of end-products from the same set of nutrients . Both these factors result in metabolic networks that are branched and interconnected . Here we propose and simulate a more realistic version of the toolbox model . The most prominent feature of the new model of pathways is the “AND” function acting on inputs of multi-substrate reactions . It reflects the constraint that a reaction cannot be carried out unless all its substrates are present . The new version of the toolbox model simulates addition of anabolic pathways aimed at production of new metabolites from those the model organism can currently synthesize ( its current metabolic core ) . The new pathways are optimal in the sense that they contain the smallest number of reactions necessary to synthesize the desired end-product . As for previous versions of the toolbox model , one can modify the rules of this model to apply to catabolic pathways but for simplicity we will limit the following discussion to anabolic pathways . The rules of the new model are: Numerical simulation of this model shows that the number of transcriptional regulators scales with the number of metabolites with power ( Figure 5 ) . This is consistent with quadratic scaling we observed and mathematically derived for a simpler model with linearized pathways composed of single-substrate reactions . The mathematical formalism derived in the previous sections is limited to tree-like universal networks and thus does not directly apply to the new model . Nevertheless , one generally expects the quadratic scaling to be limited only to critical , “large world” networks in which organisms with small genomes initially tend to acquire sufficiently long pathways . As noted before , from purely topological standpoint the KEGG network has a “small world” property making long pathways unlikely . It is important to check if the realistic treatment of multi-substrate reactions did in fact restore the “large world” property and criticality to the KEGG universal network by increasing the minimal number of steps required for connecting target metabolites to the metabolic core . To quantify the criticality of the expansion process as before we use the ratio where denotes the number of metabolites reached at step of the scope expansion starting from the initial seed subset of metabolites . As in the case of critical branching trees this ratio asymptomatically converges to 1 thus confirming the criticality of the scope expansion process . The mere existence of ∼40 steps in this process ( the x-axis in Figure 6 ) can serve as evidence in favor of “large world” character of the KEGG universal network characterized by the existence of long pathways . Unlike linearized pathways in the original version of the toolbox model [2] , branched pathways in the more realistic model from previous section are interesting objects in their own right . We identified several geometrical properties of these pathways ( see Figure 4 for illustration ) quantifying their position relative to the core network to which they were added: 1 ) nborder rxn–the number of added reactions that are connected ( as a substrate or a product ) with at least one metabolite in the core , 2 ) nbase–the number of metabolites in the core that serve as substrates to reactions in the added pathway , 3 ) nfeedback–the number of core metabolites that are products of reactions in the new pathway , 4 ) nbyproduct–the number of final metabolic products of the added pathway that are neither core metabolites nor the target , 5 ) length-the number of steps ( layers of the scope expansion process ) it takes to transform core metabolites into the target product . 4 illustrates the definition of these parameters while Figure 7 and Figure 8 plot these parameters as a function of ( the number of metabolites in the added pathway ) or ( the number of reactions in the added pathway ) . Approximately linear relationship between nborder rxn vs . ( Figure 8a ) suggests that added pathways tend to be located at or near the surface of the core metabolic network of the organism . Most of reactions in these pathways use metabolites from this core network either as substrates ( nbase ) or as products ( nfeedback ) . Further analysis indicates that “currency metabolites” ( common co-factors that serve as substrates or products of many reactions ) constitute a significant fraction ( ∼25% ) of all core metabolites involved in border reactions ( see the section “Analysis of the currency metabolites in the toolbox model” of Text S1 for details ) . On the other hand , the fact that the number of steps in a pathway ( its length ) constitutes a good fraction of its overall number of reactions implies that , in spite of these numerous “shortcut” connections to the core , added pathways remain very thin and essentially linear . That is to say , these pathways tend to work as a single “conveyor belt” sequentially converting intermediate products into each other instead of having two or more parallel “processing lines” and assembling final products of these lines only at final stages of the pathway . This finding provides an intuitive reason why models with branched and linearized pathways have similar scaling properties . One can argue that this is because pathways in both models are essentially linear . Yet , in spite of their linearity and optimality ( each has the smallest number of reactions to generate the target from the core ) added pathways in the new version of the model are very different from shortest paths on the universal network . As illustrated in Figure 9 the average pathway length is several times longer than the geometrically shortest path between the target and the core . As can be seen from Figure 7 , most of added pathways ( around 97% ) do not generate any byproducts . They only produce the intended target and nfeedback metabolites in the core network of the organism to which they were added . The relative scarcity of byproducts suggests that pathways in our model satisfy the evolutionary constrains imposed on real-life organisms . Indeed , as previously proposed in Ref [9] it makes sense to assume that evolution favors pathways that achieve a given metabolic goal using the smallest number of enzymes and at the same time striving to generate the maximal possible yield . Unnecessary byproducts not only reduce the yield of the desired metabolic target , they also might become toxic in high concentrations and thus would require extra transporter proteins to pump them out .
The small world property of complex biomolecular networks has been extensively discussed in the literature during the last decade ( see [10]–[12] for earliest reports in metabolic and protein interaction networks correspondingly ) . It was often assumed that the small world effect positively contributes to the robustness of the network by providing multiple redundant pathways for target production in metabolic networks or for propagation of signals along regulatory and protein interaction networks . In addition to its positive aspects the small world property in biomolecular networks also has a potentially negative side by facilitating system-wide propagation of undesirable cross-talk [13] . In the course of evolution different strategies appeared allowing organism to limit and attenuate these unwelcome side effects of global connectivity . The extent of small world topology in metabolic networks has been recently disputed in [14] . There it was argued that many shortcuts in simple graph representations of metabolic networks are meaningless from biochemical standpoint . By taking into account additional structural information about metabolites Arita [14] dramatically increased the diameter of the metabolic network in E . coli . In our simulations of the toolbox model we also encountered limitations of the simple graph representation giving rise to small world topology of metabolic networks . Small world by definition implies very short pathways ( or equivalently supercritical network branching with exponentially growing lists of neighbors at distance ) , which in its turn prevents the appearance of quadratic scaling in the linear toolbox model . How to reconcile this apparent contradiction ? The answer known from pioneering studies of R . Heinrich and collaborators ( see e . g . [8] , [15] , [16] ) is to altogether abandon the simple graph representation in favor of realistic treatment of multi-substrate reactions . A metabolic reaction with two or more substrates will not proceed at any rate until all these metabolites are present in the cell . This implicit “AND” function operating on inputs of multi-substrate metabolic reactions makes reaching a given metabolic target much harder task and ultimately leads to dramatically longer pathways ( Figure 9 quantifies this effect ) . These longer pathways in turn reinstate the quadratic scaling in the version of the toolbox model that was introduced in the previous section . This leads to the novel conclusion of our study that , when multi-substrate reactions are properly taken into account , the small world ( supercritical ) topology of the metabolic universe disappears in favor of the “large world” topology characteristic of critical branching networks . The increase in the effective diameter of the network due to this effect is dramatic . One goes from 3–4 steps diameter typical of a small world network of [12] , [11] to ∼8 steps of [14] and finally to 30–40 layers in the scope expansion process shown in Figure 6 ( see also Figure 6 of [8] ) . These arguments lead us to adapt the “scope expansion” algorithm by Heinrich et al [8] to pathway acquisition in the toolbox model . Not only did it restore the “large world” properties such as quadratic scaling to the model , it also made the added pathways plausible from evolutionary standpoint . Unlike linear random walk pathways on KEGG network used in [2] , pathways in the new version of the toolbox model have the smallest number of KEGG reactions to achieve their metabolic task ( production of the target metabolite from the set of metabolites already present in organism's network ) . As can be seen in Figure 7 a large fraction of these pathways also does not generate any byproducts . Accumulation of such byproducts inside a cell is potentially dangerous and would require specialized proteins to excrete them to the environment . The lack of byproducts also means that the useful yield of an added pathway is at or near its theoretical maximum . This is consistent with the fact that real biological pathways are optimized in the course of evolution to increase their yield while simultaneously reducing the number of reaction steps [7] , [17] , [18] . Optimality of metabolic pathways in central carbon metabolism was recently discussed in Ref . [17] . There it was shown that some ( but not all ) of these pathways coincide with the shortest walks in the space of possible metabolic transformations . This study also estimated a typical metabolic substrate can in principle be converted into any of the 20 different products in just one step . This quickly adds up to a very large number of biochemically feasible paths connecting metabolites to each other . However , this exponential growth does not necessarily result in a small world universal metabolic network . Indeed , evolutionary optimization leaves just a tiny fraction of these biochemically feasible reactions to be realized in any organism . The universal metabolic network formed by the union of all organism-specific metabolic networks is thus dramatically sparser than the set of all reactions allowed by the basic rules of biochemistry . Thus , as demonstrated in Ref . [8] and the present study , the number of metabolites one could generate in N steps starting from a small core network and using KEGG-listed metabolic reactions instead of expanding as grows with N much more slowly ( algebraically ) . The overall picture consistent with both our observations and those of Ref . [17] is that exponentially large , supercritical tree of all possible biochemical transformations is first pruned to an evolutionary optimized critical universal network out of which individual organisms select a subset of reactions necessary to accomplish their metabolic goals: that is to utilize nutrients in their environment and generate metabolic targets necessary for their operation . Simplified “toy” models based on artificial chemistry reactions have a long history of being used to reveal fundamental organizational principles of metabolic networks: In our study we used the real-life ( even if incomplete and sometimes noisy ) metabolic universe of all reactions in the KEGG database . The only simplifying approximations remaining in the new most realistic version of the toolbox model is random selection of metabolic targets in the course of evolution and easy availability of any subset of KEGG reactions for horizontal transfer . Both these approximations can be relaxed in later versions of the model . Another promising direction is to extend the toolbox model to artificial chemistry universal networks of Refs . [18] , [19] , [20] . While taking away from the realism of the model such extensions might help to broaden our intuition about what topological properties of the universal network determine the scaling properties of its species-specific subnetworks .
The universal network used in our study consists of the union of all reactions listed in the KEGG database . The directionality of reactions and connected pairs of metabolites were inferred from the map version of the reaction formula: ftp . genome . jp/pub/kegg/ligand/reaction/reaction ? mapformula . lst . The universal network with linearized pathways used to construct Figure 2 and Figure 3 consists of 1813 metabolites upstream of pyruvate and 2745 reaction edges out of which 1782 are irreversible and 963 are reversible . The metabolic network with branched and cyclic pathways used to construct Figure 5–9 consists of 1861metabolites located downstream from the central metabolism and reachable from it by the scope expansion algorithm of Ref . [8] . It has 2819 reactions out of which 1402 are irreversible and the remaining 1417 are reversible . Table 1 and Table 2 shows the statistics for the number of substrates and products of these reactions . The list of core metabolites is obtained from KEGG Pathways Modules in the category “central carbohydrate metabolism” and extended with “currency” metabolites including water , ATP and NAD . Simulations were done in Matlab and Octave .
|
It has been previously reported that in prokaryotic genomes the number of transcriptional regulators is proportional to the square of the total number of genes . We recently offered a general explanation of this empirical powerlaw scaling in terms of the “toolbox” model in which metabolic and regulatory networks co-evolve together . This evolution is driven by horizontal gene transfer of co-regulated metabolic pathways from other species . These pathways are part of a larger “universal” network formed by the union of all species-specific networks . In the present work we address the question of how topological properties of this universal network influence the powerlaw scaling of regulators in the toolbox model . We also generalize its rules to include reactions with multiple substrates and products , branched and cyclic metabolic pathways , and to account for optimality of metabolic pathways . The main conclusion of our analytical and numerical modeling efforts is that the quadratic scaling is the robust feature of the toolbox model in a broad range of universal network topologies . They also demonstrate why , in spite of “small-world” topology , real-life metabolic networks are characterized by a broad distribution of pathway lengths and sizes of regulons in regulatory networks .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/microbial",
"evolution",
"and",
"genomics",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"genetics",
"and",
"genomics/comparative",
"genomics",
"physics/interdisciplinary",
"physics",
"computational",
"biology/transcriptional",
"regulation",
"evolutionary",
"biology/evolutionary",
"and",
"comparative",
"genetics",
"microbiology/microbial",
"evolution",
"and",
"genomics",
"computational",
"biology/evolutionary",
"modeling",
"genetics",
"and",
"genomics/gene",
"function",
"computational",
"biology/metabolic",
"networks",
"evolutionary",
"biology/bioinformatics",
"computational",
"biology/systems",
"biology"
] |
2011
|
A Toolbox Model of Evolution of Metabolic Pathways on Networks of
Arbitrary Topology
|
Life has effectively colonized most of our planet and extremophilic organisms require specialized enzymes to survive under harsh conditions . Cold-loving organisms ( psychrophiles ) express heat-labile enzymes that possess a high specific activity and catalytic efficiency at low temperatures . A remarkable universal characteristic of cold-active enzymes is that they show a reduction both in activation enthalpy and entropy , compared to mesophilic orthologs , which makes their reaction rates less sensitive to falling temperature . Despite significant efforts since the early 1970s , the important question of the origin of this effect still largely remains unanswered . Here we use cold- and warm-active trypsins as model systems to investigate the temperature dependence of the reaction rates with extensive molecular dynamics free energy simulations . The calculations quantitatively reproduce the catalytic rates of the two enzymes and further yield high-precision Arrhenius plots , which show the characteristic trends in activation enthalpy and entropy . Detailed structural analysis indicates that the relationship between these parameters and the 3D structure is reflected by significantly different internal protein energy changes during the reaction . The origin of this effect is not localized to the active site , but is found in the outer regions of the protein , where the cold-active enzyme has a higher degree of softness . Several structural mechanisms for softening the protein surface are identified , together with key mutations responsible for this effect . Our simulations further show that single point-mutations can significantly affect the thermodynamic activation parameters , indicating how these can be optimized by evolution .
One of the most intriguing problems in biology regards the molecular mechanisms involved in adaptive capabilities for life in extreme environments . Cold-adapted organisms have an extraordinary ability to grow in and colonize environments where the temperature is close to the freezing point of water . From the viewpoint of chemical kinetics , a key problem with lowering the temperature is that the enthalpy of activation gives rise to an exponential decrease in enzyme reaction rates according to transition state theory ( 1 ) Here , krxn is the reaction rate and T the temperature , κ is a transmission coefficient , k and h are Boltzmann's and Planck's constants , respectively , and ΔG‡ is the free energy of activation . The latter quantity can be decomposed into entropic ( −TΔS‡ ) and enthalpic ( ΔH‡ ) contributions and decreasing the temperature from 37°C to 0°C typically results in a 20–250 fold reduction of the activity of a mesophilic enzyme [1] . Survival at low temperatures thus requires that the enzyme kinetics can be adapted to avoid this problem and also that protein stability is maintained in a cold environment . As a strategy to cope with the strong temperature dependence of the reaction rates , psychrophiles synthesize heat-labile enzymes possessing a high specific activity and catalytic efficiency at low temperatures [2] , [3] , [4] . It is thus well established that cold-adapted enzymes generally have reduced thermal stability compared to mesophilic orthologues , presumably to counteract the increase in structural rigidity at lower temperatures [1] , [5] , [6] . However , the change in structural stability does not seem to follow any general rule , but is rather a combination of several factors [5] . More remarkable , however , is the seemingly universal characteristic that catalyzed reactions of cold-adapted enzymes have a lower enthalpy and a more negative entropy of activation than their mesophilic and thermophilic counterparts [1] , [5] , [6] . Overall activation free energies , on the other hand , are usually similar around room temperature [5] . The lower activation enthalpy thus makes the rate less temperature dependent ( equation ( 1 ) ) and is believed to be the primary adaption in psychrophilic enzymes [2] , [7] , [8] . It has long been proposed that cold-adaptation originates from increased flexibility of the active site [6] , which could hypothetically yield lower activation enthalpies at the expense of requiring more ordering of substrates and the active site , as the reaction barrier is surmounted ( i . e . , a more negative ΔS‡ ) . However , there seems to be no strong experimental support for this hypothesis and , e . g . , X-ray analysis of cold- and warm-active trypsin did not indicate any overall flexibility differences between the two enzymes [9] . Moreover , recent computer simulations of differently adapted citrate synthases showed that the flexibility of the highly conserved active site residues was virtually identical . Instead it was found that differences in protein stiffness outside of the active site appear to be correlated with differences in thermodynamic activation parameters [10] . The origin of catalytic rate optimization in cold-adapted enzymes , in terms of actual structure-function relationships , thus remains rather obscure . Understanding such relationships would not only provide information regarding the evolutionary adaption processes , but potentially also enable rational design of enzymes adapted to low temperature . Computer simulations could provide a unique way of analyzing the reaction energetics of differently adapted enzyme orthologs . However , in order for such a strategy to be viable several criteria must be met . First , analysis of indirect or circumstantial factors ( flexibility , electrostatics , hydrophobicity etc . ) alone does not suffice for obtaining conclusive evidence . Instead reliable free energy profiles along the reaction pathway must be obtained with high precision . Second , the crucial activation enthalpy-entropy balance for different enzymes must be reproduced by the simulations and the only way to do this is to computationally obtain Arrhenius plots for the activation free energy versus temperature . This involves calculating a large number of free energy profiles at different temperatures so that activation enthalpies and entropies can also be extracted with high precision . Clearly , such extensive sampling by molecular dynamics ( MD ) simulations precludes the use of most standard QM/MM approaches , but the empirical valence bond ( EVB ) model [11] , [12] . provides a very efficient method for this purpose . Third , provided that the experimentally observed activation enthalpy-entropy balances are captured by the simulations , it must be possible to decompose these into their underlying energy components and ultimately translate them into differences between the enzyme 3D structures and fluctuations . Here , we report extensive MD/EVB free energy simulations that yield high precision Arrhenius plots for the reactions of psychrophilic and mesophilic trypsins . The calculations reproduce both experimental rates at room temperature and the characteristic relationships between activation enthalpy and entropy for the orthologous salmon and bovine enzymes . The relationship between these parameters and the 3D enzyme structures is reflected by significantly different internal protein energy changes during the reaction . This effect originates from outside of the active site where the cold-adapted salmon enzyme has a higher degree of softness , which is evident from the corresponding potential energy term . We also identify key residues for which simulations predict significantly altered thermodynamic activation parameters upon mutation .
Atomic coordinates for psycrophilic and mesophilic trypsin were obtained from the crystallographic structures with PDB entries 1BZX [13] and 3BTK [14] , respectively . All EVB calculations were performed with the molecular dynamics package Q [15] using the OPLS2005 all-atoms force field [16] , [17] . Additional simulations details are given in Text S1 . The EVB reaction surface was calibrated using the imidazole catalyzed methanolysis of formamide in water [18] as a reference reaction ( Text S1 ) . The EVB free energy profiles were calculated using the free energy perturbation ( FEP ) umbrella sampling approach described elsewhere [11] , [12] . Each enzyme and water reaction free energy profile involved 500 ps of MD simulation and compromised 51 discrete FEP steps . Thermodynamic activation parameters were obtained from Arrhenius plots based on simulations at eight different temperatures ( 275–310 K ) . At each temperature point 100 and 150 independent FEP simulations were carried out , resulting in a total simulation time of 408 and 612 ns for salmon and bovine trypsin , respectively . In addition 100 ns simulation time was performed at the reactant and transition state at 300 K for both systems . Enzyme mutations were created using the builder tool in Mastro 9 . 1 ( Schrödinger , LLC , New York , NY , 2011 ) . The mutated residues were relaxed prior to MD simulation with the clean up geometry tool in Maestro . In order to obtain reliable sampling , the simulations were repeated 20–60 times at each temperature ( 275–310 ) for the mutated model systems .
We used the reactions of the mesophilic bovine trypsin ( BT ) and the psychrophilic anionic salmon trypsin ( AST ) as models to examine the temperature dependence of reaction rates for differently temperature adapted enzymes . The energetics of the rate-limiting formation of the tetrahedral intermediate , using a Cys-Lys-Ala tripeptide as substrate , was calculated by the MD/EVB approach [11] , [12] . The results from these simulations at 300 K are shown in Fig . 1a as free energy profiles along the reaction coordinate for the two enzymes . The corresponding free energy profile for the reference reaction used to calibrate the EVB potential ( see Text S1 ) , i . e . , imidazole catalyzed formation of the tetrahedral intermediate in water [18] , is also shown . In order to attain a sufficiently high precision the calculations were averaged over up to 150 independent runs at each temperature ( see below ) . The calculated activation energies at 300 K of 18 . 2±0 . 2 kcal/mol and 19 . 0±0 . 2 kcal/mol for AST and BT , respectively , are in excellent agreement with the substrate dependent barrier of 15–20 kcal/mol [19] . This difference in activation free energies translates into a 4-fold increase in kcat for AST when compared to BT , which is in remarkable good agreement with experiments that shows 2- to 4-fold increase depending on the temperature [22] . The simulations also clearly demonstrate the large catalytic effect on the reaction for both enzymes . Compared to the uncatalyzed hydrolysis reaction in water , the transition state is found to be stabilized by over 13 kcal/mol [18] , [23] . With respect to the imidazole catalyzed reference reaction in solution the corresponding stabilization is about 7 kcal/mol [18] . Since the catalytic rates of the two trypsins at room temperature are well reproduced by the MD/EVB simulations , we can now turn to examine their temperature dependence . Eight different temperatures were chosen in the range of 275 to 310 K and 100–150 independent free energy profile calculations were carried out at each temperature to obtain high precision Arrhenius plots . Activation entropies and enthalpies were then extracted by linear regression from plots of ΔG‡/T vs . 1/T . The temperature dependence of the activation free energies is shown in Fig . 1b and it can immediately be seen that the psychrophilic enzyme ( AST ) has a significantly smaller slope than the mesophilic counterpart ( BT ) . The calculated activation parameters for BT are ΔH‡ = 20 . 4 kcal/mol and ΔS‡ = 3 . 5 e . u , while the corresponding values for AST are ΔH‡ = 9 . 9 kcal/mol and ΔS‡ = −27 . 5 e . u . This is thus a remarkable example of enthalpy-entropy compensation where the large differences in ΔH‡ are balanced by -TΔS‡ contributions at 300 K of −1 . 4 and +8 . 3 kcal/mol for BT and AST , respectively ( Table 1 ) , to yield similar activation free energies . It should be noted that an increase in the activation free energy of 1 kcal/mol directly translates into a 5-fold decrease in kcat . The fact that both the absolute rates at 300 K and the characteristic balance between activation enthalpy and entropy for the mesophilic and psychrophilic enzymes are reproduced by the computer simulations is also remarkable and raises the question of what the structural origin of this effect really is . As far as energetics is concerned it is relatively straightforward to identify the source of the difference in activation enthalpy between the two enzymes . Since ΔH‡ = ΔU‡+pΔV‡ , and the pressure-volume term is completely negligible , the activation enthalpy is determined by the corresponding change in internal ( total ) energy of the system . The latter can be decomposed into contributions from the reacting fragments ( i . e . , the EVB atoms whose interaction parameters change along the reaction ) , their interactions with the surrounding protein and solvent , and the interactions within the surrounding environment ( 2 ) Here , the subscripts r and s denote the reacting fragments and surroundings ( the protein and solvent included in the simulations ) , respectively . The last term of equation ( 2 ) involves very large energies , since it pertains to a huge number of interactions within the surrounding protein and solvent , making it practically impossible to obtain a converged value for this quantity directly from the MD simulations . However , since both ΔH‡ and can be evaluated from the trajectories with sufficiently high precision we can still get an accurate estimate of all the terms in equation ( 2 ) . Table 1 shows this breakdown of the energetics which immediately reveals that the source of the decreased activation enthalpy in the cold-adapted enzyme is not associated with a more favorable term . Instead it is a significantly lower value of that is responsible for the decrease in ΔH‡ . Hence , while the internal energy change involving the reacting groups is similar , the contribution from the surroundings is predicted to be about 9 kcal/mol more favorable for AST than BT . It would be desirable to further decompose into protein-protein , protein-water and water-water interaction contributions according to ( 3 ) but , again , the energies involved are too large to allow converged direct calculations of these averages . However , from the viewpoint of locality it is reasonable to expect that the two first terms involving protein interactions dominate the reduction in or the cold-adapted enzyme . That is , the protein interactions are likely to respond more strongly to the energy change in the active site , associated with climbing the activation barrier , since the active site is primarily embedded in the protein , which in turn is surrounded by water . At any rate , we can conclude that the reduction of activation enthalpy in the cold-adapted enzyme originates from interactions outside of the active site . This is perhaps not so strange since all residues surrounding the substrate are conserved between the two proteins , making it more likely that energetic differences are to be found farther away . The fact that the energy cost reflected by the term is lower in the cold-active than the warm-active trypsin further suggests that the surroundings of the active site are effectively softer in the salmon enzyme . In this respect , the term “softness” can be more precisely defined than protein flexibility in general , as it refers to the change in potential energy of the surroundings of the active site as the system moves along the reaction coordinate form reactants to transition state . This potential energy change can thus be viewed as reflecting an effective force constant of the surroundings , which is stiffer in the warm-active enzyme and softer in the cold-active . This brings us back to the possible role of protein flexibility in cold-adaptation . Cold-adapted enzymes are often assumed to benefit from higher flexibility to deal with the decrease in chemical rates and altered structural rigidity at low temperatures . Since the activation entropy is also more negative than for mesophilic homologs , this could be interpreted in terms of an increased flexibility of the active site in the reactant state [8] . This proposal was , however , not supported by Bjelic et al . who evaluated the positional root-mean-square fluctuations ( RMSF ) of the key residues in the active site of different temperature-adapted citrate synthases [10] . They demonstrated that the active site and substrate mobilities were virtually identical and found no indication of the cold-adapted enzyme having larger active site RMSFs compared to the heat-adapted enzymes . The fluctuations obtained with a spherical boundary model were also found to be virtually identical to those obtained with a much larger simulation system simulated using periodic boundary conditions . It should be noted that the present calculations were carried out with the entire protein immersed in a spherical droplet of water ( Fig . S1 ) . To further examine the flexibility hypothesis , we carried out additional 100 ns simulations at both the transition and reactant states for BT and AST . As in Refs . [10] and [24] , we again find that the mobility of the active site is low and practically identical in the two enzymes ( Fig . S2 ) . Furthermore , the overall protein backbone RMSFs are very similar with calculated values of 0 . 65 Å and 0 . 66 Å for BT in the reactant and transition state , respectively , while the corresponding values for AST are 0 . 61 Å and 0 . 65 Å . A plot of the average backbone positional fluctuations versus amino acid sequence ( Fig . 2a ) , however , shows as expected that there are local differences in mobility and that these mainly are found on the protein surface . For example , Tyr97 and Asp150 in AST are significantly more flexible than their corresponding BT residues . Both Tyr97 , situated in the Nβ5-Nβ6 loop , and Asp150 of the so-called autolysis loop are also conserved through different cold-adapted trypsins ( Fig . 3 ) . Moreover , further analysis of the backbone RMSFs shows that the prevalence of residues with high mobility , measured radially from the active site , differs significantly between bovine and salmon trypsin ( Fig . 2b ) . That is , while both enzymes become more flexible further away from the active site , the cold-adapted protein has a markedly higher prevalence of residues with high RMSF values beyond 10 Å from the active site . The conclusion is thus that both enzymes have a relatively rigid core and softer outer regions , but that the surface regions of the cold-adapted enzyme are , at least locally , softer than for the warm-adapted protein . Both the energetic and mobility analysis above strongly suggest that the surface of the cold-adapted enzyme is softer compared to its warm-active counterpart and the key question now is how this could be accomplished . Here , ultra-high resolution ( 0 . 75–1 . 0 Å ) crystal structures of both BT [25] and AST [26] turn out to provide valuable clues since a large number of surface bound water molecules are resolved in these structures . Such water molecules very often allow polar surface side-chains , like those of Asn , Gln , Ser and Thr , to engage in extensive H-bond networks back to the protein surface . Charged surface side-chains , on the other hand , preferentially point out into solution due to their stronger requirement for solvation by bulk water . Analysis of the key mutations identified in Fig . 3 with these high-resolution structures [25] , [26] reveals some basic principles for how the protein surface can be made softer by point-mutations . First , the surface can be softened by disrupting a water mediated H-bond network through mutation of a polar to a less polar residue . This is exemplified in the trypsins by mutation of Asn97 in BT to Tyr97 in AST , which largely abolishes such a network ( Fig . 4a ) . Second , a surface H-bonding network can also be disrupted by mutation of a polar residue into a charged one , since the latter may prefer to interact with bulk solvent . This is exemplified by the mutation of Ser110 in BT to Lys110 in AST ( Fig . 4b ) . A single or few mutations may also completely change one surface H-bond network into another such as the structurally correlated Thr21Glu , Ser150Asp and Val154Lys mutations , which drastically affect the conformation of the autolysis loop and make the active site region more solvent accessible in AST ( Fig . 4c ) . Finally , one can also identify mutations that destabilize the packing of hydrophobic surface patches by mutation of a nonpolar residue into a charged one . This is , e . g . , the case with the mutation of Val90 in BT to Arg90 in AST ( Fig . 4d ) . Of the mutations mentioned above , we will focus on Asn97Tyr and Ser150Asp since both of these are strictly conserved in the cold-adapted trypsins ( Fig . 3 ) and are also the ones showing the largest increase in flexibility compared to the warm-adapted enzyme ( Fig . 2a ) . We thus again calculated free energy profiles at different temperatures to obtain Arrhenius plots for the four cross-species mutations BTN97Y , BTS150D , ASTY97N and ASTD150S , in order to be able to predict their thermodynamic activation parameters . The resulting calculated free activation energies remain essentially unchanged by the mutations ( Table 1 ) , which underscores the general notion that mutations far away from the active site do not significantly affect catalytic rates [27] . However , what is remarkable is that the calculations predict significant changes in ΔH‡ and ΔS‡ for most mutations , but that these are again nearly perfectly compensating . Thus , both the BTN97Y and BTS150D mutations markedly lower the activation enthalpy and make the entropy more negative compared to the wild-type bovine enzyme and they become more like the cold-adapted AST . For the reverse mutations , ASTD150S renders the cold-adapted enzyme more mesophilic-like with a significantly raised ΔH‡ and a more positive ΔS‡ . This is also seen by the predicted effect the autolysis loop structure , which approaches the bovine conformation ( Fig . S3 ) . The ASTY97N mutation , on the other hand , yields relatively smaller effects on both the activation parameters and . This probably just reflects the fact that correlated mutations ( e . g . , ASTS95N ) may be needed to build up the native bovine H-bond network involving the Nβ5-Nβ6 loop ( Fig . 4a ) , so that a single point-mutation does not suffice . Since both of the residues mutated are involved in distinct H-bond networks in the bovine enzyme , which appear to rigidify the surface , it is logical that a single mutation could destroy such a network and make the surface softer . In this respect , it would seem more difficult to conversely rigidify the surface by a single mutation , as in the case of ASTY97N , if that requires the creation of a new H-bond network . It is also noteworthy here , that the single BTN97Y mutation is predicted to yield values of ΔH‡ , TΔS‡ and that are almost identical to those of native AST and kcat is predicted to increase 19-fold at 4°C for this mutation . That mutation of residue 97 , either from BT to AST or vice versa , has a pronounced effect on the backbone mobility of the Nβ5-Nβ6 loop is also evident ( Fig . 5 ) , where Asn consistently reduces positional fluctuations whereas Tyr increases them .
In this work , we have addressed the problem of the structural origin of what appears to be a universal characteristic feature of cold-adapted enzymes , namely a reduced enthalpy and more negative entropy of activation . This was done using Atlantic salmon and bovine trypsin , cold- and warm-active , respectively , as models for the phenomenon . With very extensive all-atom computer simulations , using the EVB method to describe the catalytic reaction , reliable Arrhenius plots for the temperature dependence of the activation free energy could be obtained . It should be pointed out here that there is probably no other feasible way at present to calculate Arrhenius plot from first principles . It is rather remarkable that these simulations clearly reproduce the characteristic balance between activation enthalpy and entropy for cold-adapted versus warm-active enzymes , as well as the higher catalytic rate of AST compared to BT . Nevertheless , the activation free energies are similar at room temperature , thus reflecting a near perfect compensation between the former activation parameters . The origin of the catalytically beneficial low activation enthalpy at low temperatures ( accompanied by a more negative entropy ) is found to be associated not with the active site but with the protein surface . From an evolutionary point of view this is perhaps not so surprising since any mutations in or near an optimized active site are likely to cause drastic rate reductions . What is rather surprising , however , is that it is the softness of the protein-water surface that appears to regulate the activation enthalpy-entropy balance . The simple picture that emerges is thus that the enzymes have a relatively rigid core , where the probability of successful adaptive mutations is low , surrounded by a softer outer matrix ( Fig . 2b ) whose properties can be fine-tuned by evolution . While our earlier study of citrate synthases [10] also identified the same potential energy terms as responsible for the altered activation enthalpy-entropy balance , the structural origin of the effect remained obscure . Here , with the help of ultra-high resolution crystal structures , the actual structural “mechanisms” by which the surface softness is altered could finally be elucidated . Moreover , computational predictions of the effects of surface mutations were made that strongly support these conclusions . In particular , we identified extensive H-bond networks between polar surface groups and bound water molecules in the mesophilic enzyme that rigidify the surface , and several cold-adaptive mutations soften it by disrupting these networks . However , there are also examples of mutations in the cold-adapted enzyme ( e . g . , Val90Arg and Val154Lys ) that appear to destabilize local hydrophobic surface patches . In view of the above findings , it is understandable that structural bioinformatics analysis has not yielded any consistent common descriptors of cold-adapted enzymes [5] . That is , since we identified several different types of mechanisms for surface destabilization and since the effects are often very local it is unlikely that there are distinct global descriptors that can capture them . It is also noteworthy that the computer simulations predict that the enzyme ΔH‡ and ΔS‡ activation parameters can change significantly due to a single surface mutation . Such phenomena have , in fact , also been experimentally observed for other orthologous enzymes [28] . At first this may seem as a mysterious long-range effect on catalysis , but it should be emphasized that the activation free energies , and hence the catalytic rates , remain essentially unchanged . Instead it is the surface properties that are altered by such point mutations and mutations that soften the surface can apparently reduce the activation enthalpy of the catalyzed reaction at the expense of a more negative activation entropy . Such surface alterations are , however , beneficial for adaptation to low temperatures as they simultaneously make the rate more temperature insensitive and presumably also counteract the structural rigidity imposed by the reduction in temperature . It is , of course , important to also try to address the generality of the present findings . In this respect , it should first be recalled that the characteristic trends with regard to activation enthalpy and entropy for cold-adapted enzymes appear to be completely universal , without known exceptions [5] . Two different types of enzymes ( citrate synthases [10] and trypsins ) have now been analyzed by extensive reaction simulations , which reproduce the observed behavior of warm- and cold-adapted orthologs , and which identify protein surface softness as the key variable . There is also other circumstantial evidence pointing towards surface properties , and flexibility in particular [29] , [30] . Furthermore , the active site residues are basically always conserved between highly similar orthologous warm- and cold-adapted enzymes , which makes the idea that active site fluctuations would be substantially different very unlikely . It does therefore not seem far-fetched to assume that the difference in protein surface properties found here is likely to be a general feature of cold-adapted enzymes . It is further noteworthy that significantly altered kinetics and thermostability due to amino acid changes at a few sites distant from the active site have also been observed in dehydrogenases that are rate-limited by conformational changes rather than by chemistry [29] , [30] , as in the case of trypsin [19] . While the effects discussed herein pertain to the catalytic rates of the enzymes , their influence on thermostability is more difficult to assess . As mentioned , reduced thermostability is also an apparent universal characteristic of enzymes adapted to cold environments The net stability of folded proteins normally on the order of 10 kcal/mol , and is the result of large compensatory contributions . This , of course , makes it difficult to point out a single factor to explain differences in thermostability . However , previous studies of trypsin [31] indicate that the stability of a few loops and the C-terminal helix are important factors when explaining the difference in thermostability of cold- and warm-adapted trypsins . While our results also identify the same loop regions as important for adaptation to cold , experimental characterization is needed to examine whether these amino acid substitutions only change the catalytic rate or whether they affect thermostability as well .
|
Cold-adapted organisms require specialized enzymes to maintain functional integrity at low temperatures , and psychrophiles express heat-labile enzymes that possess a high specific activity and catalytic efficiency at low temperatures . The high catalytic rates are achieved by enzyme adaptations yielding lower activation enthalpies and entropies than for mesophilic homologs , thereby solving the problem of the exponential rate decrease with falling temperature . However , the structural mechanisms behind this universal property of cold-adapted enzymes remain unknown . By extensive computer simulations , which reproduce both the experimental reaction rates and the characteristic temperature dependence of activation free energies , we show that it is the softness of the protein-water surface that regulates the activation enthalpy-entropy balance . Structural mechanisms behind this phenomenon are identified and our simulations show that single mutations can significantly affect the thermodynamic activation parameters , indicating how these can be optimized by evolution .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"physics",
"biochemistry",
"computer",
"and",
"information",
"sciences",
"computer",
"modeling",
"biocatalysis",
"thermodynamics",
"catalysis",
"free",
"energy",
"biology",
"and",
"life",
"sciences",
"chemistry",
"physical",
"sciences",
"biophysics",
"biophysical",
"simulations"
] |
2014
|
Protein Surface Softness Is the Origin of Enzyme Cold-Adaptation of Trypsin
|
Epstein-Barr virus ( EBV ) infection often occurs in early childhood and is asymptomatic . However , if delayed until adolescence , primary infection may manifest as acute infectious mononucleosis ( AIM ) , a febrile illness characterised by global CD8+ T-cell lymphocytosis , much of it reflecting a huge expansion of activated EBV-specific CD8+ T-cells . While the events of AIM have been intensely studied , little is known about how these relate to asymptomatic primary infection . Here Gambian children ( 14–18 months old , an age at which many acquire the virus ) were followed for the ensuing six months , monitoring circulating EBV loads , antibody status against virus capsid antigen ( VCA ) and both total and virus-specific CD8+ T-cell numbers . Many children were IgG anti-VCA-positive and , though no longer IgM-positive , still retained high virus loads comparable to AIM patients and had detectable EBV-specific T-cells , some still expressing activation markers . Virus loads and the frequency/activation status of specific T-cells decreased over time , consistent with resolution of a relatively recent primary infection . Six children with similarly high EBV loads were IgM anti-VCA-positive , indicating very recent infection . In three of these donors with HLA types allowing MHC-tetramer analysis , highly activated EBV-specific T-cells were detectable in the blood with one individual epitope response reaching 15% of all CD8+ T-cells . That response was culled and the cells lost activation markers over time , just as seen in AIM . However , unlike AIM , these events occurred without marked expansion of total CD8+ numbers . Thus asymptomatic EBV infection in children elicits a virus-specific CD8+ T-cell response that can control the infection without over-expansion; conversely , in AIM it appears the CD8 over-expansion , rather than virus load per se , is the cause of disease symptoms .
Epstein-Barr Virus ( EBV ) is a ubiquitous gamma herpesvirus associated with occasional severe primary infections , several malignancies and significant pathology in immunosuppressed hosts . It does not , however , cause significant morbidity in the majority of those infected . In The Gambia most children are infected during childhood , in contrast to most developed countries where the majority of primary infections occur at a later age , often in adolescence [1 , 2] . It is estimated that between a quarter and up to three quarters of those infected in adolescence will develop a sometimes-severe disease , AIM [3 , 4] . Paradoxically , those infected during childhood tend to have minor self-limiting illnesses that often go undetected [5] . It is not fully understood why individuals that contract EBV during childhood are usually asymptomatic and do not develop AIM . Of note , most of the published literature regarding the immunopathogenesis of primary EBV infection is derived from studies of AIM , rather than asymptomatic infections . Many studies in adults have characterised cellular immune responses ex vivo during AIM , both among CD8+ and to a lesser extent CD4+ T-cell subsets [6–13] . The EBV-specific CD8+ T-cell response is hugely amplified , such that total CD8+ T-cell numbers in the blood may reach five to ten-fold higher than usual . Indeed individual lytic antigen reactivities ( typically against epitopes within the immediate early ( IE ) and some early ( E ) proteins ) can account for up to 40% of the highly expanded CD8+ T-cell population , and individual latent antigen reactivities ( typically against epitopes from the EBV nuclear antigen 3A , 3B , 3C family ) occupying up to 5% . These CD8+ T-cells display a phenotype consistent with recent antigen stimulation , being perforin-positive with direct ex vivo cytotoxic function [14–16] and express the activation marker CD38 and cell cycling marker Ki-67 [8 , 10 , 14 , 17] . What drives these expansions in AIM is unclear , but factors such as an initial lack of natural killer cell control [18] , cross-reactive recognition by clonotypes in pre-existing CD8+ T-cell memory [19] , and genetic factors [20–22] , including polymorphism of the IL-10 promoter [23] , have all been proposed . Whether the cellular response to early EBV infection in asymptomatic children shows features of disruption similar to those described in AIM has been difficult to investigate , mostly because donors without clinical symptoms can’t be readily identified . However , an understanding of how EBV infection is controlled with minimal immunopathology , i . e . without the development of AIM , is important , as AIM is associated with an increased risk of EBV-associated diseases such as EBV-positive Hodgkin lymphoma and multiple sclerosis [24 , 25] . Of the few published studies on asymptomatic primary EBV infections , Silins et al identified four adult patients undergoing silent seroconversion within a vaccine trial [26] . Interestingly some of these had high EBV loads yet did not have massive T-cell expansions and , where studied , most did not have the distorted T-cell receptor ( TcR ) repertoires usually seen in AIM; however , their EBV-specific T-cell response was not studied . As to infections during childhood , an early report suggested that this occurs with a serological picture distinct from AIM and without obvious lymphocytosis [27] , while another small study of children aged 20–35 months detected EBNA3A , B and C-specific CD8+ T-cell responses in the blood without addressing issues of viral load or hyper-expansion [28] . A detailed study of the EBV-specific immune response and EBV dynamics in asymptomatically infected children utilising modern immunological and virological tools is lacking . The present work follows a cohort of 114 Gambian children longitudinally over six months , using serology and viral load to determine EBV status . It describes the EBV-specific CD8+ T cell responses in those that have had a relatively recent primary EBV infection without any obvious clinical history , and additionally captures six children undergoing silent seroconversion .
Gambian children of an age , 14–18 months , likely to be undergoing primary EBV infection were recruited for study from an infant vaccination clinic . Blood samples were collected from the children at baseline ( visit one ) , they were vaccinated one week later against Diptheria , Tetanus , Pertussis , Hepatitis B and Haemophilus influenzae B ( visit two ) . A five-millilitre blood sample was collected a week later ( visit three ) primarily to monitor vaccine responses and a further sample at six months ( visit four ) . Of 120 children screened , six were ineligible due to concurrent illnesses or malnutrition at screening whereas 114 were enrolled , of which 99 remained in the study until completion at six months . The study dropouts did not significantly differ in age , sex , weight , haemoglobin , leucocyte and lymphocyte counts , compared to those who continued to participate ( S1 Table ) . Initially children were tested for their EBV-VCA antibody status and categorised into one of three groups: non-infected ( IgM−IgG− ) , established infection ( IgM−IgG+ ) or very recently infected ( IgM+IgG+/− ) . At visit one , 71 children had established EBV infection as judged by the presence of VCA-specific antibodies of IgG but not IgM class ( Fig . 1 ) . Another four children showed evidence of recent infection , with three having IgM VCA-specific antibodies only and one having both IgM and IgG VCA-specific antibodies . The remaining 39 children appeared to be non-infected , having no detectable VCA-specific antibodies or viral genomes in their peripheral blood mononuclear cells ( PBMCs ) . At visit four ( six months later ) 17 of these 39 initially EBV non-infected donors had become VCA IgM−IgG+ , another two had become VCA IgM+IgG+ , 13 remained VCA IgM−IgG− , while seven dropped out of the study . The four initially IgM+ children had now become IgM- and had developed VCA-specific IgG . All children , including those with VCA-specific IgM antibodies , were asymptomatic for classical symptoms of AIM ( fever , lymphadenopathy , malaise ) prior to recruitment and at subsequent visits , based on maternal history and clinical evaluation . Overall , 62% of children showed serological evidence of being EBV infected at baseline , rising to 86% among those remaining in the study six months later . An analysis of VCA IgG titre in a subset of 25 pairs of samples from children at visit one and four showed no significant difference in titre ( p = 0 . 774 , S1A Fig ) . Since acquisition of EBV in similar African cohorts begins between six and twelve months after birth [29 , 30] , it is likely that at least some children who were IgG reactive to VCA within the cohort were infected with EBV within the last six months prior to recruitment . To examine for evidence of recent infection among these donors , EBV genome loads in PBMCs were measured by qPCR analysis . Fig . 2 shows viral genome load data from PBMCs collected from 70 IgM−IgG+ donors at baseline , 58 of these donors six months later , and six very recently infected IgM VCA reactive donors ( some of whom also had VCA-specific IgG antibodies ) . Genome load data from Caucasian adolescent patients undergoing primary symptomatic EBV infection , AIM , assessed using the same qPCR assay are also included for comparison . Almost all of the children who had IgG antibodies to VCA at baseline had high EBV genome levels in their PBMCs , ranging up to two million genomes per million PBMCs . Indeed , for the whole cohort of IgM− IgG+ children sampled at visit 1 , the median load of 3000 genomes per million PBMCs ( IQR 900 to 8000 genomes per million PBMCs ) was not significantly different to that observed in AIM patients . When PBMCs from these same children were assessed for genome loads six months later , a narrower range of values was found and the median load was eight to tenfold lower than at baseline . The decreasing virus loads observed over these two time points suggests that these donors were establishing their virus host balance following recent EBV infection . Comparing virus load to VCA IgG titre in a subset of 25 children at the two time points showed no correlation between load and titre ( S1B and C Fig ) . When the samples from the six IgM+ donors were analysed , these also showed high EBV genome loads with a median value of 8000 genomes per million PBMC , slightly higher than but not significantly different from loads in IgM-IgG+ positive children measured at baseline and again similar to that seen in AIM patients . Such data are consistent with these IgM+ children having been very recently EBV-infected . Primary symptomatic infection with EBV is associated with dramatic expansions in the frequency and absolute number of lymphocyte subsets , especially CD8+ lymphocytes [6] . Evidence of disruptions to the lymphocyte compartments of the three groups of children ( IgM− IgG− , IgM− IgG+ or IgM+ IgG+/− ) were studied by determining absolute numbers of lymphocytes within the CD3 , CD4 , CD8 and CD19 subsets . Fig . 3 shows results of absolute cell counts and , for comparison , counts of equivalent subsets from six Caucasian AIM patients . Dramatic expansions of the CD3+ and CD8+ ( but not CD4+ ) T-cell numbers and a contraction of B cell numbers were seen in AIM patient samples . However , no obvious or significant expansion of lymphocyte subsets was observed when comparing uninfected children with the two EBV-infected groups . Furthermore , no significant changes in the CD4:CD8 ratios were observed in PBMCs from a subset of 14 children over the six month study period ( p = 0 . 76 , S2 Fig ) . This indicated that there was little disruption to peripheral lymphocyte subsets in children at these different stages of asymptomatic EBV infection . To further understand the dynamics of CD8+ T-cell responses and virus loads at early stages of infection , MHC-class I tetramers were used to assess the frequency of EBV-specific T-cells in the PBMCs of children with VCA-specific IgM− IgG+ antibodies at baseline and six months later and compared to their virus loads at these time points . Children with relevant EBV-specific responses were identified by screening PBMCs from visit one for responses by ELISpot to pools of peptides containing peptide-epitopes known to be presented by HLA types frequent within the Gambian population . This identified 14 children with responses which could be assessed with HLA-A*0201 , HLA-B*0801 or HLA-B*3501MHC class I tetramers . Virus loads from this subset of donors were representative of the overall population shown in Fig . 2 at the two time points and their loads were significantly decreased at the second time point ( Fig . 4A ) . Frequencies of EBV-specific responses in each of the children’s paired visit one and visit four PBMCs samples were obtained by staining with an appropriate tetramer which would identify an immunodominant response . These tetramers presented epitopes derived from the immediate early lytic cycle protein BZLF1 protein , either the HLA-B*0801 presented RAKFKQLL peptide or the B*3501 presented EPLPQGQLTAY peptide , or from the early lytic cycle protein BMLF1 , the HLA-A*0201 presented GLCTLVAML peptide . Analysis of the frequency of tetramer-specific CD8+ T-cells in PBMCs from this group is shown in Fig . 4B . This shows that at the first time point EBV-specific responses are made , but their frequency of up to 2 . 5% of CD8+ T cells is not obviously increased as is seen in AIM patients , in whom previous reports have documented up to 40% of the total CD8+ T-cell population as being EBV-specific ( Callan et al . 1998 ) . Analysis of the visit four time point , collected six months later , showed that responses were still present , however they were on average significantly lower than those detected at visit one . This reduction in EBV-specific T-cell numbers with time , coupled with a falling virus load illustrates a pattern consistent with recent EBV infection and establishment of a long term carrier state in these children . In all but one of these 14 children there were sufficient cell numbers to examine the changing phenotype of the T-cell response over time . Both total CD8+ T-cells and tetramer-positive CD8+ T-cells were examined for markers known to be expressed by CD8+ T-cells responding to acute EBV infection; namely activation status as defined by co-expression of CD38 and HLA-DR , cycling status as defined by the expression of Ki-67 , and apoptosis sensitivity as indicated by loss of Bcl-2 , which is down regulated in activated EBV-specific cells in AIM patients . Of note CD38 and HLA-DR co-staining for activation status was used as lymphocytes from young children may constitutively express CD38 which is progressively lost with age [31] . The graphs on the left hand side of Fig . 5 summarises the results of this analysis while the flow cytometry plots show representative analyses for each stain at the two time points; note that the top flow plots combine data from all CD8+ T cells in black , with the tetramer-positive population in red , while the middle and bottom flow plots show the Ki-67/tetramer and Bcl-2/tetramer profiles . At the first time point there were significantly more HLADR+CD38+ ( p = 0 . 002 ) and Ki-67+ ( p = 0 . 01 ) EBV-specific CD8+ T-cells compared to the total CD8+ T-cell population , while tetramer positive CD8+ T-cells expressed lower levels of the anti-apoptotic marker Bcl-2 ( p <0 . 0001 ) . The percentage of HLADR+CD38+ EBV-specific CD8+ T-cells in the IgM−IgG+ children declined significantly over time ( p = 0 . 003 ) , whereas cellular expression of Bcl-2 significantly increased ( p = 0 . 013 ) , with a non-significant decline seen in frequency of cells expressing Ki-67 . Again , the phenotypic analysis of these VCA antibody IgG+ donors suggests recent EBV infection and the establishment of a long term carrier state . To get the clearest understanding of the early events in EBV-specific CD8+ T-cell responses in children undergoing asymptomatic infection , the EBV tetramer-specific responses and phenotype of these cells were examined in children with IgM VCA-specific antibodies . Of the six IgM+ children found in this study , three were suitable for study with the tetramer panel for frequency and phenotype analysis at the different time points . However the three other donors did not show expanded numbers of CD8+ T cells ( Fig . 3 ) and showed , at most , small changes in the frequencies of CD8+ T cells from 40 . 9% , 36 . 8% and 50% at visit one when they were IgM+ , to 27 . 3% , 32 . 5% and 50 . 1% at visit four respectively when they were IgG+ . Of the samples from children which could be analysed with tetramers , 082 and 007 were seronegative at the first time point , but had developed IgM+ VCA-specific antibodies at six months . Although there was no significant increase in the absolute numbers of CD8+ T-cells before and after EBV acquisition , both donors 082 and 007 showed increases in the frequencies of CD3+ CD8+ lymphocytes from baseline values of 18 . 2% and 19 . 8% to values of 50 . 3% and 40 . 8% respectively six months later . As shown in Fig . 6 , tetramer analysis of samples at baseline when the donors 082 and 007 were VCA seronegative showed no tetramer-specific staining . However at six months , when the children were VCA IgM+ IgG+ , the HLA-B*0801 donor 082 made a substantial response with 6 . 9% of their CD8+ T-cells being specific to the RAK-epitope while the HLA*0201 donor 007 made a small response of 0 . 43% CD8+ T-cells to the GLC-epitope . In both cases , tetramer positive cells were highly activated with the majority of EBV-specific cells co-expressing HLA-DR and CD38 . In both of these donors a substantial frequency of EBV-specific cells were in cycle; interestingly in the case of child 007 a large proportion of non-tetramer-specific CD8+ T-cells were also in cycle , likely representing other EBV-specificities , possible bystander activation or coincident infection with another pathogen . In the EBV-specific CD8+ T-cells from these children there was little if any expression of Bcl-2 ( Fig . 6 ) . The third child studied , the HLA-B*0801 donor 061 , had VCA-specific IgM+ antibodies at the first time point and made a substantial response to the RAK epitope which allowed serial monitoring of this response ( Fig . 7 ) . The frequency of PBMC CD8+ T cells did not change over time with levels being 38% at visit one , 37% at three two weeks later , and 37% at visit four with no significant expansion of absolute numbers ( see Fig . 1 ) . However , as shown in Fig . 7 , tetramer analysis at the first time point detected a large RAK-specific response of 15 . 9% of CD8+ T-cells , with 32 . 2% of these being activated , 14 . 8% in cycle and few expressing Bcl-2 . After two weeks the tetramer-specific frequency had decreased to 5 . 4% of CD8+ T-cells with no associated decrease in activation marker and Ki-67 expression , nor increase in Bcl-2 expression by the tetramer-specific cells . By six months , the frequency of RAK-specific cells had decreased to 1 . 75% of CD8+ T-cells . Here the RAK-specific cells phenotypically resembled the EBV-specific cells from the VCA-IgG+ donors at visit 4 , with only 5% expressing activation markers , 5% of cells in cycle and 15% expressing Bcl-2 ( Fig . 7 ) . Overall these findings suggest that in asymptomatic primary EBV infection , the frequency of activated EBV-specific cells in the CD8 population can be substantial but this occurs without significant expansion of the CD8 compartment as a whole .
Current understanding of the immunological changes seen during primary EBV infection is almost exclusively derived from studies of AIM in adolescents and adults . These may not apply to the situation in asymptomatic primary EBV in early childhood , when a large proportion of infections occur . In this study , virological and immunological parameters of African children , studied during a time period when they undergo primary asymptomatic EBV infection , were examined to better understand the pathogenesis of primary asymptomatic EBV infection . Children who had EBV infection established for at least some months , as judged by the presence of VCA IgG antibodies but not IgM , have high virus loads , comparable to AIM patients , and their EBV-specific CD8+ T-cells show evidence of recent activation . These loads dropped significantly when tested six months later but were still elevated when compared to loads in other populations such as UK carriers , but were similar to those detected in healthy older children from The Gambia [32] . This suggests that in this situation , the virus set point is established several months after primary infection and that it is higher than in EBV-infected carriers in the UK . What determines the high virus set point detected in these children is unclear . One factor known to increase EBV loads is exposure to malaria , with children living in malaria endemic regions have higher EBV loads compared to those living in areas of sporadic transmission [32 , 33] . How EBV loads in children in the present study relate to loads in other African childhood cohorts is difficult to determine due to differences in assays and sample sources used to quantify EBV loads [33 , 34] . Malaria infection is thought to increase EBV loads either through promoting B cell proliferation [35] or altering T cell responses [36] . However the children in this study showed no evidence of acute malaria and data at the time of the study demonstrated a low prevalence of malaria in The Gambia [37 , 38] . Others have suggested virus load may be related to the age at infection , with African infants infected shortly after loss of maternal antibodies having higher virus loads than those infected later [34] . In this context , comparing virus loads to western populations should be interpreted with some caution as the timing of infection in these latter populations is relatively delayed and is particularly dependent on the ethnic group studied [39] . A clearer picture of the early immune response dynamics comes from studying children with IgM+ VCA-specific antibodies , who are likely to have had very recent EBV infection . Although these children had high virus loads in their PBMCs , akin to those seen in studies of symptomatic AIM [2 , 3 , 40] , they showed no obvious physical signs of infection . There was no significant expansion of lymphocytes or the CD3+ or CD8+ T-cell compartment; a finding consistent with those of others who have studied lymphocyte compartments in primary asymptomatic EBV infections [1 , 26 , 27] . Nevertheless , there was an ongoing virus-specific CD8+ T cell response in the IgM+ children with , in one case , greater than 15% of CD8+ T cells directed against a single dominant EBV epitope . This increased frequency of EBV-specific T cells found in the periphery in the presence of high virus loads may call in to question their role in control of the infection . However we have previously found that in AIM patients EBV-specific T cells lack expression of lymphoid homing receptors such as CCR7 and CD62L which allow access to tissues such as the tonsil . Virus replication and transformation of B lymphocytes , in AIM patients at least , occurs in this tissue and appears poorly controlled by the inefficient recruitment of EBV-specific T cells to this site [10] . As activated EBV-specific T cells express low levels of CCR7 and CD62L , we propose that the activated EBV-specific T cells in the children are similarly inefficiently recruited to the tonsil and virus replication and transformation poorly controlled at this site , allowing higher loads of virus to be detected in the presence of these strong responses . The phenotype of the antigen-specific CD8+ T-cells from the IgM+ VCA antibody children was consistent with what has been described in AIM donors , being highly activated ( HLADR+CD38+ ) , in cycle ( Ki-67+ ) and pro-apoptotic ( low Bcl-2 expression ) [8 , 10 , 14 , 17 , 41 , 42] . Why these children do not develop the expanded numbers of CD8+ T-cells observed in AIM patients after EBV infection is not clear . Possible reasons for the AIM associated hyper-expansion have included the development of heterologous immunity where an existing response to an epitope coded by a previously encountered pathogen cross reacts with another from EBV , amplifying the pool of T-cells responsive to EBV challenge and potentially inducing an exaggerated response . Children with less antigenic exposure would have a more limited repertoire of T-cells capable of responding in this heterologous manner [19] . Alternatively two recent studies comparing the incidence of AIM in monozygotic compared to dizygotic twins , and first- , second- or third-degree relatives have shown concordance in the development of AIM , suggesting that there may be a genetic component underlying disease development [20 , 22] . Perhaps more importantly , an emerging concept in the control of early EBV infection comes from studies using immunodeficient mice reconstituted with human haematopoietic cells , which repopulate human NK and T-cell repertoires . Depletion of NK cells in this model followed by challenge with EBV recapitulates an AIM-like response including splenomegaly , increased plasma levels of the pro-inflammatory cytokine IFN-γ and increased CD8+ T-cell numbers and frequencies [18] . Currently there is a lack of clarity in the literature as to the dynamics and role of NK cell responses in AIM and so the immediate relevance of these experimental findings to natural infection remain to be resolved . Some studies have indicated that there are inverse relationships between NK cell numbers and both severity of symptoms and virus load [43] , while others have shown a positive correlation between NK cell numbers both with virus load and severity of symptoms in AIM patients [3] . However NK cells are a heterogeneous population and so a key question arising from these studies is whether there is a difference in subsets of NK cells in terms of numbers or function between individuals who go on to develop AIM compared to children or others who don’t develop this disease . It is important to recognise that the immune system of children is different in comparison to adolescents where AIM is typically seen . Neonates are born with high levels of the immunosuppressive cytokine IL-10 , high levels of plasma immunosupressive factors such as adenosine , and have Th2 and Th17 skewed immunity , all of which decline by 1–2 years of age to near adult levels [44] . This “infant-adapted” immune profile evolves in the first few years of life , with a gradual increase in Th1 capacity , maturing of B cell and antibody responses , and development of T and B cell immunological memory [44] . By the time infants were recruited into the study at 14–18 months of age , they would have a considerably matured immune system compared to birth but some differences would persist compared to adolescents . Such immunological differences could have contributed to the asymptomatic primary infection that is seen in the children compared to adolescents . Throughout this study we have used the VCA serological status as a guide to when infection occurred , consequently the precise timing of the primary EBV infection cannot be determined . However , from previous data on infants at the same study site , only 18% of infants were found to be EBV infected at nine months of age [29] and one can , therefore , assume that the majority of children would have been infected six months to one year prior to recruitment . Furthermore , the IgM+IgG± donors were likely infected within the last 120 days and may be at a different stage of infection . Secondly , ensuring children are truly asymptomatic in this setting can be challenging as reliance on maternal perception may not be reliable and careful clinical studies of infants or children have described AIM like symptoms in some instances [45] . To combat this , the children all underwent a health screen by the study clinician , including baseline clinical observations such as weight , height , temperature and heart rate and where indicated a rapid malaria test . Recent work by Balfour et al . has suggested that 89% of primary EBV infections in a cohort of University students display some symptoms , although they may not fulfil all classic criteria for AIM , which is different to our observations in Gambian children [3] . In summary , this study supports the notion that AIM is an immunopathological disease and that symptoms are caused by the significantly expanded CD8+ T-cell responses to the virus . It provides clear evidence that during primary asymptomatic infection EBV-specific responses are indeed activated and can occupy a significant percentage of the circulating CD8+ T cell pool . However these responses appear able to contain the infection without the massive expansion that characterises AIM . Conversely then , the symptoms of AIM appear to derive from the absolute CD8 expansion rather than from the virus infection per se .
The Gambian Government/ MRC Laboratories Joint Ethics Committee approved this study . Participants were enrolled after individual written informed consent was obtained from the participant's parent/guardian . SCC 1206 . This study was conducted in a peri-urban Medical Research Council ( MRC ) UK clinic , Sukuta , situated within the Government Sukuta Health Centre , serving a low-income population living in crowded conditions . A cohort of 120 children aged between 14 and 18 months were screened when they attended the local government health centre for their routine booster vaccination of diphtheria , tetanus , whole cell pertussis ( DTwP ) combined vaccine . All children were screened by enquiring about a maternal history of recent illness ( e . g . fever ) and a clinician examination for signs and symptoms of infectious mononucleosis , including weight and baseline observations ( temperature , heart rate , weight and length ) . Any child found to be unwell ( observations outside normal clinical range for age or maternal report of recent illness ) or had a weight below that specified on the local Infant Welfare Card Growth Chart were not recruited into the study . ( n = 6 ) . Of the children not recruited , none of these showed clinical features suggestive of infectious mononucleosis . Five millilitres of blood was collected from each child into vacutainers containing EDTA ( BD ) . A 500μl aliquot was removed and used to obtain a full blood count on each child using a M-series M16/M20 Haematology Analyser ( Medonic , Sweden ) . A further 250μl aliquot was removed and whole blood flow cytometric staining performed . The remaining blood was layered on to 4mls of Lymphoprep ( Axis-Shield , UK ) in 15ml Leucosep tubes ( Greiner Bio-One , UK ) . Following centrifugation , the plasma layer was removed , and cryopreserved in 2ml aliquots and stored at −70°C for downstream serology . The lymphocyte interphase was harvested and washed . Cells were counted and re-suspended in freezing medium ( FCS ( Sigma-Aldrich ) supplemented with 10% ( v/v ) dimethyl sulfoxide ( DMSO ) ) at approximately 5 x 106/ml . Children were brought back one week later to receive the Pentavalent vaccination ( DTwP , Hep B , Hib ) ( Easy Five Panacea Biotec ) . They were subsequently invited to return a week after vaccination and again at six months to undergo further blood sampling . AIM patients were recruited from a cohort of young adults ( 18–25 years old ) collected at the University of Birmingham , UK . All patients gave written informed consent to donate samples and experiments were approved by the South Birmingham Local Research Ethics Committee ( reference number 07/Q2702/24 ) . Patients were defined as AIM by having tonsillitis/sore throat , high lymphocyte counts and being heterophile antibody positive . Mononuclear cells were harvested from blood specimens and stored as described above . EBV genome loads were assayed by quantitative real-time PCR , as described elsewhere [46] . DNA extraction was performed from 1x106 PBMCs using QIAmp DNA Blood Mini kit ( Qiagen ) . IgG and IgM reactivity to EBV Viral Capsid Antigen ( VCA ) were measured using a previously described in-house immunofluorescence assay at the Institute for Cancer Studies , Birmingham [47 , 48] and the MRC-University of Glasgow Centre for Virus Research , University of Glasgow [49 , 50] respectively . For children a 100 μl volume of whole blood for each donor was stained with antibodies to the following surface markers: CD3 PE , CD4 PerCP , ( BD Biosciences ) , CD8 efluor450 and CD27 APCalexafluor750 ( Ebioscience ) for 30 min at 4°C . Red blood cells were then lysed using 1:10 FACS Lysing Solution ( BD Biosciences ) and incubated for 10 min at room temperature . Cells were then washed twice in FACS buffer ( PBS , 5% BSA , 5% EDTA ) and re-suspended in Cytofix ( BD , Biosciences ) . Samples were acquired on a Cyan ADP flow cytometer using Summit software ( Beckman Coulter ) at MRC Gambia . Lymphocyte subsets from AIM patients were identified by staining with antibodies specific to: CD19 FITC , CD4 PE ( Biolegend ) , CD27 APC elfluor 780 , CD3 efluor 450 ( eBioscience ) and CD8 qDot 655 ( Invitrogen ) . Samples were stained for 30 min on ice , washed and analysed immediately on an LSR-II flow cytometer ( BD Biosciences ) . Data was analysed using Flow-Jo software ( Treestar Inc ) . Tetramers were used to identify and analyse the surface marker phenotype of epitope-specific CD8+ T-cells . From the aforementioned IFN-γ ELISPOT data we selected the following epitopes , B*0801 RAKFKQLL , B*3501 EPLPQGQLTAY and A*0201 GLCTLVAML for tetramer manufacture , as they were frequent targets of the immune response . Markers of activation ( CD38 & HLDR ) , proliferation ( Ki-67 ) and the anti-apoptotic marker , Bcl-2 were assessed . Tetramers were validated for specificity against HLA-matched and mismatched seropositive and seronegative donors . Tetramer staining was performed on cryopreserved PBMCs as described elsewhere . Cells were thawed and stained with LIVE/DEAD fixable Aqua Dead Cell Stain for 30 min at 4°C , washed and stained with 1μg of tetramer-PE for 15min at 37°C . Following two further washes , surface staining with CD3-Qdot655 , CD4-Qdot605 , CD8-Qdot705 , CD14-V500 , CD19-V500 , CD38-APC and HLA-DR-Alexafluor700 were performed . Following fixing and permeabilisation as described above , intracellular staining with Ki67-Alexafluor488 and Bcl-2 ( B-cell lymphoma 2 ) -V450 was performed . Fluorescence minus one samples were included to aid gating during subsequent flow cytometric analysis . A comparison of expression of the above phenotypic markers on EBV-specific and the total CD8+ T-cell populations were performed . Compensation for fluorescence ‘spill-over’ was performed using the BD CompBead Anti-Mouse Ig set ( BD Biosciences ) and the antibodies described above . Briefly , antibodies were added to separate tubes containing one drop each of Anti-Mouse Ig beads and the negative control beads ( which do not bind κ light chain-bearing immunoglobulin ) . Following a 30 min incubation at 4°C , beads were washed and re-suspended in FACS buffer . All statistical analyses were performed using Graphpad Prism version 5 . 0 for Macintosh ( GraphPad Software , San Diego California , USA , www . graphpad . com ) . Comparisons between variables were performed using the Mann-Whitney U test ( two-tailed ) and for non-parametrically distributed data , the Wilcoxon matched pairs test ( for comparisons between total CD8+ and virus-specific CD8+ T-cells made within individuals ) was used . Correlations between non-normally distributed data were made using the Spearman’s rank correlation coefficient .
|
Primary infection with EBV , a common human herpesvirus , is typically asymptomatic in childhood but , if occurring in adolescence or later , often presents as AIM . This febrile illness is characterised by high virus loads in the blood and an exaggerated EBV-specific CD8+ T-cell response that pushes total CD8+ T-cell numbers well above normal levels . By contrast , very little is known about the events of asymptomatic primary infection . We therefore studied young Gambian children at an age at which many acquire EBV , monitoring them over six months for evidence of EBV infection by virus load in the blood , virus-specific IgM and IgG antibody status , and virus-specific CD8+ T-cell responses . Focusing on IgM-positive children with very recent EBV infection but no history of symptoms , we found that they carried a virus load equivalent to that seen in AIM patients and also mounted a classical virus-specific CD8+ T-cell response . However , that response , though it could occupy at least 15% of the circulating CD8+ T-cell pool , occurred without the huge global expansion of CD8 numbers seen in AIM . This work reinforces the idea that the host’s exaggerated CD8+ T-cell response , rather than the virus load per se , leads to the symptoms of AIM .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Early Virological and Immunological Events in Asymptomatic Epstein-Barr Virus Infection in African Children
|
The group model is a useful tool to understand broad-scale patterns of interaction in a network , but it has previously been limited in use to food webs , which contain only predator-prey interactions . Natural populations interact with each other in a variety of ways and , although most published ecological networks only include information about a single interaction type ( e . g . , feeding , pollination ) , ecologists are beginning to consider networks which combine multiple interaction types . Here we extend the group model to signed directed networks such as ecological interaction webs . As a specific application of this method , we examine the effects of including or excluding specific interaction types on our understanding of species roles in ecological networks . We consider all three currently available interaction webs , two of which are extended plant-mutualist networks with herbivores and parasitoids added , and one of which is an extended intertidal food web with interactions of all possible sign structures ( +/+ , -/0 , etc . ) . Species in the extended food web grouped similarly with all interactions , only trophic links , and only nontrophic links . However , removing mutualism or herbivory had a much larger effect in the extended plant-pollinator webs . Species removal even affected groups that were not directly connected to those that were removed , as we found by excluding a small number of parasitoids . These results suggest that including additional species in the network provides far more information than additional interactions for this aspect of network structure . Our methods provide a useful framework for simplifying networks to their essential structure , allowing us to identify generalities in network structure and better understand the roles species play in their communities .
Networks are a useful tool to understand patterns of interactions in an ecological community . As ecologists have collected more and more network data , the size of published networks has grown dramatically , with many networks now containing hundreds of species . To make sense of these increasingly complex data , we need tools to simplify the network down to its essential structure , allowing us to identify general patterns of interaction in the community . The group model ( equivalent to the stochastic block model from the social science literature , [1] ) is a useful way to simplify and understand ecological networks . It has previously been common to characterize species in terms of their ecological niches , that is , by the resources or predators of a given species . Species with identical niches were considered “trophic species” , and ecological networks were often simplified by combining them [2] . However , this approach is highly sensitive to small changes or errors in the network structure , since a single missing or false interaction can change which species may be combined . The group model [3] models the concept of ecological equivalence [4] ( distinct from the term as used in neutral theory ) . Species are considered to be ecologically equivalent if their predators and prey are equivalent , who are equivalent if their predators and prey are equivalent , and so on . In other words , species are grouped together if they eat and are eaten by the same other groups . This recursive definition implies that species which are far from each other in the network may still impact each other’s grouping . This reflects the ecological reality of the complex ways in which species in a network influence each others’ dynamics , for example , via trophic cascades or apparent competition [5 , 6] . Since ecologically equivalent species prey on and are preyed on by the same other groups , species within a group can be thought of as filling the same role in the community , and may be expected to operate in the community in similar ways . The group structure is also able to capture both modular ( compartmental ) and anti-modular ( i . e . , trophic levels ) aspects of the network , both of which are found in ecological networks . Thus , the group model is a useful way to gain a coarse-grained view of ecological dynamics and the niches that are filled in the community . A limitation of the group model is the fact that it can only group species based on a single interaction type ( usually predator-prey interactions , although it could in principle be applied to any one interaction type ) . Of course , species in ecological communities interact in diverse ways , and different interaction types operate simultaneously to influence community dynamics [7 , 8] . Although ecologists have traditionally built separate networks for each interaction type , such as food webs ( containing only feeding interactions ) , or plant-pollinator and plant-seed-disperser networks ( containing only mutualistic interactions ) [9–14] , there is a growing recognition that different interaction types may work in concert to influence communities . Both empirical and simulation studies have demonstrated the complex ways in which mutualisms and antagonisms may interact [7 , 15 , 16] . For example , recent work has begun to explore the possible effects of including cheaters in mutualistic networks [15 , 17 , 18] , modelling communities with multiple interaction types [19 , 20] , and combining mutualistic networks and food webs [21–23] . Here , we extend the group model from unsigned ( single interaction type ) to signed directed adjacency matrices , allowing ecologists to study the general structure of merged interaction networks . Using this extension of the group model , species in a group tend to interact with other groups in the same way . We demonstrate one possible use of this method by considering how including or excluding different interaction types changes our understanding of group structure in three interaction webs ( the only three such networks currently available ) . Despite the growing body of work on potential impacts of merging networks with multiple interaction types , it is unknown whether these merged networks provide new , meaningful information about species roles at the network level . While it is intuitive that more types of interaction data would provide more ( or more accurate ) information about the roles species play in their communities , it is valuable to study this question directly . Clearly , species groupings will be contingent on the species and interactions that are included in the network . Adding interactions may reinforce , refine , or contradict the previous understanding of species roles ( Fig 1 ) . We study how our understanding of species roles changes based on different types using three networks of two types . The Tatoosh mussel bed network is an intertidal food web with additional interaction types included . This network contains feeding ( +/- ) , competitive ( -/- ) , mutualist ( +/+ ) , commensal ( +/0 ) , and amensal ( -/0 ) interactions . For this network , we compare how species group based on all interaction types , only trophic interactions , and only nontrophic interactions . The other interaction networks , from Doñana Biological Reserve [24] and Norwood farm [25] , are terrestrial networks which include plants , plant mutualists , plant herbivores , and in the Norwood web , parasitoids which parasitize herbivores . These networks are structurally different from the Tatoosh web in that they are almost entirely multipartite; that is , they are composed of “layers” of species which only interact with the layers above and below ( i . e . , mutualists interact with plants , plants interact with mutualists and herbivores , and herbivores interact with plants and parasitoids ) . In these networks , only plants are involved in both feeding and mutualistic interactions , so we consider how the grouping of plants is affected by including mutualists , herbivores , or both . For the Norwood web , we also consider the effect of including or excluding parasitoids on plant groupings . Using this framework , we study if and how omitting specific interaction types changes our understanding of network structure and species roles .
The Makah Tribal Council has granted permission to the Wootton lab for access to Tatoosh Island . A food web composed of S species may be represented by an adjacency matrix A , where Aij is 1 if j consumes i , and 0 otherwise . Similarly , interaction networks may be represented by a signed adjacency matrix where Aij is 1 if the growth rate of j positively depends on the presence of i , -1 if its growth rate negatively depends on i , and 0 otherwise . Such a matrix may be thought of as containing the signs of the community matrix ( the Jacobian evaluated at equilibrium ) , as opposed to a matrix of zero-sum energy or nutrient flow throughout the system ( sensu [26] ) . Some interactions , such as competition for carbon or another nutrient , may be considered an indirect interaction which is the product of two consumer-resource interactions ( two direct consumer-carbon interactions in this case ) . In this example , carbon would be incorporated into the differential equations underlying the community matrix . Since we are interested in how species group within an interconnected network , we require that the complete interaction networks are a single weakly connected component ( that is , isolated subgraphs were removed ) . Interaction data for Tatoosh Island were collected from the intertidal middle zone based on observed interactions and natural history information . This middle zone on Tatoosh is dominated by the mussel Mytilus californianus . This mussel-dominated band is defined from below by the presence of Pisaster ochraceus , which consumes M . californianus [27] , and from above by physiological constraints , such as time spent submerged [28] . The signed interaction network contains 110 taxa and 1898 interaction pairs ( 869 +/- , 5 +/+ , 208 +/0 , 492 -/0 , and 324 -/- ) . This dataset is available on Dryad ( DOI:10 . 5061/dryad . 39jv1 ) The largest weakly connected component was taken from Doñana Biological Reserve and Norwood Farm ( data made available in [24] and [25] , respectively ) . The Doñana network contains 391 species total , with 170 plants , 207 mutualists ( 576 mutualistic interactions ) , and 14 herbivores ( 221 feeding interactions ) . The Norwood network contains 445 species , with 91 plants , 251 mutualists ( 569 mutualistic interactions ) , 62 herbivores ( 570 herbivory interactions ) , and 43 parasitoids ( 367 parasitic interactions ) . Two species were classified in two categories: one which interacted both as a mutualist and as an herbivore , and another as both a mutualist and a parasitoid . Because taxonomically similar species are generally expected to fill similar roles in a community [29] ( but see [30] ) , taxonomic data provide a potential natural grouping . Tatoosh taxa were classified to kingdom and phylum , and plants in the Doñana and Norwood webs were classified to the order level . Taxonomic levels were chosen to have a number of groups that was similar to the number of groups found by the group model for the complete networks . The high phylogenetic diversity of the Tatoosh system meant that taxonomic groupings beyond the phylum level included too many groups to provide useful information about the system . Taxonomic data for all three networks were gathered from the Integrated Taxonomic Information System ( ITIS ) database and Encyclopedia of Life ( see S1 Text for details ) . Consider an interaction web with S species and L links , K of which are positive and L − K negative . The data can be represented using a signed directed adjacency matrix N . What is the probability of obtaining N by chance ? A simple model of random signed network structure is similar to an Erdős-Rényi random graph with S species and a fixed probability c of connecting any two nodes , with an additional probability π that a link is positive . Then the probability of obtaining exactly N using this model is: P ( N ( S , L , K ) | c , π ) = c L π K ( 1 - c ) Z ( 1 - π ) L - K ( 1 ) where Z = S2 − L is the number of zeros in the matrix . This likelihood is maximized when c ^ = L L + Z and π ^ = K L . Now to see this in the context of the group model , consider N when divided into two groups , X and Y . If N is a mutualistic web , these groups might correspond to plants and pollinators . Now the random network process involves eight probabilities: cxx , the probability of a species in group X connecting to another species in group X , πxx , the probability of a link between two species in X being positive , cxy , the probability of a species in X connecting to a species in Y , and so on for cyx , cyy , πyy , πxy , and πyx , which are defined similarly . Note that cxy and cyx are not necessarily equal ( nor are πxy and πyx ) , since N need not be symmetric . Then the probability of obtaining N given the two groups is: P ( N ( S , L , K ) | c i j , π i j , i , j ∈ x , y ) = ∏ i ∈ ( X , Y ) ∏ j ∈ ( X , Y ) c i j L i j π i j K i j ( 1 - c i j ) Z i j ( 1 - π i j ) L i j - K i j ( 2 ) Analagous to Eq 1 , this likelihood is maximized when c ^ i j = L i j L i j + Z i j and π ^ i j = K i j L i j for all combinations of groups . This can be generalized to g groups as follows: P ( N ( S , L , K ) | c i j , π i j , i , j ∈ 1 : g ) = ∏ i = 1 g ∏ j = 1 g c i j L i j π i j K i j ( 1 - c i j ) Z i j ( 1 - π i j ) L i j - K i j ( 3 ) When g = 1 , this is equivalent to Eq 1 . When g = S , each species is in its own group , and the likelihood is 1 . Such a grouping is not very informative , so we need to perform model selection . Using a uniform prior ( such that the probability of each model is 1 2 ) , it is possible to analytically calculate a Bayes factor to compare two groupings . For groupings G1 and G2 , the Bayes factor is given by: B = P ( N | G 1 ) P ( N | G 2 ) ( 4 ) where P ( N∣Gi ) is the marginal likelihood ∫ 0 1 ⋯ ∫ 0 1 P ( c i j , π i j , i , j ∈ 1 : S | G i ) P ( N | c i j , π i j , i , j ∈ 1 : g , G i ) d c 11 … d c g g d π 11 … d π g g ( 5 ) which can be analytically integrated to give: ∏ i = 1 g ∏ j = 1 g K i j ! Z i j ! ( L i j - K i j ) ! ( 1 + L i j ) ( 1 + L i j + Z i j ) ! ( 6 ) Because there are many possible groupings to choose from , we compared the marginal likelihoods of the groupings when searching for the best grouping , rather than explicitly calculating B for each pair . We searched for the optimal grouping using Metropolis-Coupled Markov Chain Monte Carlo ( MC3 ) with a Gibbs sampler ( see S1 Text ) . It is not feasible to exhaustively search the space of all possible groupings , so the best groupings found are not guaranteed to be the optimal ones , but for simplicity , we refer to them as “best groupings” throughout . The entropy of a partition A is an information theoretic measure of the information content or uncertainty of that partition , measured in nats [31] . A partition where all species are in the same group would have low entropy , because we can be quite certain of which group any given species belongs to . In contrast , a partition with many groups would have higher entropy , since it is difficult to make an a priori guess about the group identity of a given species . Entropy is calculated as: H ( A ) = - ∑ a ∈ A p ( a ) ln ( p ( a ) ) ( 7 ) This metric is known as Shannon entropy , commonly used in ecology to measure the diversity of a community [32] . The joint entropy of two partitions A and B is similarly defined: H ( A , B ) = - ∑ a ∈ A ∑ b ∈ B p ( a , b ) ln ( p ( a , b ) ) ( 8 ) This can be thought of as the union between H ( A ) and H ( B ) , since it sums over all joint probabilities of the two entropies . Note that for all entropies , 0 ln ( 0 ) is given to be 0 , so that including values with probability zero does not change the entropy [31] . To measure the similarity between two partitions , we then wish to know how much entropy the partitions share . This is known as the mutual information ( MI ) , which quantifies the reduction in entropy of partition B when partition A is known . It is calculated as M I A B = H ( A ) + H ( B ) - H ( A , B ) ( 9 ) This can be thought of as the intersection between H ( A ) and H ( B ) . Converting this measure into probabilities gives us M I A B = - ∑ a ∈ A p ( a ) ln ( p ( a ) ) - ∑ b ∈ B p ( b ) ln ( p ( b ) ) + ∑ a ∈ A ∑ b ∈ B p ( a , b ) ln ( p ( a , b ) ) ( 10 ) = - ∑ a ∈ A ∑ b ∈ B p ( a , b ) ln ( p ( a ) ) - ∑ a ∈ A ∑ b ∈ B p ( a , b ) ln ( p ( b ) ) + ∑ a ∈ A ∑ b ∈ B p ( a , b ) ln ( p ( a , b ) ) ( 11 ) = ∑ a ∈ A ∑ b ∈ B p ( a , b ) ln ( p ( a , b ) p ( a ) p ( b ) ) ( 12 ) To see how this is calculated for a partition generated by the group model , see Box 1 . Significance of MI values was estimated based on a randomization test . To estimate how likely it was to get an equal or higher MI by chance , each of the two partitions were shuffled , such that the randomized partitions conserved the number of species in each group ( and therefore the upper bound on the MI , see S1 text for details ) , but not their identities . The MI was then calculated for the randomized partitions . This process was repeated one million times , and the p-value was estimated as the probability of getting an MI greater than or equal to the observed MI for the two partitions . Since the probability of getting a given MI is based both on the entropies and the groupings , it is possible to get a low p-value for a relatively low MI , or a high p-value for a high MI . Code for calculating partition similarity , obtaining taxonomic data , and running the search algorithm are available on GitHub at https://github . com/esander91/SignedGroupModel .
Both the partitions for the network with all interactions and the network with trophic interactions grouped species in a similar way ( Fig 3 ) . Though the complete web grouping divided taxa into more groups than the trophic grouping did ( 19 and 13 groups , respectively ) , many of these additional groups were simply nested within groups from the trophic one ( Fig 4A ) . Many groups corresponded strikingly well to known ecologically relevant groups in this community , including predatory snails ( n = 4 ) , kelps ( n = 5 ) , limpets ( n = 4 ) , and foraging birds ( n = 3; Fig 5 ) . The complete grouping was also quite similar to the nontrophic grouping . In contrast to the trophic partition , which captured the general structure of the complete grouping across the entire web ( S1 Fig ) , the nontrophic partition captured portions of the complete one very precisely , but grouped many species into one of two broad groups . Although the nontrophic network contained more interactions than the trophic one overall , these interactions were unevenly spread across species; in particular , sessile species tended to competitively interact with other species , while mobile species often only interacted with a few other species in a nontrophic fashion . As a result , many sessile species ( particularly algae and barnacles; see S2 Fig and S1 Table ) were organized into similar groups as in the complete grouping , while most other species were placed into one of two large groups which were sparsely connected to the rest of the network . The trophic and nontrophic groupings were less similar to each other than to the complete grouping ( Fig 4C ) , but were much more similar to each other than expected by chance . Jackknife resampling of the complete network showed that group structure was robust to removal of individual species , as measured by ratio between the MI for the Jackknifed and original groupings and the maximum MI possible given their entropies ( mean M I M I M A X = . 99 , σ = . 014; see S1 Text for methodological details ) . Plants in the complete Doñana network grouped in a similar way to both the herbivore-removal and mutualist-removal networks . The herbivore-removal and mutualist-removal partitions were much less similar to each other than to the complete partition , although still more similar than expected by chance ( Figs 6 and 7 ) . The herbivore-removal grouping contained much more information about the complete grouping than the mutualist-removal one did , possibly because mutualists greatly outnumbered herbivores in this network , both in number of species ( 207 and 14 species , respectively ) and interactions with plants ( 576 and 221 interactions ) . When parasitoids were excluded from the network , results for the Norwood community were qualitatively similar to Doñana . Mutualist-removal and herbivore-removal groupings were similar to the grouping with both mutualists and herbivores ( but not parasitoids ) , but were less similar to each other ( Figs 8 and 9 ) . Interestingly , removing herbivores in this network changed group structure more than removing mutualists , even though there were many more mutualists than herbivores ( 251 and 62 species ) , and mutualists and herbivores had almost exactly the same number of interactions with plants ( 569 and 570 interactions ) . Including parasitoids in the network markedly changed the resulting group structure . The complete grouping remained similar to the herbivore-removal grouping ( which also removes parasitoids , since they only interact with herbivores ) . However , the mutualist-removal partition was no more similar to the complete one than expected by chance . Surprisingly , the partition for the mutualist-parasitoid-removal was more similar to the complete partition than either the herbivore or mutualist removal groupings . Taxonomic grouping provided some information about complete groupings for all three networks . The Tatoosh complete grouping is almost perfectly nested within the species classification by kingdom ( Figs 3 and 10 ) . However , because this classification is so broad , it provides less information than phylum , even though the phylum grouping and complete grouping are dissimilar in many areas . In the Doñana and Norwood webs , taxonomic order was significantly similar to the complete groupings ( Figs 6 and 8 , respectively ) , but this similarity was not even across orders: some orders strongly grouped together in the complete groupings , while many others were scattered between several groups ( Fig 10 ) .
The extended group model is a general method for identifying functionally equivalent nodes in signed directed networks . We have discussed the method as applied to ecological interaction webs , but the methodology could also be used to study the structure of networks of gene regulation [43 , 44] , sensors [45] , and even social networks which incorporate both positive and negative social interactions [46] . The generality of the method does not detract from its usefulness in ecology; in fact , the model is able to identify highly specific ecological roles . This model is a new and useful exploratory tool to understand and compare the coarse-grained structure of ecological communities .
|
Ecological interactions are highly diverse even when considering a single species: the species might feed on a first , disperse the seeds of a second , and pollinate a third . Here we extend the group model , a method for identifying broad patterns of interaction across a food web , to networks which contain multiple types of interactions . Using this new method , we ask whether the traditional approach of building a network for each type of interaction ( food webs for consumption , pollination webs , seed-dispersal webs , host-parasite webs ) can be improved by merging all interaction types in a single network . In particular , we test whether combining different interaction types leads to a better definition of the roles species play in ecological communities . We find that , although having more information necessarily leads to better results , the improvement is only incremental if the linked species remain unchanged . However , including a new interaction type that attaches new species to the network substantially improves performance . This method provides insight into possible implications of merging different types of interactions and allows for the study of coarse-grained structure in any signed network , including ecological interaction webs , gene regulation networks , and social networks .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
What Can Interaction Webs Tell Us About Species Roles?
|
The small genome of polyomaviruses encodes a limited number of proteins that are highly dependent on interactions with host cell proteins for efficient viral replication . The SV40 large T antigen ( LT ) contains several discrete functional domains including the LXCXE or RB-binding motif , the DNA binding and helicase domains that contribute to the viral life cycle . In addition , the LT C-terminal region contains the host range and adenovirus helper functions required for lytic infection in certain restrictive cell types . To understand how LT affects the host cell to facilitate viral replication , we expressed full-length or functional domains of LT in cells , identified interacting host proteins and carried out expression profiling . LT perturbed the expression of p53 target genes and subsets of cell-cycle dependent genes regulated by the DREAM and the B-Myb-MuvB complexes . Affinity purification of LT followed by mass spectrometry revealed a specific interaction between the LT C-terminal region and FAM111A , a previously uncharacterized protein . Depletion of FAM111A recapitulated the effects of heterologous expression of the LT C-terminal region , including increased viral gene expression and lytic infection of SV40 host range mutants and adenovirus replication in restrictive cells . FAM111A functions as a host range restriction factor that is specifically targeted by SV40 LT .
SV40 large T antigen ( LT ) is a multifunctional viral protein that plays a central role in orchestrating productive viral infection as well as cellular transformation . Discrete regions of LT are required for binding to specific host proteins and provide specific functions . The LXCXE motif ( residues 103–107 ) binds to the retinoblastoma family of proteins RB ( RB1 ) , p107 ( RBL1 ) and p130 ( RBL2 ) to promote cell cycle entry . The N-terminal J domain ( residues 1–82 ) binds specifically to heat shock protein chaperone HSC70 ( HSPA4 ) and contributes to efficient viral replication as well as inactivation of p107 and p130 growth suppressing activities [1] , [2] . The LT DNA binding domain ( DBD; residues 131–251 ) binds specifically to the SV40 DNA origin of replication . The central domain ( residues 260 to 627 ) contributes to LT hexamer formation , contains intrinsic ATPase and helicase activity , and binds p53 [3]–[5] . The C-terminal region ( residues 627–708 ) contains no known structural domains but does undergo specific post-translational modifications , including acetylation of lysine residue 697 ( K697 ) and phosphorylation of threonine 701 ( T701 ) , the latter required for LT binding to FBXW7 [6] , [7] . In addition , an intact LT C-terminal region is required for the host range and adenovirus helper functions of SV40 [8] , [9] . Viral host range is defined as the set of cells , tissues and species that a virus can productively infect . There are a wide variety of cellular host range restriction factors as well as counter strategies employed by viruses to overcome them . Sometimes virally encoded proteins bind directly to specific host proteins to overcome host range restriction . SV40 host range mutant viruses , all of which contain deletions or truncations in the C-terminal region of LT , express lower levels of mRNA and protein for early ( LT ) and late ( VP1 ) genes compared to wild type virus and fail to support lytic infection in restrictive cell types [10] , [11] . Heterologous expression of the C-terminal region of LT in trans leads to increased early and late gene expression of host range mutant virus and rescues the ability of these mutant viruses to induce lytic infection in restrictive cells [10] , [12] . In addition , the C-terminal region of LT is required for the adenovirus helper effect; human adenoviruses are unable to replicate in certain monkey cell lines unless SV40 is also present [13] . The LT C-terminal region contributes a discrete activity that supports replication of SV40 and adenovirus in restrictive cell lines although it is uncertain whether these activities reflect the same function . Here , we examined host interactome and transcriptome perturbations induced by full-length and discrete functional domains of LT . The resulting data provides a global view of LT-host cell interactions and highlights cellular pathways perturbed by the presence of LT . Notably , we identified FAM111A , a previously uncharacterized cellular protein that binds specifically to the C-terminal region of LT . We provide evidence that this interaction contributes to SV40 host range and adenovirus helper functions .
The C-terminal region of LT is required for efficient viral gene expression and replication in the African green monkey kidney ( AGMK ) CV-1P cell line [8] , [9] . The SV40 host range mutant virus HR684 lacks the C-terminal 24 residues of LT and has significantly reduced expression of early ( LT ) and late viral ( VP1 ) genes compared to wild type virus in CV-1P cells ( Figure 1A; [10] ) . Heterologous expression of the LT C-terminal 82 residues ( C-TERM; residues 627 to 708 ) markedly increased levels of HR684 LT and VP1 in these cells [10] . Since LT C-TERM could support increased viral gene expression in trans , we suspected that this LT fragment could bind to a specific host cell factor and thereby increase viral gene expression . Proteomic analysis was not possible in AGMK cells because whole genome and proteome sequences were not available . Instead , we tested several human cell lines for the ability of the LT C-terminal region to increase levels of host range mutant viral genes . Increased levels of HR684 LT and VP1 were observed in U-2 OS but not in HeLa or T98G cells when co-expressed with LT C-TERM ( Figure 1A and Figure S1 in Text S1 ) . Given the ability of C-TERM to increase HR684 gene expression in U-2 OS and CV-1P cells , we selected the U-2 OS cell line to further analyze of the host range phenotype . To examine the effect of the LT C-terminal region on LT and VP1 levels , HR684 viral DNA was transfected into U-2 OS cells that stably expressed the C-TERM construct or empty vector ( Figure 1B ) . LT could be detected at 24 hours and VP1 at 48 hours after transfection in both cell lines . While levels of LT and VP1 decreased at 72 and 96 hours respectively in the vector control cells , both LT and VP1 show persistent expression at 72 and 96 hours after transfection in the LT C-TERM containing cell line ( Figure 1B ) . This result indicates that the LT C-terminus functions at least in part to sustain viral gene expression in U-2 OS cells . Discrete functional domains within the SV40 LT protein bind to diverse host cell proteins ( Figure 2A ) . We generated LT expression constructs encoding epitope-tagged fusions of full-length LT as well as fragments corresponding to computationally- and functionally-defined domains . Full-length LT ( T1 ) , the LT N-terminal region encoded by residues 1 to 135 ( T6 fragment ) or residues 1 to 350 ( T8 fragment ) , and the LT C-terminal region between residues 260 and 708 ( T16 fragment ) were stably expressed in U-2 OS cells ( Figure S2 in Text S1 ) . We determined the effects of full-length LT and various fragments on global gene expression . Cells expressing T1 and T16 constructs showed significant differential expression changes of multiple target genes compared to control . In contrast , cells expressing the N-terminal T6 and T8 constructs showed minimal changes in gene expression compared to control . To identify patterns of host transcriptional perturbation common across all comparisons between the set of LT constructs and controls , we applied model-based clustering to construct clusters from the 430 most frequently perturbed host genes ( Table S1 in Text S1 ) . Of the 14 identified clusters , 9 exhibited significantly enriched GO terms ( Figure 2B ) . Heterologous expression of T1 or T16 led to increased expression of genes involved in the cell cycle ( cluster C2 ) , regulated by the DREAM and MYB-MuvB complexes [14] , [15] , and decreased expression of genes in cluster 3 ( Figure 2B and Table S2 in Text S1 ) , enriched for p53 target genes . We compared the transcriptional perturbations induced by T1 in U-2 OS cells with an earlier study performed of T1 in normal human diploid IMR90 fibroblasts [16] . We found that the p53- and the DREAM-regulated pathways were similarly perturbed in both cell types ( Figure 2B ) [16] . To assess the biological significance of the expression profiles we applied gene set enrichment analysis ( GSEA ) [17] . A significant enrichment for increased expression of DREAM and MYB-MuvB gene sets was observed in T1-expressing cells ( Figure 2C ) . In contrast , there was significant enrichment for decreased expression of p53 genes and MYB-MuvB genes in the T16-expressing cells . These results suggest that full-length LT ( T1 ) and the C-terminal T16 portion exert distinct transcriptional perturbations in U-2 OS cells . Given the effects of LT on cellular and viral gene expression , we sought to identify host proteins that bind to LT . We used multidimensional protein identification technology ( MudPIT ) to analyze preparative scale immunoprecipitations by mass spectrometry [18] . We pooled the results of five independent experiments from cells expressing full-length T1 . In total , we identified 89 proteins that co-purified with T1 and were detected in at least three out of five independent affinity purification experiments ( Figure 3A and Table S3 in Text S1 ) . We detected several previously reported interactors of LT including FBXW7 and p53 [19] . In addition , we identified co-complexes with RB family proteins and E2F/DP1 transcription factors , but not any of the five proteins contained within the MuvB core complex ( LIN9 , LIN37 , LIN52 , LIN54 and RBBP4 ) that bind to p130 ( RBL2 ) in the DREAM complex [14] . Immunoprecipitation for p130 co-precipitated the MuvB subunit LIN37 in the absence of LT but not when LT was present ( Figure 3B ) , further supporting the conclusion that LT disrupts the DREAM complex . Consistent with previous reports that the N-terminal J domain binds HSC70 [1] , [2] , [20] , [21] , we observed specific association with HSPA4 . We also identified HSPH1 ( Hsp105 ) , STUB1 ( CHIP ) and members of the BAG protein family , including BAG2 , BAG3 and BAG5 . BAG proteins bind to HSC70 to inhibit ubiquitination of misfolded proteins by STUB1 [22] , [23] . We also detected LT association with members of the PAF1 transcription elongation complex PAF1C , including CDC73 , PAF1 , CTR9 and WDR61 [24]–[26] . We detected a previously unreported association of LT with the uncharacterized protein FAM111A ( family with sequence similarity 111 , member A; LOC63901; Gene ID: 63901 ) . FAM111A was reproducibly detected in all five full-length LT ( T1 ) replicates but not in the corresponding negative controls ( Figure 3C and Table S3 in Text S1 ) . The T16 and C-TERM LT fragments also showed association with FAM111A ( Figure 3C ) , indicating that the C-terminal region of LT was sufficient for association with FAM111A . We also identified LT-associated proteins using iTRAQ stable isotope labeling ( see Supplementary experimental procedures and Table S4 in Text S1 ) . Full-length LT ( T1 ) showed associations with RB , p107 , p53 , and FAM111A ( Figure 3D ) . In addition , the T1 and T8 constructs bound p107 while T1 and T16 fragments bound to p53 and FAM111A . The extensive sequence coverage of FAM111A ( 62 . 4%; Figure S3 in Text S1 ) and high normalized spectral abundance factor ( dNSAF ) values observed in the MudPIT analyses of the T1 affinity purification ( Figure 3C ) , as well as the strong enrichment relative to the negative control by iTRAQ analysis , suggest that LT binds FAM111A efficiently . We tested FAM111A binding to LT in a yeast two-hybrid ( Y2H ) assay . The LT constructs T1 , T16 and C-TERM bound to FAM111A either as bait or prey in Y2H , while T8 could not ( Figure 4A ) consistent with the mass spectrometry analyses . To determine where LT bound to FAM111A , we generated fifty N- and C-terminal deletion constructs of FAM111A and tested them as bait or prey by Y2H against full-length LT ( Figure S4 in Text S1 ) . The C-terminal half of FAM111A ( residues 336 to 611 ) was necessary and sufficient for interaction with LT ( Figure 4B ) . This region of FAM111A contains a trypsin-like serine peptidase domain including the conserved catalytic triad of histidine , aspartate , and serine residues [27] . Homologs of FAM111A exist in several mammalian species including mouse , rat , and rhesus monkey . FAM111A is also highly similar to its paralog human FAM111B ( Gene ID: 374393 ) with 43% identity in the C-terminal 330 residues encompassing the peptidase domain and trypsin-like catalytic triad . To confirm that LT could bind to FAM111A in human U-2 OS cells , we performed immunoprecipitations with antibodies specific for FAM111A or FAM111B . An antibody for FAM111A detected a 70 kDa band that was reduced upon shRNA-mediated knockdown of FAM111A ( Figure 5A ) . An immunoprecipitation for LT co-precipitated FAM111A and the reciprocal immunoprecipitation for FAM111A co-precipitated LT ( Figure 5A ) . FAM111A also co-precipitated the LT fragment C-TERM ( Figure 5B ) . Given the similarity between FAM111A and FAM111B we tested if LT could bind to FAM111B . However , we were unable to detect co-precipitation of FAM111B by LT in U-2 OS cells ( Figure 5C ) . This result is consistent with the MudPIT and iTRAQ analyses that only detected FAM111A and not FAM111B in association with LT . To extend our observations to AGMK cells , we immunoprecipitated LT and FAM111A from lysates prepared from CV-1P , BSC40 and U-2 OS cells infected with wild type SV40 virus ( Figure 5D ) . LT co-precipitated FAM111A from CV-1P and BSC40 cells as well as U-2 OS cells . Notably , expression of FAM111A was similar in CV-1P and BSC40 cells and LT was able bind to FAM111A in both cell types . We next examined how LT C-terminal mutations affect binding to FAM111A . The SV40 point substitution mutants T701A and K697R show wild-type host range activity , while the SV40 host range mutants HR684 and dl1066 cannot produce plaques in CV-1P cells [9] , [10] . FAM111A co-precipitated wild-type LT as well as T701A and K697R mutants from CV-1P cells , but binding to the host range mutant LT HR684 and dl1066 was substantially reduced ( Figure 5E ) . We sought to characterize FAM111A expression . Differential cellular extraction revealed that FAM111A was present in the nuclear and cytoplasmic fractions of U-2 OS cells ( Figure 6A ) . Prior work revealed that the FAM111A promoter was bound by the DREAM complex in G0 or quiescent T98G cells [14] . Given that expression of DREAM target genes are regulated in a cell cycle-dependent manner , we examined mRNA expression profiles of cell cycle synchronized T98G cells . FAM111A levels were reduced in serum-starved G0 cells and increased 20 hr after serum addition when cells were enriched for S phase ( Figure 6B and Figure S5 in Text S1 ) . We identified 79 genes that exhibited cell cycle expression profiles similar to that of FAM111A ( Pearson correlation coefficient R>0 . 9 ) . This FAM111A gene set was significantly enriched for the GO term “M phase of mitotic cell cycle” ( Figure 6C ) . Similar to the mRNA levels , FAM111A protein levels were lowest in serum starved T98G cells and increased as cells progressed towards the G2/M phase of the cell cycle ( Figure 6D ) . The pattern of FAM111A protein expression more closely resembles the expression patterns of late cell cycle genes such as CDK2 and Cyclin B1 than early cell cycle genes such as E2F1 , Cyclin E and p107 ( Figure 6D ) . Binding of LT to p53 and RB serves to inactivate their growth suppressing functions . By analogy , LT binding to FAM111A might serve to inactivate the host range restriction function of FAM111A , thereby promoting increased and sustained viral gene expression . If so , then expression of the SV40 LT C-terminal region should have the same effect on virus replication as reduced FAM111A expression . Cells expressing LT C-TERM showed eight-to-ten fold increases of early ( LT ) and late ( VP1 ) viral transcripts from the HR684 viral DNA relative to cells without LT C-TERM ( Figure 7A ) . Knockdown of FAM111A also resulted in an eight-to-ten-fold increase in early ( LT ) and late ( VP1 ) viral mRNA expression compared to non-targeting siRNA controls ( Figure 7A ) . To evaluate longer-term effects of FAM111A knockdown on viral gene expression , we generated U-2 OS and CV-1P cell lines stably expressing two different shRNAs specific ( sh-1 or sh-2 ) for FAM111A or vector control . The reduction in FAM111A mRNA and protein expression mediated by sh-2 was slightly more effective than sh-1 in both human and monkey cells ( Figures 7B and 7C ) . In control cells transfected with HR684 viral DNA , LT expression was detectable 48 hours after transfection but was markedly reduced by 96 hours . In contrast , LT expression persisted for 96 hours after transfection in U-2 OS ( Figure 7B ) and CV-1P ( Figure 7C ) cells depleted of FAM111A by sh-1 or sh-2 . Decreased FAM111A levels results in sustained host range mutant viral gene expression consistent with the effects of expression of the LT C-terminal region ( Figure 1B ) . We examined the effects of depletion of FAM111A on lytic infection by host range mutant virus . DNA corresponding to wild-type SV40 or host range mutants HR684 and dl1066 was transfected into CV-1P cells expressing shRNAs targeting FAM111A . Wild-type SV40 was capable of inducing plaque formation in control CV-1P cells and in cells containing sh-1 or sh-2 against FAM111A ( Figure 7D ) . Although the relative number of plaques produced by wild-type SV40 was not markedly affected by depletion of FAM111A , the size of the plaques were consistently larger and appeared more rapidly than in the vector control cell line ( Figure 7D and data not shown ) . The two host range mutant viruses could not induce lytic infection in the CV-1P vector control cell line ( Figure 7D ) , but could form plaques in the two FAM111A-depleted CV-1P cell lines . A single burst assay quantified the effect of FAM111A depletion on virus yield in restrictive CV-1P cells . Cells expressing shRNAs targeting FAM111A or vector control were infected with wild-type SV40 or host range mutant virus , dl1066 , at a multiplicity of infection ( MOI ) of 3 . Virions were harvested at several intervals and quantified by plaque assay in permissive BSC40 cells . The wild-type SV40 virus yield was similar in the presence or absence of FAM111A . In contrast , the host range virus yield was negligible in control CV-1P cells but was comparable to wild-type SV40 virus yield when FAM111A was depleted with either sh-1 or sh-2 ( Figure 7E ) . AGMK cells can support human adenovirus replication only when co-infected with SV40 [28] . It has been long recognized that the C-terminal region of LT contains a helper function that permits human adenovirus infection of monkey cells [12] , [29] . Since depletion of FAM111A or expression of the C-terminal region of LT can overcome the host range restriction in CV-1P cells , we investigated the role of FAM111A in adenovirus infection . Knockdown of FAM111A supported Adenovirus 5 ( Ad5 ) replication as measured by increased amounts of the adenoviral hexon protein in FAM111A-depleted CV-1P cells ( Figure 8A ) . Infection with Ad5 led to plaque formation in the FAM111A-depleted but not in the control CV-1P cell lines ( Figure 8B ) . These results indicate a critical role for FAM111A in restriction of SV40 and adenovirus replication .
The propensity of viruses to replicate in host cells depends on their ability to manipulate key host defenses . The multifunctional SV40 LT viral protein encodes discrete domains required for viral replication including origin DNA binding , helicase activity , and the ability to hijack critical host proteins . The LT C-TERM domain is necessary for evading host range restriction in AGMK cells . We demonstrate that the host protein , FAM111A , plays a critical role in restricting viral replication , and that the LT C-terminal region binds to FAM111A to overcome this effect . Proteomic identification of LT associated proteins confirmed several known co-complex associations including p53 and RB ( Figure 9 ) . LT bound to all three members of the RB family of proteins . In contrast , LT was unable to bind to any of the MuvB subunit proteins ( LIN9 , LIN37 , LIN54 , LIN52 , or RBBP4 ) , indicating that LT can disrupt the p130-containing DREAM complex . In keeping with the ability to disrupt the DREAM complex , LT led to increased expression of DREAM target genes ( Figure 2B ) . Most intriguingly , we identified an interaction of the LT C-terminal region with FAM111A and provided several lines of evidence that this interaction contributes to the host range function of LT . FAM111A binds specifically to LT , as demonstrated by two mass spectrometry approaches , Y2H analysis , and reciprocal co-immunoprecipitation . Furthermore , we demonstrated that expression of the LT C-terminal domain or depletion of FAM111A in restrictive CV-1P cells led to sustained viral gene expression and infectious virion formation by host range mutant SV40 viruses . This data strongly supports the model that FAM111A functions as a host range restriction factor that is specifically counteracted by binding to the C-terminal region of LT . In addition , FAM111A depletion enabled human Ad5 to replicate in AGMK cells consistent with the model that FAM111A contributes to viral host restriction . The observation that host range mutant viruses can productively infect permissive BSC40 cells but not restrictive CV-1P cells suggests that these cells differ in a factor that determines cellular susceptibility to viral infection . However , levels of FAM111A were not appreciably different between CV-1P and BSC40 cells , and LT could bind to FAM111A in both cell types . It is possible that small differences in FAM111A levels or activity could affect viral replication or the efficiency of host range restriction . For example , we observed that FAM111A mRNA and protein levels are regulated in a cell cycle dependent manner with the lowest expression during G0 or the quiescent phase with peak expression during G2/M phase . It is possible that differences in the proliferation rate or the cell cycle dependent expression of FAM111A in BSC40 and CV-1P cells could account for the restrictive phenotype . LT-mediated inhibition of FAM111A activity to promote viral replication is consistent with our observations that loss of FAM111A expression by RNAi-mediated knockdown rescues the host range phenotype . FAM111A is predicted to contain a trypsin-like serine peptidase domain . The conservation of the catalytic triad in the FAM111A primary sequence suggests that the protein may act as a specific peptidase . In a simple model , LT binding could inhibit the FAM111A peptidase activity . Although LT binds to the peptidase domain , there is no evidence that LT itself undergoes proteolysis or is a substrate of FAM111A and the exact role of FAM111A remains to be elucidated . It should be noted that several known LT-interacting proteins , including RB , p53 , FBXW7 and CDC73 , are bona fide tumor suppressors . The FAM111A locus has been recently associated with prostate cancer susceptibility in a subset of the Japanese population [30] raising the possibility that FAM111A may play a role in tumorigenesis .
BSC40 ( gift from J . Pipas , University of Pittsburgh ) , CV-1P [31] , T98G [32] and U-2 OS [33] cells were cultured in Dulbecco's modified Eagle's medium ( DMEM ) ( Cellgro ) supplemented with 10% Fetal Clone-I serum ( HyClone ) , penicillin and streptomycin . Cells were transfected using Lipofectamine 2000 transfection reagent ( Invitrogen ) according to the manufacturer's protocol . SV40 genomic DNA ( strain 776 ) was cloned into the BamH1 site of pBluescript KS ( Stratagene ) for propagation in bacteria . Wild type LT cDNA was transiently expressed from the pSG5 vector . The C-terminal fragment of LT was transiently expressed from the pVAX1 expression vector ( Invitrogen ) . The C-TERM construct contained LT residues 627–708 in frame with an N-terminal hemagglutinin ( HA ) epitope tag ( YPYDVPDYA ) and the SV40 nuclear localization signal ( NLS ) ( SPKKKRKVED ) cloned into the pWZL retroviral vector [10] , [34] . Full-length and truncated LT containing N-terminal HA and FLAG epitope tags were expressed from the pMSCV retroviral vector ( gift from Matthew Sowa and Wade Harper [35] ) . siRNA oligonucleotides were purchased from Dharmacon . Lentiviral vectors ( pGIPZ ) with shRNA directed against FAM111A were obtained from Open Biosystems . The sequences of siRNA and shRNA are provided in Supplementary experimental procedures in Text S1 . The following antibodies were used: LT mouse monoclonal antibodies PAb419 and PAb901 [36]; HA mouse monoclonal antibody HA-11 ( Covance ) ; VP1 rabbit polyclonal antibody supplied by N . Christensen ( Pennsylvania State University ) ; FAM111A antibodies BL8623 and BL8624 , FAM111B antibodies BL8627 and BL8630 and SET antibodies were generated by Bethyl Labs . Antibodies to p107 , Cyclin E , CDK2 , Cyclin B1 and E2F1 were obtained from Santa-Cruz Biotechnology . Tubulin and Lamin A/C antibodies were obtained from Cell Signaling . For affinity purification followed by mass spectrometry , an anti-HA affinity matrix ( Pierce ) was used in combination with the HA eluting peptide ( Roche ) and anti-FLAG beads ( Sigma ) were used in combination with 3XFLAG peptide ( Sigma ) . Whole cell lysates were prepared in EBC buffer ( 50 mM Tris-HCl [pH 8 . 0] , 150 mM NaCl , 0 . 5% Nonidet P-40 ) supplemented with protease inhibitor cocktail set I ( Calbiochem ) and phosphatase inhibitor cocktail ( Sigma ) . The Subcellular Protein Fractionation Kit for Cultured Cells was used for nuclear/cytoplasmic fractionation was used ( Thermo Scientific ) . Membranes were blocked and incubated with the appropriate primary antibody in TBS-T overnight at 4°C . Detection of proteins was performed with horseradish peroxidase-conjugated secondary goat antibody ( Pierce ) in TBS-T and enhanced chemiluminescence ( Pierce ) . For immunoprecipitations , whole cell lysate was incubated with antibodies and protein A-Sepharose beads overnight at 4°C . Immune complexes were washed four times with EBC and boiled in sample buffer . See Supplementary Data . Yeast two-hybrid matrix-style experiment with LT and FAM111A as bait or prey was essentially as previously described [37]–[40] . Cells at 80% confluency were infected with wild type SV40 diluted in DMEM supplemented with 2% Fetal Clone-I serum ( HyClone ) , penicillin and streptomycin for two hours . BSC40 and CV-1P were infected at a multiplicity of infection ( MOI ) of 0 . 125 and U-2 OS at MOI of 0 . 5 . SV40 plaque assays were as previously described [41] , [42] with additional details in Supplemental Data . For Ad5 infection , cells were seeded at 400 , 000 per well on 6 well plates and infected with serial dilutions of Ad5 diluted in DMEM containing 2% FBS for 2 hours . Plaque assays were as described for SV40 . For detection of the hexon protein , cells infected with Ad5 were stained 48 or 72 hours after infection ( Adeno-X rapid titer kit , Clontech ) . Cell images were quantified using ImageJ software . U-2 OS cells were transfected with control siRNA and total RNA was extracted using TRIzol ( Invitrogen ) and purified in RNeasy columns ( Qiagen ) . RNA integrity was determined using a Bioanalyzer ( Agilent ) . Gene expression was assayed using Human Genome U133 Plus 2 . 0 arrays ( Affymetrix ) in a single batch . Microarray intensities were normalized using robust multi-array averaging ( RMA ) through the affy package in R/Bioconductor . Differential expression was determined using the limma package [43] . The complete set of expression profiling microarray data can be accessed from the Gene Expression Omnibus ( GEO ) repository GSE40567 . To select genes for clustering , differential expression was tested between all pairwise comparisons and all genes whose expression changes were statistically significant in two or more comparisons were retained ( p<0 . 05 after Benjamini-Hochberg correction for multiple testing ) . Next , all genes that were differentially expressed in any T1 , T6 , T8 or T16-expressing cells compared to the vector control cells were adjoined to the previous set of genes . This resulted in a final set of 430 unique HUGO gene symbols . The expression profile of each gene was determined by taking the median expression levels of all probesets annotated to that gene . All the profiles were mean-centered and scaled by the standard deviation before using the mclust package to cluster the genes [44] . Reverting to the original RMA-normalized data , the gplots package was used to visualize the heatmap of fold changes for each gene relative to vector control . Previous microarray profiling of IMR90 normal human fibroblasts transduced with either GFP or SV40 LT was incorporated into the heatmap in the following way . Data from Human Gene 1 . 0 ST arrays ( Affymetrix ) was preprocessed as described [16] . The genes in each of the fourteen clusters that also had a corresponding probeset on the Human Gene 1 . 0 ST array were included in the heatmap . The IMR90 column on the heatmap ( Figure 2B ) shows the log of the fold change of the SV40 LT-transduced IMR90 cell lines relative to GFP . The three columns of fold changes ( T1 , T16 , and IMR90 ) were hierarchically clustered and the dendrogram was constructed by optimal leaf ordering using the seriation package in Bioconductor . Functional enrichment was determined using FuncAssociate 2 . 0 . Enrichment for custom gene sets was computed using Fisher's exact test , and p-values were corrected for multiple testing using the Benjamini-Hochberg method . GSEA was run using the Java-based desktop application . Probesets were collapsed to gene symbols using median levels . Four combinations of parameters were tried for each run of GSEA: genes were ranked by either signal-to-noise ratio or by t-test , and the p-value was estimated by permuting either sample or gene labels . Only GSEA runs that resulted in significant p-values across all four parameter sets were retained for further interpretation . Therefore , although the enrichment score traces and p-values depicted in the figures correspond specifically to t-test ranking and gene set permutation , these gene sets were significant among all parameter combinations tried . The DREAM and B-MYB/MuvB gene sets were extracted from [14] , [15] and the p53 target gene set corresponds to the “V$P53_02” gene set in the Molecular Signatures Database ( MSigDB ) .
|
Viruses have evolved numerous mechanisms to counteract host cell defenses to facilitate productive infection . Simian Virus 40 ( SV40 ) replication depends on specific interactions between large T antigen ( LT ) and a wide variety of host cell proteins . Although the LT C-terminal region has no evident enzymatic activity , mutations or deletions of this region significantly reduce the ability of the virus to replicate in restrictive cell types . Here , we identified host proteins that bind to LT and determined that the LT C-terminal region binds specifically to FAM111A . This physical interaction was required for efficient viral replication and sustained viral gene expression in restrictive cell types . In addition , RNAi-mediated knockdown of FAM111A levels in restrictive cells restored lytic infection of SV40 host range mutants and human adenovirus . These results indicate that FAM111A plays an important role in viral host range restriction . Our study provides insights into the viral-host perturbations caused by SV40 LT and the interaction of viruses with host restriction factors .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biochemistry",
"protein",
"interactions",
"virulence",
"factors",
"and",
"mechanisms",
"proteins",
"virology",
"biology",
"microbiology",
"host-pathogen",
"interaction"
] |
2012
|
Identification of FAM111A as an SV40 Host Range Restriction and Adenovirus Helper Factor
|
Trypanosoma cruzi is the etiologic agent of Chagas disease throughout the Americas . Few population-level studies have examined the epidemiology of canine infection and strain types of T . cruzi that infect canines in the USA . We conducted a cross-sectional study of T . cruzi infection in working hound dogs in south central Texas , including analysis of triatomine vectors collected within kennel environments . Paired IFA and Chagas Stat-Pak serological testing showed an overall seroprevalence of 57 . 6% ( n = 85 ) , with significant variation across kennels . Dog age had a marginally significant effect on seropositivity , with one year of age increase associated with a 19 . 6% increase in odds of being seropositive ( odds ratio 95% CI 0 . 996–1 . 435; p = 0 . 055 ) . PCR analyses of blood revealed 17 . 4% of dogs harbored parasite DNA in their blood , including both seronegative and seropositive dogs . Molecular screening of organs from opportunistically sampled seropositive dogs revealed parasite DNA in heart , uterus , and mammary tissues . Strain-typing showed parasite discrete typing units ( DTU ) TcI and TcIV present in dog samples , including a co-occurrence of both DTUs in two individual dogs . Bloodmeal analysis of Triatoma gerstaeckeri and Triatoma sanguisuga insects collected from the kennels revealed exclusively dog DNA . Vector infection with T . cruzi was 80 . 6% ( n = 36 ) , in which T . gerstaeckeri disproportionately harbored TcI ( p = 0 . 045 ) and T . sanguisuga disproportionately harbored TcIV ( p = 0 . 029 ) . Tracing infection status across dog litters showed some seropositive offspring of seronegative dams , suggesting infection of pups from local triatomine vectors rather than congenital transmission . Canine kennels are high-risk environments for T . cruzi transmission , in which dogs likely serve as the predominant parasite reservoir . Disease and death of working dogs from Chagas disease is associated with unmeasured yet undoubtedly significant financial consequences because working dogs are highly trained and highly valued .
Chagas disease in humans and dogs is caused by the hemoflagellate protozoan Trypanosoma cruzi . Active transmission cycles of the parasite occur across the southern USA , where infected triatomine ‘kissing bug’ vectors and wildlife reservoirs co-occur [1–4] . Canines in particular have been shown to be important reservoir and maintenance hosts throughout the Americas ( see [5] for a comprehensive review ) . Although epidemiological studies of canine infection with T . cruzi in the southern USA are limited , cases are widespread , especially in Texas [4 , 6–11] . The first cases of canine T . cruzi infection in the USA were documented in Texas [9] , and a recent retrospective study reported cases from across the state [8] . Studies have revealed anti-T . cruzi seroprevalences of 7 . 5% in stray dogs and 8 . 8% in shelter dogs across Texas [10 , 12] . However , given variation in clinical presentation in infected dogs , which ranges from asymptomatic to acute death or chronic heart disease [13] , the veterinary implications of canine T . cruzi infections are uncertain . The absence of a canine vaccination or canine antiparasitic treatments against T . cruzi further complicates clinical case management . Infection with T . cruzi can occur through the introduction of infected triatomine insect feces into skin lesions as the bug defecates on the host during or shortly after blood feeding . Oral transmission to dogs and wildlife may result from consumption of infected bugs or infected prey species [6 , 14 , 15] . Although congenital transmission in canines has been documented [16–18] , the frequency with which this occurs is unknown . Accordingly , owners of seropositive breeding bitches are left with little information to guide breeding programs , except for the option of removal of positive females from breeding roles [19] . Although serologic testing is a common tool for diagnosing T . cruzi infections in dogs , parasitemia is known to peak as early as two weeks prior to detectable antibody levels . In dogs , experimental studies indicate that parasitemia occurs within days to four weeks after initial infection [20–23] , with development of anti-T . cruzi antibodies detected at 15 days to 4 weeks post infection [23–25] . Further , T . cruzi genetic strain differences may play a role in disease outcomes in canines [21 , 25–27] , with genetic variation occurring across geographic regions [28] , yet there have been few investigations of which strains infect dogs in the USA [29] . Veterinarians and dog owners are faced with increasing diagnoses of canine T . cruzi infections , but a limited ability to understand the veterinary and public health consequences . Our objective was to compare multiple serological and molecular biology techniques to detect and characterize T . cruzi infections in a cross-sectional analysis of working hound dogs in a parasite-endemic region . We documented an active T . cruzi transmission cycle in kennels in south central Texas .
Research use of all samples from dogs was secondary to collection for diagnostic purposes; the Texas A&M University Institutional Animal Care and Use Committee granted a formal waiver of ethical approval . This study was motivated by unexplained deaths of several dogs from a large network of working hound dogs used for various scent detection functions , mainly across Texas . Several dogs died within a short time period , and postmortem histopathologic findings indicated that canine T . cruzi infection was the probable cause of the deaths . A representative histopathology report from a T . cruzi-seropositive six-year old female hound that died in August 2013 showed myocarditis and epicarditis—lesions consistent with chronic Chagas disease—although no protozoal amastigotes were observed in the myocardium or any other tissue examined ( kidney , mediastinal lymph node , lung , liver , or spleen ) . Using a cross-sectional study design , we assessed and sampled 86 working dogs from three multi-dog kennels in the network: 26 dogs from kennel A , 31 dogs from kennel B ( where the sudden deaths and T . cruzi infection diagnosis had occurred ) , and 29 dogs from kennel C , which comprised all dogs in residence at these kennels . All dogs were Coonhounds , most were bred by the facilities , and ages ranged from approximately 6 months to 13 years . Dogs were housed in indoor-outdoor , open air , cement/concrete kennels located within a 50 km radius of each other in south central Texas counties . Canines had limited travel history , mainly within Texas . General physical examinations ( auscultation , rectal temperature , mucous membrane color , and generalized palpation ) were conducted , and blood samples were collected between July and September , 2013 . Over the following several months , opportunistic postmortem samples of blood and other tissues ( heart , mammary gland , testicle , uterus ) were collected from dogs euthanized for reasons unrelated to this study . Pedigree lineage records were analyzed to determine relationships among sampled dogs ( i . e . , dams and littermates ) . Triatomine bugs were opportunistically collected from kennels in the network by kennel staff and pest control operators in summer 2013 . Serum aliquots were tested for anti-T . cruzi antibodies using indirect fluorescent antibody ( IFA ) testing at the Texas Veterinary Medical Diagnostic Laboratory ( TVMDL; College Station , TX ) . All samples were screened for the presence of anti-T . cruzi antibodies at 1:20 , 1:80 and 1:160 dilutions . According to TVMDL protocols , titer values of 20 or greater were considered positive for antibody . The remaining serum was stored at -20°C until analyzed using the Chagas Stat-Pak chromatographic dipstick test ( ChemBio , NY ) . The Chagas Stat-Pak test has previously been used for antibody-detection test in dogs [12 , 25 , 30] , and may offer an economical alternative for rapid screening of population , as had been suggested of a similar rapid test [31] . Stored serum samples were tested according to manufacturer’s instructions and any development of a band at 15 minutes was considered positive for antibody . Band strength was noted as faint , medium , or bold . Samples positive using both IFA and Chagas Stat-Pak dipstick tests were considered seropositive in the calculation of population-level seroprevalence . An extraction kit ( E . Z . N . A . Tissue DNA kit , Omega Bio-Tek , Norcross , GA ) was used to extract DNA from 250 μL of clotted blood from dogs for which serology testing was also performed . Extracted DNA was analyzed using qPCR to detect parasite DNA . Samples were first screened for presence of T . cruzi DNA using the real-time PCR Cruzi 1/Cruzi 2 primer set and Cruzi 3 probe [32 , 33] . This PCR amplifies a 166-bp region of a repetitive nuclear DNA sequence , and is sensitive and specific for T . cruzi when compared to other PCR techniques [34] . A Stratagene MxPro3000 instrument ( Agilent Technologies , Santa Clara , CA ) was used to amplify DNA under previously described thermocycling parameters [32] , except with a 3-minute initial denaturation . Reactions consisted of 5 μL of template DNA , primers at a final concentration of 0 . 75 μM each , 0 . 25 μM of probe , and iTaq University Probes Supermix ( BioRad Laboratories , Hercules , CA ) , in a total volume of 20 μL . Machine-calculated thresholds and reaction curves were visually checked to assure successful amplification . Samples producing cycle threshold ( Ct ) values of less than 34 were considered potential positives and subjected to further testing for confirmation and discrete typing unit ( DTU ) -typing . A multiplex qPCR was used to confirm T . cruzi infection and determine T . cruzi DTU based on amplification of the nuclear spliced leader intergenic region ( SL-IR ) with the use of a panel of DTU-specific probes [35] . Reactions were 20 μL total volume using a QIAGEN Multiplex PCR Kit ( QIAGEN , USA ) , run using the following protocol: 15 minutes at 95°C followed by 40 cycles of 95°C for 30 seconds and 60°C for 1 minute . Reactions were run on a BioRad CFX96 ( Hercules , CA , USA ) . Both FAM and HEX dyes were used as previously described [35]; however , due to differing instrument capabilities , our reactions differed from published protocol [35] by substituting Cy5 and Tex615 dyes ( Integrated DNA Technologies , Inc . , Coralville , IA , USA ) for Quasar670 and CAL Fluor Red610 , respectively . Samples that yielded amplification curves on both the Cruzi 1/2/3 qPCR and the SL-IR qPCR were interpreted as PCR-positive in our analyses . Samples that fluoresced with FAM were classified as TcI , whereas samples that fluoresced with Tex615 were classified as TcIV; samples that fluoresced with both FAM and Tex615 were classified as mixed TcI/TcIV . Although different TcIII isolates have previously resulted in fluorescence of either Quasar670 alone or both Quasar670 and CAL Fluor Red610 , the TcIV isolates previously tested were shown to only cause CAL Fluor Red610 fluorescence [35] . None of the samples we tested resulted in Quasar670 ( here , Cy5 ) fluorescence . Supported by the subset that we definitively typed using TcSC5D gene sequencing ( below ) , we classified samples with CAL Fluor Red610 ( here , Tex615 ) fluorescence as TcIV . In addition to probe-based DTU-typing , as an additional method to investigate strain-typing , a subset of samples were amplified using a primer set that amplifies a region of the TcSC5D gene , a putative lathosterol/episterol oxidase [36] . The 832-bp amplicons were visualized on 1 . 5% agarose gel with ethidium bromide , and sequenced using Sanger sequencing ( Eton Bioscience Inc . , San Diego , CA , USA ) . Geneious version 8 [http://www . geneious . com [37]] was used to visually review chromatographs and sequences , align forward and reverse sequences , and examine locations of key SNPs to determine DTUs [36] . Negative controls were included in each set of DNA extractions and PCR reactions . Positive controls included T . cruzi DNA extracted from a TcI isolate Sylvio X10 CL4 ( ATCC 50800 , American Type Culture Collection [ATCC] , Manassas , VA , USA ) , an untyped isolate cultured from a published Texas canine case [38] , a TcIV isolate from an infected Texas raccoon [39] , and TcIV isolates from T . sanguisuga and T . gerstaeckeri from Texas . Samples positive using both IFA and Chagas Stat-Pak dipstick tests were considered seropositive in the calculation of population-level seroprevalence . Blood samples classified positive by Cruzi 1/2/3 and SL-IR qPCRs were considered positive in calculation of population-prevalence of T . cruzi DNA in blood samples . To evaluate the relationship between potential risk factors and positive canines , bivariable analyses were performed using chi-squared , Fisher’s exact tests , or t-tests . Variables assessed were kennel ( A , B , or C ) , age , and sex . Logistic mixed effect regression models were built using the lme4 package in Program R [40] to further investigate risk factors with p < 0 . 25 in the initial screening and risk factors with justification for inclusion based on published data . To determine the variation in positive dogs across age and sex , kennel was included as a random effect . To determine the variation in positive dogs across kennels , age was included as a random effect . Factors with values of p < 0 . 05 were considered significant . Odds ratios and 95% confidence intervals were calculated . Separate models were built for anti-T . cruzi antibody status and blood T . cruzi PCR status . Tissues collected opportunistically from euthanized dogs were preserved in 10% neutral buffered formalin . Formalin-preserved samples were submitted for histopathologic examination with routine hematoxylin and eosin staining at the TVMDL and reviewed by a pathologist . Additionally , DNA was extracted from approximately 1 cm3 pieces of various fresh tissues using the same methods as the molecular processing of dog blood samples as described above . Given variation in parasite localization within tissues , up to five independent subsamples per tissue were tested . Bugs were identified to species using morphologic features [41]; sex and evidence of a recent bloodmeal were noted . After bugs were washed in 10% bleach solution and rinsed in distilled water , sterile instruments were used to dissect the bugs and isolate hindgut material . DNA was extracted from hindguts and tested for T . cruzi DNA and DTU determination using the same methods described above . Two-sample tests for equality of proportions with continuity corrections were used to compare the proportions of TcI and TcIV between infected T . gerstaeckeri and T . sanguisuga . In order to determine the source of recent bloodmeals , hindgut DNA was subjected to PCR amplification of host cytochrome B sequences using previously published primers and cycling conditions [42 , 43] . Reactions included 3 μL template DNA , primers at final concentrations of 0 . 66 μM each , and FailSafe PCR Enzyme Mix with PreMix E ( Epicentre , Madison , WI ) in a final reaction volume of 50 μL . Amplicons were visualized on 1 . 5% agarose gel with ethidium bromide , and sequenced using Sanger sequencing ( Eton Bioscience Inc . , San Diego , CA , USA ) . Resulting sequences were compared to existing sequences using Basic Local Alignment Search Tool ( National Center for Biotechnology Information , US National Library of Medicine ) . Sequences of the TcSC5D genetic region amplified from samples are available at NCBI GenBank; accession numbers are KX594832-KX594840
General physical examinations ( auscultation , rectal temperature , mucous membrane color , and generalized palpation ) of the dogs at the time of sampling did not reveal any significant findings suggestive of clinical presentation of T . cruzi infection . The birthdate was known for 80 of the 86 dogs in the three kennels . Ages ranged from 6 months to 13 years , with a mean and median of 3 . 96 years and 3 . 58 years , respectively . Mean age and standard deviation at kennels A , B , and C was 4 . 05 ± 3 . 31 years ( n = 26 ) , 3 . 77 ± 2 . 64 years ( n = 27 ) , and 4 . 06 ± 2 . 68 years ( n = 27 ) , respectively . There were 15 dogs less than 1 year old ( 18 . 6% of 80 ) . There were 39 males ( 45 . 3% ) and 47 females ( 54 . 7% ) . At the time of the cross-sectional blood sampling , there were seven dams with a total of seventeen offspring that were included in the study ( Fig 1 ) . A total of 56 of 86 ( 65 . 1% ) dogs had an antibody titer value of 20 or greater on IFA , and 53 of 85 ( 62 . 4% ) dogs were reactive on the Chagas Stat-Pak ( Table 1 ) . Combined , 49 of 85 dogs were positive on both antibody detection tests , yielding a seroprevalence of 57 . 6% . A single sample was positive on IFA with an antibody titer of 160 , but was not tested on the Chagas Stat-Pak and was therefore not included in the overall seroprevalence estimate . There were 10 dogs positive on only one test and negative on the other; these dogs with discordant results were considered seronegative for the purpose of this study . Of these ten discordant samples: 4 dogs were negative on IFA but had faint ( positive ) lines on the Chagas Stat-Pak , and 6 dogs were negative on Chagas Stat-Pak but had IFA titer values of 20 ( 3 dogs ) , 80 ( 1 dog ) , and 160 ( 2 dogs ) . Overall seroprevalences at each kennel were: 46 . 2% at kennel A ( n = 26 ) , 71 . 0% at kennel B ( n = 31 ) , and 53 . 6% at kennel C ( n = 28 ) . Analysis of canine serostatus in relation to lineage revealed both positive and negative littermates born to positive and negative dams . Four 20-month old pups were seropositive , despite the concurrently-tested dam being seronegative ( Fig 1 ) . In addition to the main 86 dogs of this study , a litter of young pups was opportunistically sampled to gather data on serostatus and PCR status of neonatal pups born to a seropositive dam . A litter of six pups was born to a female who tested serologically positive ( ≥ 1:160 ) nine months previously , and these pups were serially sampled twice over 4 weeks . The dam accidentally smothered one of the pups one day after birth . Blood samples from that pup were PCR negative for parasite , although testing on Chagas Stat-Pak gave a faint ( positive ) band . The other five pups had blood sampled two weeks after birth and tested on Chagas Stat-Pak: two gave very faint ( positive ) bands and three were negative . At one month of age , all five were negative on Chagas Stat-Pak . None of the two week or one-month samples were PCR positive . PCR analysis of 86 DNA extracts of blood clots revealed 15 ( 17 . 4% ) positive samples . There were 23 samples with Ct values less than 34 on the initial Cruzi 1/2/3 qPCR and subjected to attempted amplification using the SL-IR PCR , of which 15 samples tested positive using the SL-IR PCR assay . Overall prevalences of PCR-positive dogs at each kennel were: 15 . 4% at kennel A ( n = 26 ) , 25 . 8% at kennel B ( n = 31 ) , and 10 . 3% at kennel C ( n = 29 ) . Of the 15 PCR positive blood samples , the SL-IR assay revealed 9 TcI , 5 TcIV , and 1 TcI/TcIV mixed infections . Amplification and sequencing of the TcSC5D gene DNA target was successful in five blood samples . TcSC5D and SL-IR results were congruent , except one of the TcIV TcSC5D findings was characterized as a mixed TcI/TcIV by SL-IR PCR . Using the serological positivity criterion of being positive on both IFA and Stat-Pak assay , serology and PCR findings categorized 12 of 85 dogs ( 14 . 1% ) as both seropositive and PCR positive ( Table 2 ) , 37 dogs ( 43 . 5% ) as seropositive and PCR negative; 3 dogs ( 3 . 5% ) as PCR positive and seronegative; and 33 dogs ( 38 . 8% ) as both seronegative and PCR negative . One dog sample with a 160 IFA titer was not run on Chagas Stat-Pak; that dog was PCR negative . Two dogs that were PCR positive but did not meet the positivity criterion on both serological assays were positive on IFA with antibody titers of 20 and 80 . Of the 37 seropositive , PCR-negative dogs , only one was less than 1-year old , indicating that seropositive dogs were not positive due to maternal antibodies . Bivariable analysis of putative risk factors for canine seropositivity indicated that kennel ( p = 0 . 146 ) and age ( p = 0 . 077 ) , but not sex ( p = 0 . 535 ) , were associated with p-values below the threshold significance level for inclusion in the regression model ( Table 3 ) . Bivariable analysis of putative risk factors for blood PCR-status indicated that neither kennel ( p = 0 . 310 ) , age ( p = 0 . 344 ) , nor sex ( p = 0 . 863 ) were associated with p-values below the threshold significance level for inclusion in the regression model ( Table 3 ) . Nonetheless , all putative risk factors were retained in the regression models based on previous work [14 , 44–46] . In logistic regression models to predict serostatus while including kennel as a random effect , dog age was marginally positively ( p = 0 . 055 ) associated with seropositivity , where one year of increase in canine age was associated with a 19 . 6% increase in the odds of being seropositive , and there was no effect of sex ( p = 0 . 855 ) ( Table 4 ) . The odds of being seropositive were 6 . 6 ( 95% CI 1 . 32–32 . 88 ) times greater for dogs in kennel B than in the referent kennel ( kennel A ) ( p = 0 . 022 ) . In the logistic regression models to predict blood PCR status , none of the putative risk factors of kennel ( p = 0 . 251 and p = 0 . 670 ) , age ( p = 0 . 296 ) or sex ( p = 0 . 637 ) were significant . A total of five tissue samples opportunistically collected from four IFA-positive dogs were examined histologically . Three of four cardiac samples had lesions consistent with chronic canine trypanosomiasis ( Table 5 ) , although no amastigotes were observed in any of the sections . Lesions included cardiomyofiber degeneration ( ranging from minimal to moderate ) , with accumulations of lymphocytes , plasma cells , and rare macrophages . One uterine tissue section was viewed; no amastigotes or significant lesions were observed ( Table 5 ) . T . cruzi DNA was detected in heart , blood , uterus , and mammary gland tissues collected opportunistically from multiple serologically-positive dogs ( Table 5 ) . Three dogs did not have detectable parasite DNA in tested uterine tissue , whereas four of the five samples from the body of the uterus of one dog were PCR positive . Three dogs did not have detectable parasite DNA in tested testicular tissue . One dog ( Dog 88 ) had multiple parasite positive tissues , including heart , blood , uterus , and mammary gland . One dog ( Dog 432 ) had PCR-positive blood and heart tissue . Cardiac and uterine samples from one dog ( Dog 88 ) revealed TcIV in uterine and mammary gland tissue and TcI in cardiac tissue; these results were congruent between SL-IR and TcSC5D DTU-typing methods . One dog ( Dog 432 ) had TcI in heart tissue and TcIV in blood . A total of 44 bugs ( Table 6 ) were opportunistically collected in summer 2013 from the network of working dog kennels , including the three kennels that housed dogs tested in the cross-sectional serological study . Bugs included 16 adult Triatoma gerstaeckeri and 28 adult T . sanguisuga . Of the 36 insects that were tested for T . cruzi , 29 ( 80 . 6% ) were positive , including 16 of 16 ( 100% ) tested T . gerstaeckeri and 13 of 20 ( 65% ) T . sanguisuga . The proportion of infected T . gerstaeckeri that harbored TcI was significantly greater than that of infected T . sanguisuga ( χ2 = 4 . 026; p = 0 . 045 ) , whereas the proportion of infected T . sanguisuga that harbored TcIV was significantly greater than that of T . gerstaeckeri ( χ2 = 4 . 765; p = 0 . 029 ) ; Table 6 ) . Based on visual examination , 30 of the 44 bugs had evidence of a recent bloodmeal in their guts . Of 24 bugs with sufficient bloodmeal volume for successful bloodmeal PCR and Sanger sequencing , all 24 had ≥97% identity to Canis lupus familiaris ( domestic dog ) .
Over half ( 57 . 6% ) of a population of working hound dogs were seropositive for T . cruzi , and 17 . 4% harbored parasite DNA in their blood ( Table 2 ) . Additionally , we documented parasite DNA in heart , mammary , and uterine tissues in dogs from this network . A high ( 80 . 6% ) infection prevalence was found in triatomines recovered from the kennels , and the only bloodmeal host detected in these bugs was dog . Coupled with the documented history of multiple deaths due to Chagas disease in these working dogs , our findings highlight a key role of dog kennels as a nidus of T . cruzi transmission . While Chagas disease impacts many types of dogs across the southern USA , including pet dogs and stray dogs , there is an additional unmeasured yet undoubtedly important financial consequence when Chagas disease impacts working dogs because they are highly trained and have a significant economic worth that results from the value of the duties they perform . The 57 . 6% seroprevalence in these kenneled working dogs is much higher than the 8 . 8% seroprevalence found in a general population of dogs across Texas [12] . This difference is similar to findings in Louisiana , in which kenneled hunting dogs had a seroprevalence of 51 . 6% , which was higher than the 22 . 1% seroprevalence reflected in a general population of dogs in the surrounding area [30] . This population of working dogs we sampled was selected due to recent deaths in a focal kennel and is not representative of all kennels or south Texas . Dogs were considered at high risk to acquiring T . cruzi infection due to the presence of infected vectors in the kennel environment as well as outdoor working settings . All sampled dogs were intensively trained for pack tracking of lost and missing persons; when working , dogs trail intently while running and are undistracted by their surroundings . Therefore , although dogs may also encounter vectors outside the kennel environment , it is most plausible that infection was acquired within the kennels . Older dogs were more likely to be seropositive , with approximately 19 . 6% increase in odds of seropositivity with each additional year of age , although the trend was only marginally significant ( p = 0 . 055; Table 4 ) . Increasing infection with age has been previously reported [7 , 14] and is expected , since older dogs have had longer opportunity to be exposed to T . cruzi and develop life-long seropositivity . Our findings of higher seroprevalence in older dogs are suggestive of an ongoing transmission cycle in these kennels , rather than an emerging recent phenomenon . It is unclear why dogs in kennel B were more likely to be seropositive than dogs in kennel A; one potential risk factor not examined in this study was additional outdoor kennels at kennel B that possibly served as refugia for triatomine bugs . In this study , both serological diagnostic approaches ( IFA and Chagas Stat-Pak ) resulted in similar population-level estimates of seroprevalence ( 65 . 1% vs . 62 . 4% , ( Table 1 ) . However discordant results ( positive and negative results on the same sample across different testing platforms ) occurred in 10 dogs , the majority of which were negative on one test and only faintly positive ( faint band or 20 endpoint titer ) on the other . Although dogs with discordant results were interpreted as seronegative in our study , at least some of these dogs were infected , based on PCR-positive results in 2 dogs , likely reflecting acute infections . This observation underscores the importance of using personal protective equipment when handling canine blood even from seronegative individuals . While the Stat-Pak has not been validated using dogs with known infection histories , this test has shown high sensitivity ( 87 . 2–100% ) and specificity ( 93 . 2–98 . 6% ) in human samples when compared with other serological techniques [47–49]; however , others have found considerable variation and lower sensitivity ( 26 . 6–87 . 5% ) [50] . It is difficult to compare canine infection prevalence across studies because data from the same diagnostic tools may be interpreted differently . For example , whereas we interpreted any development of color to indicate a positive result for the Stat-Pak as per manufacturer’s instructions , others have considered faint band development as negative [30] . Further , in our study , serum dilutions for IFA of 1:20 or higher were considered positive for antibody as per TVMDL reporting standards . In other canine T . cruzi studies , however , dilutions interpreted as positive included those equal to or greater than 1:128 or 1:160 [4 , 8] ( however , see [51] ) ; and one previous study found chronically infected dogs produced positive results of serum dilutions ranging from 1:120 to 1:320 [20] . Faint bands on the Stat-Pak , low antibody titers on the IFA , and discordant results across multiple testing platforms may result from T . cruzi strain type variation , weak immune response , an early , rising antibody response to a recent infection , and variation in test sensitivity or specificity . Imperfect diagnostics and the absence of a gold standard indeed represent one of the major challenges in canine Chagas disease research . Prior research of experimental chronic Chagas disease in dogs has demonstrated that multiple extractions and PCRs are needed to ensure detection of T . cruzi DNA from whole blood samples [52] , although another study found that T . cruzi DNA was more likely to be detected the blood clot ( which was used in this study ) than buffy coat or whole blood samples [53] . One study conducting controlled experimental reinfection in dogs found that parasitemia was not as common in reinfections as it was in initial infection , and that parasitemia profile varied depending upon the individual dog [23] . While PCR of samples does not confirm the presence of whole , viable parasites in the blood , findings of parasite DNA in the blood suggest that positive dogs could potentially be infectious to blood-feeding insect vectors . It is likely the relatively high prevalence of T . cruzi DNA found in this study ( 17 . 4% ) reflects the timing of the blood sampling ( late July ) corresponding with the time of year kissing bugs are most likely to be encountered in Texas [54] . With potential for continued exposure to kissing bugs and repeat infections with T . cruzi , it is possible that dogs with positive serological and positive PCR results ( Table 2 ) could have been recently reinfected . Additional diagnostic difficulties are the result of parasitemia waning after initial infection [55] or lower parasitemic peaks from reinfection [23] . In comparing molecular and histology results , we found that all four PCR-positive hearts subjected to histology were associated with lesions consistent with chronic T . cruzi infection , although no amastigotes were seen in heart samples ( Table 5 ) . The lack of apparent amastigotes is not surprising , however , given that experimental studies have shown parasites are not always histologically detected in cardiac tissue of chronically infected dogs [20 , 23] . Further , T . cruzi strain type can also influence level of cardiac damage and presence of amastigotes [22] , although more research is needed on pathology variation owing to parasite strain . There is an interest , particularly in the canine breeding community , in whether T . cruzi can be sexually transmitted between dogs . We used PCR testing to evaluate testicle and uterine samples from seropositive dogs . None of four testicle samples were positive , but small sample size and conflicting reports in previous literature [56–58] leave us unable to draw firm conclusions . Of four uteri tested , we detected a single positive uterus in which four of the five samples taken from the body of the uterus were positive . The mix of positive and negative samples suggests that T . cruzi distribution in the tissue is not uniform . The detection of parasite DNA in uterine tissue supports previous reports of transplacental transmission of this parasite in dogs and in humans [17 , 59] . In addition to reproductive tissues , the potential for transmammary transmission has been suggested by others [14 , 16] . Our finding of T . cruzi DNA in mammary gland tissue was in a dog that also had evidence of parasites in heart and uterine tissue . Congenital transmission of T . cruzi in dogs has been shown , with one study finding circulating antibodies in 45-day old pups born to experimentally infected dams [17] , and is a concern to breeders . We conducted a cross-sectional study with concurrent sampling of dams and their offspring , and the infection status of dams was not specifically known at the time of whelping . Although this study design limits the ability to draw conclusions about congenital transmission in this setting , the observed patterns of infection across family lines are useful for inferring transmission pathways . For example , we found both seropositive and seronegative littermates from a seronegative dam , supporting the likelihood of local vector-borne transmission . In contrast , we observed several seropositive bitches associated with both seropositive and seronegative offspring; scenarios for which congenital transmission cannot be rule out . None of the PCR-positive dogs in this study were the offspring/dams of any other concurrently-sampled PCR-positive dogs . Additionally , although no parasite DNA was detected in six young pups from a seropositive dam , faint bands produced on the Chagas Stat-Pak test on early blood draws might be the result of maternal antibodies circulating in the pups . Additional research is needed regarding congenital transmission rates and relation between maternal and self-produced anti-T . cruzi antibodies . We found higher infection prevalence ( >80% ) in kissing bugs than recent statewide estimates of 63% and 51% [54 , 60] . Bloodmeal analysis of the triatomines revealed all evaluated bugs had fed on dogs ( Table 6 ) . Other studies in the USA have found evidence of dog blood in triatomines , including bugs associated with houses and dog kennels [61–64] . High infection prevalence in vectors collected from canine quarters , combined with evidence of bugs feeding on dogs , supports vector-host contact and parasite transmission to dogs . Combined with our findings of PCR positive dog blood , it is likely that dogs are the source of vector infection and serve as the main reservoir in this setting . Prevention of canine infection with T . cruzi relies heavily on vector control . Integrated pest management strategies consisting of pesticide use , barrier methods ( netting or mesh around kennels ) , and physical management of dogs ( moving dogs to indoor facilities at night ) have been employed in different areas around Texas . We found TcI and TcIV infections in dogs , including three dogs infected with both DTUs . These dogs may have been re-infected given that the vector populations in the same areas also harbored both strains ( Table 6 ) . Previous strain typing efforts of limited dog samples from the USA have shown almost exclusively TcIV infections [29 , 65] , although one TcI/TcIV mixed infection was documented in a USA dog [29] , and mixed strain infections have been documented in dogs in Columbia [66] . T . gerstaeckeri were disproportionately infected with TcI ( p = 0 . 045 ) , whereas T . sanguisuga were disproportionately infected with TcIV ( p = 0 . 029 ) , in contrast to previous findings of only TcI in limited T . sanguisuga samples from the eastern USA [29] . Differing host preferences and geographic distribution of these triatomine species [60] may put geographically disparate dog populations at risk of acquiring different strains of T . cruzi . Previous research suggests that parasitemia , antibody development , and disease may vary according to strain type of T . cruzi , as well as length of infection and infected host species [21 , 22 , 67 , 68] . We found DTU determination using the probe-based qPCR [35] was more useful than the TcSC5D gene target assay , likely because the latter assay was developed using DNA from pure parasite culture [36] was not optimized for use in field-collected samples with mixed DNA populations [see [39 , 69]] . Although dogs have been shown to be important T . cruzi reservoirs in areas of Latin America [66 , 70 , 71] , with one model of a rural Brazilian village estimating that an infected dog could infect one triatomine per day [72] , the ecological settings of dogs in central Texas may limit their importance as reservoirs of human infections . Dogs in central Texas are typically housed either in a kennel separate and somewhat distant from the human dwelling or indoors in a house constructed with screens and doors that limit bug entry . The infected Texas dogs likely serve as reservoirs within the kennel setting , serving to infect bugs that can subsequently infect other dogs . However , the infectiousness of dogs to bugs has been shown to vary widely , and depends upon a variety of factors , including: dog body condition , coinfections , dog history of infection , vector competence , bug bloodmeal size , and bug feeding duration ( [73] , see [5] for a comprehensive review ) . Given our findings of high infection prevalence in dogs and vectors that fed on dogs , we conclude canine kennels represent a high-risk environment for T . cruzi transmission .
|
The parasite Trypanosoma cruzi can cause Chagas disease in humans and dogs . The parasite is typically spread through the feces of a blood-sucking ‘kissing bug’ insect . Despite many documented cases in dogs across Texas , there are few population-level research studies investigating canine Chagas disease in multi-dog kennels in Texas . We sampled a total of 86 dogs from three kennels in south central Texas . We found 58% of the dogs had antibodies indicating they had been exposed to the parasite , and 17% of the dogs had parasite DNA circulating in their blood . We found that over 80% of kissing bugs collected from the kennels were infected . Further , using a bloodmeal analysis technique , we detected that all tested bugs had recently fed on dog blood . Our findings indicate that dog kennels can serve as an environment where kissing bugs and dogs interact , and that outdoor , multi-dog kennels in areas with kissing bugs are high risk areas for Trypanosoma cruzi transmission in dogs .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"united",
"states",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"geographical",
"locations",
"vector-borne",
"diseases",
"vertebrates",
"parasitic",
"diseases",
"dogs",
"parasitic",
"protozoans",
"animals",
"mammals",
"organisms",
"north",
"america",
"texas",
"protozoans",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"serology",
"artificial",
"gene",
"amplification",
"and",
"extension",
"molecular",
"biology",
"hematology",
"people",
"and",
"places",
"trypanosoma",
"cruzi",
"trypanosoma",
"blood",
"anatomy",
"physiology",
"biology",
"and",
"life",
"sciences",
"amniotes",
"polymerase",
"chain",
"reaction"
] |
2017
|
Epidemiology and Molecular Typing of Trypanosoma cruzi in Naturally-Infected Hound Dogs and Associated Triatomine Vectors in Texas, USA
|
Kaposi’s sarcoma-associated herpesvirus ( KSHV ) is causally associated with Kaposi’s sarcoma , primary effusion lymphoma ( PEL ) and multicentric Castleman’s disease . The IFIT family of proteins inhibits replication of some viruses , but their effects on KSHV lytic replication was unknown . Here we show that KSHV lytic replication induces IFIT expression in epithelial cells . Depletion of IFIT1 , IFIT2 and IFIT3 ( IFITs ) increased infectious KSHV virion production 25-32-fold compared to that in control cells . KSHV lytic gene expression was upregulated broadly with preferential activation of several genes involved in lytic viral replication . Intracellular KSHV genome numbers were also increased by IFIT knockdown , consistent with inhibition of KSHV DNA replication by IFITs . RNA seq demonstrated that IFIT depletion also led to downregulation of IFN β and several interferon-stimulated genes ( ISGs ) , especially OAS proteins . OAS down-regulation led to decreased RNase L activity and slightly increased total RNA yield . IFIT immunoprecipitation also showed that IFIT1 bound to viral mRNAs and cellular capped mRNAs but not to uncapped RNA or trimethylated RNAs , suggesting that IFIT1 may also inhibit viral mRNA expression through direct binding . In summary , IFIT inhibits KSHV lytic replication through positively regulating the IFN β and OAS RNase L pathway to degrade RNA in addition to possibly directly targeting viral mRNAs .
Kaposi’s sarcoma-associated herpesvirus ( KSHV , HHV8 ) is causally associated with Kaposi’s sarcoma ( KS ) , primary effusion lymphoma ( PEL ) and multicentric Castleman’s disease ( for a review , see reference [1] ) . KSHV maintains a persistent latent infection in B lymphocytes , from which it occasionally reactivates , enters a lytic cycle of replication , and produces infectious virions . Transmission occurs by both sexual and nonsexual contact as well as blood and organ transfusion . Cell-mediated immunity is important for limiting KSHV reactivation and pathogenesis . KSHV infection activates several pattern recognition receptors ( PRRs ) , including cGAS , IFI16 , RIG-I , NLRP1 , and several Toll-like receptors ( TLRs ) which play an important role in promoting the innate immune response [2–7] . KSHV pathogen associated molecular patterns ( PAMPs ) recognized by the innate immune system remain to be fully characterized but are primarily thought to reside on viral glycoproteins and nucleic acids [8] . Much of the work done on the innate immune response to KSHV has used systems in which the cellular response to incoming virus has been examined . These studies have shown that cytoplasmic and endosomal viral nucleic acids may be detected by one or more PRRs [9] and that viral glycoproteins may activate PRRs upon viral entry [9–12] . Several DNA sensors may be important in recognition of viral DNA , including cyclic GMP synthetase ( cGAS ) and IFI16 . IFI16 has recently been shown to act in the nucleus to activate a nuclear inflammasome in response to KSHV and EBV infection [2 , 13 , 14] . cGAS mediated recruitment of STING and IRF3 activation requires association with a ribonucleoprotein complex which is remodeled by foreign DNA [15] . Several components of these cytoplasmic innate immune pathways are involved in the innate immune response to KSHV . TLR9 appears to act as a sensor for incoming KSHV DNA and partly contributes to the activation of IFN-α [10] . Both NLRP1 , a protein component of the inflammasome , and IFI16 may restrict KSHV reactivation , since depletion of NLRP1 or IFI16 results in increased lytic replication [3 , 16] . Similarly , RIG-I , a cytosolic RNA sensor , may be important in limiting KSHV reactivation from latency as KSHV lytic replication was enhanced in RIG-I-/- cells . The importance of these pathways is emphasized by the fact that KSHV counteracts the host response via vIRF-1 , LANA and ORF52 [4 , 6 , 17] . The PAMPs displayed by herpesviruses are predicted to differ considerably depending on the stage and type of their replicative cycles . During latency , herpesviruses exist as chromatinized nuclear episomes . During stringent latency , few , if any , lytic proteins or RNAs are synthesized , and virion DNA is not produced [1] . In contrast , once reactivation occurs , and the virus enters the lytic replicative cycle , abundant amounts of viral mRNAs and non-coding RNAs are produced and newly replicated genomes are produced , encapsidated , and egress from the nucleus prior to tegumentation , final envelopment and transit through the plasma membrane . Thus , while virion DNA is not expected to be exposed to endosomes and cytoplasm in the same context as during primary infection of the cell , there is nevertheless ample of opportunity for virus components such as mRNAs , non-coding RNAs , and viral proteins to be detected by cytoplasmic PRRs and trigger innate immune responses . We therefore wished to extend our study of host cellular factors to cytoplasmic PRRs that could contribute to establishment of an antiviral state and restrict lytic KSHV replication and reactivation from latency . These include human IFIT proteins that have been recently demonstrated to play important roles in the inhibition of several viruses besides herpesviruses [18 , 19] . The IFN induced tetratricopeptide repeat containing proteins ( IFITs ) are among the most highly interferon-induced proteins [20] . They constitute a family of related genes that have been identified in a wide variety of mammals from mouse to man [21] . The human genes , encoded on chromosome 10 , are IFIT1 ( ISG56 ) , IFIT2 ( ISG54 ) , IFIT3 , IFIT5 and IFIT1B [21] . Several of the IFITs have been implicated in an important antiviral response pathway dependent on recognition of foreign RNAs . In mice , Ifit1 , a homolog of human IFIT1B , specifically recognizes uniquely modified viral RNAs that lack 2'O-methylation of their 5' mRNA caps ( cap0-mRNAs ) [22 , 23] . Viruses that replicate in the cytoplasm that have either “snatched” a cap or encode their own 2”O-methyltransferase may thereby evade recognition as non-self [24] . In humans however , IFIT1 protein differs significantly from IFIT1B , and may play a broader role in antiviral function , with different RNA binding specificities [21] . IFIT1 forms a tripartite complex with IFIT2 and IFIT3 and binds to 5’ mRNA caps [25] . In addition to inhibiting replication of viruses that are predicted to have 2’O-methylated caps , IFIT1 inhibits papillomavirus replication by binding its E1A protein [26] . IFIT1 inhibits translation of viral mRNAs by preferentially binding their 5’ cap and preventing association with eukaryotic initiation factors [23] . In addition , IFIT1 has been shown to affect protein translation by interacting with eukaryotic initiation factor eIF3 and may thereby inhibit additional virus families by different mechanisms [18] . It was therefore of interest to determine whether IFIT proteins could inhibit KSHV , whose replication strategy differs considerably from other virus families in which IFITs have been shown to exert antiviral activity . In this study , we examined the effect of IFIT proteins on KSHV replication by depleting IFITs under conditions whereby highly efficient KSHV lytic replication and infectious virion production was enabled in epithelial cells . KSHV lytic gene expression , DNA replication and virion production were enhanced by depletion of IFITs . Further , IFIT expression , which was undetectable during latent infection , was induced during the course of KSHV reactivation and lytic replication . Using deep sequencing of mRNA , we analyzed the effects of IFITs on KSHV and cellular transcript accumulation during lytic KSHV replication . In addition to IFIT effects on the viral transcriptome , we discovered an unexpected positive effect on the expression of other members of the interferon-induced response that is predicted to amplify the antiviral effect of IFIT proteins .
IFIT genes are strongly induced by several viral infections , including the betaherpesvirus hCMV [18 , 19 , 27] . Although several innate immune pathways are induced by KSHV infection or reactivation , it was not known whether IFITs were induced by KSHV lytic replication during reactivation from latency . We therefore examined the status of IFIT protein expression in iSLK/Bac16 cells stably transduced with a doxycycline-inducible viral transactivator , KSHV ORF50/Rta [59] . These Rta-inducible SLK cells ( iSLK ) are stably and latently infected with the Bac16 KSHV strain that expresses hygromycin resistance and GFP , and robust and synchronous reactivation of KSHV from latency is achieved by doxycycline treatment . [30] . Infected cells were 100% GFP positive when maintained under hygromycin selection ( S1 Fig ) . iSLK/Bac16 cells were treated with doxycycline and cells were harvested at serial time points from 0–72 hrs at 12 hr intervals . We measured IFIT1 and IFIT3 protein levels in iSLK cells by Western blotting . Both IFIT1 and IFIT3 were not detectable in uninduced cells but were expressed after KSHV reactivation ( Fig 1 and S2 Fig ) . IFIT1 protein was first detectable at 36hr post induction ( p . i . ) and continued to increase to 72 hr . IFIT3 was first detectable at 12 hr and reached peak expression by 48 hr . ( Fig 1A and 1B ) . In order to confirm that exogenous Rta expression itself did not affect IFIT expression , we also assessed IFIT1 and IFIT3 expression in doxycycline treated or untreated uninfected iSLK cells . Neither IFIT1 and IFIT3 was detectable at 48 hr and 72 hr ( S3 Fig ) . Induction of KSHV Bac16 lytic replication was confirmed by immunoblotting for ORF57 which was expressed by 12hr after induction ( Fig 1C ) . qPCR also showed IFIT1 , IFIT2 and IFIT3 expression peaking by 36hr which was consistent with the results of Western blotting ( Fig 1D ) . These results clearly demonstrate that KSHV lytic replication induces IFIT expression . In order to confirm induction of IFITs as a result of KSHV replication and to determine their cellular location , we examined lytically induced KSHV infected cells by immunofluorescence microscopy . iSLK cells were grown on glass coverslips and treated with 1 μg/ml doxycycline to induce virus lytic replication . Cells were fixed at 48hr , 72hr and 96hr post induction . Immunofluorescence staining for IFIT3 was performed and revealed cytoplasmic expression ( Fig 2A ) . A small percentage of cells were IFIT3 positive before induction , possibly in cells which undergo spontaneous lytic gene expression . However , the percentage of cells expressing IFIT3 increased progressively after lytic induction , and was approximately 45-fold higher by 96 hrs ( Fig 2B ) . These results confirmed the immunoblotting data and demonstrate that IFIT3 is expressed in the cytoplasm . IFIT1 exhibited similar cytoplasmic localization to IFIT3 ( S4 Fig ) . In order to investigate IFIT1 , IFIT2 and IFIT3’s effect on KSHV lytic replication and reactivation from latency , we measured virion production in KSHV infected cells ( iSLK/Bac16 ) after depletion of IFITs . IFIT depletion was carried out by lipid-mediated transfection of iSLK cells with siRNA specific for IFITs ( Fig 3A–3C ) . Cells were transfected with siRNAs 6 hr prior to inducing lytic replication . We first measured IFIT RNA abundance at 48 hr post induction by qPCR ( Fig 3A ) . Approximately 90% depletion of IFIT mRNAs was apparent in cells induced to permit KSHV replication . We next examined expression of IFIT1 and IFIT3 by immunoblotting from samples harvested at 48 hr post induction . Approximately 90% depletion of IFIT1 protein was achieved by 48 hr post-transfection as assessed by densitometry of the Western blot ( Fig 3B ) . There was approximately 73% depletion of IFIT3 ( Fig 3C ) . In order to assess the effect of IFIT depletion on KSHV reactivation and virion production , cells were transfected with either IFIT siRNAs or control siRNA , and 6 hours later KSHV reactivation was induced by addition of doxycycline . Virion-containing supernatant was harvested at 120 hours after induction of lytic replication . Infectious virus production was measured by infection of 293T cells with serial dilutions of virus supernatant followed by flow cytometry of infected cells . Virus titer in the supernatant can thus be accurately quantitated as GFP-transducing units [29] . As shown in Fig 3D and S5 Fig ( IFIT KD and virion titration repeated in a separate experiment ) , IFIT depletion led to a marked increase in virion production ( 25–32 fold ) , compared to control siRNA-transfected cells induced in parallel . There was no microscopically detectable release of infectious virus in the absence of doxycycline from either IFIT depleted cells or in control cells , indicating that Rta is still absolutely required for lytic replication . These data indicate that the IFITs act as a restriction factor for KSHV virus production . IFIT1 , IFIT2 and IFIT3 may form a tripartite complex and cooperate in RNA binding [23 , 25 , 30] . Therefore , we performed virion titration as was done in the previous experiments to examine the effect on virion production of individual depletion of each IFIT . Individual depletion of the three IFITs had a similar effect on virion release , with each IFIT depletion leading to a marked increase in virion production ( about 22–25 fold ) ( Fig 3E ) . The magnitude of this effect is similar to that observed upon depletion of all three IFITs together . These data suggest that each of the three IFITs is important for restriction of KSHV virus production . We next wished to ask at which stage of KSHV lytic replication IFITs might be exerting an inhibitory effect on KSHV virion production . To determine whether the IFIT effect was due to inhibition of KSHV DNA replication , we measured KSHV genome abundance by qPCR on DNA samples from cells that were induced to replicate after depletion or mock depletion of IFITs . The results demonstrated that intracellular KSHV genome copy numbers increased at least 9-fold upon IFIT KD ( Fig 3F ) . However , this increase was not as large as the increases observed in infectious virus titer ( Fig 3D , S5 Fig ) , suggesting that IFITs may affect other steps in the lytic KSHV cycle in addition to DNA replication . IFIT depletion enhanced KSHV virion production ( 25–32 fold ) while KSHV DNA copy number increased only 9-fold , suggesting that IFITs may restrict expression of late genes that are needed for virion formation , egress or infectivity . In order to assess the global effect of IFITs on KSHV lytic gene expression , we performed high-throughput deep sequencing of mRNA from KSHV-infected cells in which IFITs were depleted prior to induction of lytic replication . KSHV-infected iSLK cells were transfected with either control siRNA or IFIT siRNAs as was done in the previous experiments to examine the effect on virion production . Six hours after siRNA transfection , cells were treated with doxycycline to induce KSHV lytic replication , and cells were harvested at 48 hours post induction , RNA was isolated , oligo-dT selected , and processed for deep sequencing . The effects of IFIT KD on lytic cycle transcription were compared to the transcriptional profile of induced cells transfected with control siRNA . A comparison of the transcriptional profiles is presented in Fig 4A . Consistent with its effect on virus production , IFIT KD was associated with broad enhancement of KSHV lytic gene expression . Approximately two thirds of genes demonstrated increased expression: 6 genes increased > 2-fold and 35 genes increased 1 . 2-2-fold . 30 genes did not exhibit increases ( less than 20% change ) while 15 genes decreased 20% -85% ( Fig 4B ) . Of the six genes whose expression was increased more than two-fold , 5 are involved in lytic KSHV DNA replication: ORF56 , the helicase primase; ORF54 , the deoxyuridine triphosphatase; ORF6 , the single-stranded DNA-binding protein; ORF70 , thymidylate synthase; and ORF57 , a post-transcriptional regulator that preferentially enhances mRNA accumulation of several genes involved in DNA replication [31] . The sixth gene ORF47 , encodes glycoprotein L . The preferential enhancement of genes involved in KSHV DNA replication is consistent with the effect of IFIT KD on KSHV DNA replication shown above . However , the increased expression of the majority of KSHV lytic genes suggested that IFITs may have general effects beyond inhibition of specific KSHV genes . In order to determine whether IFITs have a generally inhibitory effect on KSHV replication , we set out to examine the effect of IFIT KD in the BCBL1 KSHV-infected primary effusion lymphoma ( PEL ) cell line . The TREx BCBL1-Rta cell line employed carries a doxycycline-inducible Rta gene , allowing robust KSHV lytic replication upon doxycycline treatment [32] , kind gift of Jae Jung . TRExBCBL1-Rta cells were treated with doxycycline and cells were harvested at serial time points from 0–48 hrs at 12 hr intervals . IFIT1 and IFIT3 protein were not detectable ( S6A Fig , S6B Fig ) by Western blotting although KSHV ORF57 was strongly induced ( S6C Fig ) . IFIT expression in iSLK Bac16 cells was easily detectable under the same conditions . These results indicated that the innate immune response in TRExBCBL1-Rta is different from that in iSLK/Bac16 , consistent with prior reports that IFIT expression is nonfunctional in a large percentage of cancer cell lines and primary cancer cells [33 , 34] . Regardless , we asked whether KD of IFIT mRNAs could affect KSHV gene expression even though IFIT protein expression was undetectable by immunoblotting . Lentiviruses containing shIFIT1 were constructed and tested by infecting induced iSLK/Bac16 . Three clones of shIFIT1s ( 258 , 316 and 581 ) achieved efficient knockdown of IFIT1 in iSLK/Bac16 without cellular toxicity ( S6D Fig ) . These three lentiviral preparations were combined and used to infect TRExBCBL1-Rta at an MOI of 15 . After infection by lentivirus , cells were either treated or mock-treated with doxycycline and harvested at 48hr post induction . Flow sorting was performed to confirm transduction by GFP-expressing lentivirus . Lysates from control or IFIT1 shRNA transduced TREX BCBL1-Rta cells showed no differences in ORF57 ( early ) or K8 . 1 ( late ) lytic protein expression ( S6G Fig and S6H Fig ) . These results indicate that IFITs are not expressed and do not restrict KSHV lytic replication in BCBL1 cells . IFIT depletion enhanced KSHV DNA replication and upregulated viral mRNA expression , contributing to increased virion production . We also wished to determine the effects of IFIT KD on cellular gene expression . There were 99 cellular genes whose transcript abundance decreased by 50% or more upon IFIT KD when KSHV lytic replication was induced . Analysis of this gene set by GO Enrichment Analysis ( http://geneontology . org/page/go-enrichment-analysis ) [35–37] showed high enrichment in genes assigned to the type 1 interferon pathway . 11 of the 99 genes whose transcript abundance decreased on IFIT KD were assigned to this pathway ( Table 1 ) . This represents a 35 fold enrichment over expected ( p value 6 . 94 X 10−14 ) . Among these ISGs , the OAS family ( OAS1 , OAS2 , OAS3 , OASL ) was most significantly enriched . We performed qPCR for these OAS genes to validate and confirm the RNA Seq data ( Fig 5A ) . All OAS genes were significantly down regulated ( p<0 . 0002 ) to less than 17% after IFITs were depleted compared to mock depletion , suggesting that IFITs may affect KSHV replication through the OAS-RNase L pathway . Since several other ISGs belonging to the type 1 interferon pathway were also downregulated by IFIT KD , we performed qPCR to measure the expression of the most upstream regulator , IFN β . Lytic replication of KSHV induced IFN β expression in the control transfection ( Fig 5B ) . IFN β expression was downregulated significantly ( p<0 . 0001 ) to 7% in IFIT KD cells compared to the control ( Fig 5B ) . Therefore , IFITs appear to enhance IFN β production and downstream ISG expression during KSHV lytic replication , and depletion of IFITs results in a blunted type 1 interferon response . The 2’ , 5’-oligoadenylate ( 2-5A ) synthetase ( OAS ) -RNase L system is an interferon-induced antiviral pathway . Induction of OAS proteins by IFN and viral replication leads to synthesis of 2’5’ oligoadenylates which activate RNase L . OASs ( OAS1 , OAS2 , OAS3 ) synthesize 2’-5’ oligoadenylates and activate RNase L leading to degradation of viral and cellular RNAs , thereby restricting viral infections [38] . Activated RNase L preferentially cleaves target RNAs produced by viruses as well as several cellular RNAs at specific sites [39 , 40] . To confirm that the OAS expression that was inhibited by IFIT KD was functionally relevant , we compared RNase L activity in IFIT depleted and mock-depleted cells . Donovan et al . have demonstrated that RNA cleavage by activated RNase L can be quantitatively measured by RtcB-ligase assisted qPCR [40] . In this assay , RtcB ligase , which is capable of ligating 2’ , 3’-cyclic phosphates ( generated by RNase L cleavage ) to 5’OH RNAs , is used to ligate all such ends in the total cellular RNA pool to an RNA-DNA adapter with a 5’OH group . The ligated RNA is then reverse-transcribed and the cDNA is analyzed by qPCR . By using forward primers complementary to specific individual cleavage sites in known RNase L targets , measurement of the cleavage at each such site is achieved , serving as a quantitation of RNase L activity ( Fig 6A ) . In order to serve as an internal control , U6 RNA , which has a naturally occurring cyclic 2’3’ phosphate , was also analyzed and used for normalization . We prepared and purified recombinant RtcB ( S7 Fig ) and then performed a ligation-PCR assay to measure the effect of OAS downregulation on RNase L activity . The RtcB analysis demonstrated that upon IFIT KD , RNase L cleavage decreased significantly at site 36 in tRNA-His , site 27 in non-protein-coding RNA RNY4 and site 30 in non-protein-coding RNA RNY5 . The differences between RNase L directed cleavage in the presence and absence of IFITs were statistically significant as shown in Fig 6B–6D . Consistent with lower cleavage activity of RNase L , total RNA yield increased significantly upon IFITs KD compared to the control ( Fig 6E ) . Thus , IFITs may inhibit KSHV replication through the OAS-RNase L pathway . IFIT1 has been shown to preferentially recognize certain types of capped RNA [12 , 41] . Recently published data indicate that IFIT3 stabilizes IFIT1 and increases its affinity for cap0 mRNAs ( See Fig 7 for a diagram of the various types of RNA 5’ caps ) [41–43] . However , IFIT1 is also capable of binding cap1 mRNAs ( Fig 7 ) , albeit at lower affinities [41] . It was therefore possible that IFIT1 might recognize and inhibit translation or stability of KSHV mRNAs directly in addition to indirect effects mediated via other ISGs as shown with OASs . We performed immunoprecipitation experiments to determine if IFIT1 bound viral mRNA specifically or preferentially . iSLK cells were treated with doxycycline to induce KSHV replication , lysates were harvested at 48 hr post induction and immunoprecipitated with IFIT1 and IFIT3 antibodies . Immunoprecipitated RNAs were isolated and measured by qPCR . As shown in Fig 8A , viral RNAs were enriched 5~11-fold in the immunoprecipitation using IFIT1 and IFIT3 antibodies compared to control IPs . The cellular GAPDH mRNA was enriched 4 . 8-fold . It should be noted that the degree of binding of individual RNAs to IFITs was not related to their overall abundance . For example , ORF6 , which was present at 330 FKPM was enriched similarly to K4 , which was highly abundant at 50 , 000 FKPM ( Fig 8B ) . Uncapped MT-ADP6 RNA , a cellular mitochondrial transcript [44] , as well as uncapped snoRNAs U15 [45] and U16 [46 , 47] were enriched only 1 . 7-fold ( Fig 8A ) . U6 RNA , which has a gamma-monomethyl phosphate cap [48] , was enriched in only 2 . 2-fold in the immunoprecipitates ( Fig 8A ) . U1 , U2 and U5 have a trimethylation cap [49] , and they were similarly enriched less than 1 . 8-fold ( Fig 8A ) . All these snoRNA are expressed at more than 10 , 000 copies per cell which is a much higher abundance compared to cellular genes [50 , 51] . Therefore , IFIT1 and IFIT3 , while preferentially recognizing cap0 structures , are also capable of binding to both viral and cellular capped mRNA ( cap1 or cap2 ) . However , their ability to bind uncapped , monomethyl capped , or trimethyl capped RNAs appears to be extremely limited .
In this study we examined the effects of cellular IFIT1 , IFIT2 and IFIT3 on KSHV lytic replication . Heretofore , IFIT proteins have been primarily implicated in antiviral responses against RNA viruses [18 , 19] . IFIT1 , IFIT2 and IFIT3 form multimeric complexes that initially were shown to bind 5’ tri-phosphate RNAs [25] . Such RNAs are produced by several negative strand RNA viruses such as Rift Valley virus , vesicular stomatitis virus and influenza virus against which IFITs exhibit antiviral activity [19] . However , IFITs were subsequently shown to also preferentially bind cap0 mRNAs which lack 2’O-methylation at the first and second transcribed ribonucleotides ( as seen in cap1 and cap2 mRNAs , Fig 7 ) . Many viruses that replicate in the cytoplasm , including flaviviridae , poxviridae and coronaviridae , have evolved enzymes to independently perform 2’O-methylation of their mRNAs . Mutants of these viruses that have lost 2’O methyltransferase activity exhibit increased susceptibility to IFIT dependent immune responses , suggesting that IFITs allow discrimination between self and non-self RNAs [12 , 52] . Although IFIT induction by RNA viruses is common , herpesviruses may also induce IFIT gene expression , by direct or indirect mechanisms . Indeed , IFIT2 and IFIT3 were identified as hCMV induced genes ( cigs ) over twenty years ago by the use of differential display [27] . HSV infection also leads to IFIT induction , albeit less strongly than CMV infection [27] . It was therefore of interest to determine whether a gammaherpesvirus such as KSHV could induce IFIT gene expression . KSHV , similar to other herpesviruses which undergo lytic replication in the nucleus , are presumed to have cap structures similar if not identical to host mRNAs [24] . Nevertheless , IFIT1 has also been shown to exert inhibitory effects on translation independent of mRNA sequestration by interacting directly with eIF3 [18] , also raising the question of whether IFITs could establish an antiviral state that would inhibit KSHV virion production . We first established that KSHV reactivation and lytic replication results in IFIT induction . While all three IFIT mRNAs were measurably induced upon KSHV lytic replication , we were only able to detect increased expression of IFT1 and IFIT3 proteins . While the IFIT2 antibodies we employed were able to detect exogenously overexpressed IFIT2 , we did not detect IFIT2 protein expression by either immunoblotting or immunofluorescence microscopy . This may be due to minimal IFIT2 protein induction as a consequence of KSHV replication but our finding that IFIT2 depletion enhanced KSHV production suggests that functional IFIT2 is present and the failure to detect IFIT2 protein is likely due to inadequately sensitive IFIT2 antibodies . Nevertheless , it is clear that although KSHV lytic reactivation occurs from the nucleus , PAMP exposure sufficient to engage PRRs and induce ISG expression occurs . Future studies to examine the nature of the non-self signatures , whether DNA , RNA or protein , that evoke the innate immune response to reactivating herpesviruses , and whether the PRRs that recognize them are nuclear and/or cytoplasmic , will be very informative . Both IFIT1 and IFIT3 proteins that we detected by IF studies were localized to the cytoplasm in KSHV infected cells . Although virtually all studies have focused on the interaction of IFITs with cytoplasmic RNAs , it has been suggested that IFIT1 may also have transcriptional activating functions [53] . We examined the potential role of IFITs as KSHV inhibitory proteins by knocking down IFITs and then inducing KSHV lytic replication . As expected , IFIT expression was minimal in the absence of KSHV replication , and expression of IFITs was also severely curtailed after siRNA treatment . Blocking IFIT production resulted in a 25–30 fold increase in infectious KSHV virion production . Consistent with these findings , IFIT KD also led to increased lytic KSHV mRNA accumulation . Importantly , the increase in mRNA abundance , while not completely equal amongst all KSHV mRNA , was nevertheless broad , with over 65% of lytic mRNAs increasing in abundance . However , these findings pose a difficulty in interpretation due to the fact that herpesvirus late gene transcription , including that of KSHV , is dependent on DNA replication [54] . Since the most highly IFIT restricted KSHV mRNAs encode proteins that are either essential or important for KSHV lytic DNA replication , the broad inhibitory effect on KSHV mRNAs may be partly indirect , with late gene repression by IFITs due to the inhibitory effect of IFITs on viral DNA replication . We confirmed that IFITs do have an inhibitory effect on DNA replication by directly measuring KSHV DNA abundance in the presence and absence of IFITs . KSHV is thought to most likely enter the human host by oral epithelial cell infection [55] and has been demonstrated to infect a variety of human epithelial cells including oral keratinocytes as well as epithelial cell lines [56 , 57] . These experiments were carried out in the iSLK/Bac16 cell line , an epithelial cell line that supports efficient KSHV lytic replication and has served as a model for KSHV infection and reactivation from latency [58] . We also examined the potential role of IFITs in KSHV reactivation from latent infection in a PEL cell line , BCBL1 [59] . However , cells from BCBL1 did not express detectable IFITs and it was therefore not possible to determine whether IFITs may play a role in restricting KSHV replication in B lymphocytes . As loss of ISG expression is not uncommon in many human tumors , these findings do not rule out the possibility of IFITs playing a physiological role in regulating KSHV replication in B lymphocytes in vivo . These data also raised the possibility that IFITs exerted at least some of their antiviral function by other indirect mechanisms that did not depend on specific targeting of KSHV mRNAs , especially since there are no known differences in cap structures between herpesvirus mRNAs and host cellular mRNAs [24] . Our analysis of the cellular transcriptome suggested an effect of IFITs on the type I interferon pathway , as several ISGs , including several known to be important for establishing an antiviral state , decreased upon IFIT KD in comparison to infected cells in which IFITs were not depleted . These findings do not differentiate between transcriptional or post-transcriptional effects of IFITs in enhancing ISG expression . However , the fact that interferon mRNA levels were decreased in the absence of IFITs suggests that the simplest model for a positive feedback loop maintained by IFITs might be an effect on type I IFN transcription or RNA stability . A recent report implicated IFIT1 in nuclear regulation of transcription , both acting to negatively regulate the inflammatory response as well as enhancing IFN β1 transcription [53] . The IFN β 1 response to pathogens was also blunted in IFIT1-depleted cells in this study . Our data support a model in which IFITs maintain an antiviral state by promoting enhanced IFN and ISG production . OAS proteins are established components of the innate immune response to viruses . OASs synthesize 2’-5’ oligoadenylates and activate RNase L leading to degradation of viral and cellular RNAs and thereby block viral infections as well as amplification of IFN α/β by RNase L-generated small RNAs [38] . Because OAS mRNAs were the most highly downregulated upon IFIT KD , we examined whether this correlated with a functional decrease in potential antiviral activity . By using the RtcB-ligase assay , which measures the production of cyclic 3’ phosphate moieties at specific RNase L cleavage sites , we were able to determine that ISG KD does lead to a functional decrease in RNase L activity . The observed decrease in RNase L activity is consistent with the generalized decrease in KSHV lytic mRNA abundance and in total cellular RNA . We also examined the ability of IFITs to bind several types of capped RNA . By direct immunoprecipitation of IFIT proteins , we found that cap1/2 host cell mRNAs and KSHV mRNAs were widely represented in IFIT immunoprecipitates . Despite the fact that IFIT complexes were originally isolated by using triphosphate uncapped RNAs as bait , we found very little representation of naturally uncapped RNAs such as certain mitochondrial or snoRNAs [44–47 , 60] in IFIT immunoprecipitates . Whether interactions of IFITs with cap1/2 mRNAs exert negative effects on their stability or translation or if such interactions could even have positive effects on target mRNA remains to be determined . Although the enhanced affinity of IFIT1 complexes for cap0 mRNAs has been adduced as evidence of a PAMP recognition by IFITs that allows them to distinguish between self and non-self , the fact remains that the majority of viral mRNAs are 2’-O-methylated . In addition , some viruses , such as parainfluenza virus , whose mRNAs are 2’-O-methylated , are nevertheless inhibited by IFITs [61] . Our findings demonstrating that IFITs exert antiviral effects on KSHV , a herpesvirus , which has neither genomic RNAs nor atypically capped mRNAs , provide further evidence that IFITs may have antiviral effects beyond direct sequestration of mRNA . Although IFIT complexes do not appear to be highly enriched for specific transcripts , we cannot rule out the possibility that sequestration by IFITs may have varying effects on different targets , affording a degree of specificity . Variation in the effects of IFIT binding to different targets could arise from intrinsic differences in translatability or stability of individual target mRNAs , especially lytic herpes virus transcripts , which are primarily intronless [62 , 63] . In summary we have shown that IFITs exert an antiviral effect on a herpesvirus which does not express any of the putative pathogen associated RNA signatures expressed by RNA viruses . In addition , in infected cells , IFITs do associate with canonically capped viral and cellular mRNAs that are not known to be possess cap0 structures . Further , IFIT depletion led to decreases in IFNβ as well as several other antiviral effectors of the interferon pathway , suggesting that IFITs may possess broad antiviral effects by virtue of their ability to amplify the interferon response . Finally , by virtue of the IFITs ability to interact with canonically capped cellular and viral transcripts , they may also affect both host cell and viral gene expression by direct effects on mRNA .
293T cells ( kind gift of Lori Frappier , University of Toronto ) were grown at 37°C in Dulbecco’s modified Eagle’s medium ( DMEM ) supplemented with 10% fetal bovine serum ( FBS ) and glutamine . iSLK cells [58] ( gift of Don Ganem , UCSF ) were maintained in DMEM containing 10% charcoal stripped FBS ( Sigma ) and 1% glutamine with 250μg/ml G-418 and 1μg/ml puromycin . iSLK cells were infected with WT KSHV derived from bacmid BAC16 , expressing eGFP and hygromycin resistance [28] . Bac16 KSHV infected iSLK cells were maintained in 1 . 2 mg/ml hygromycin , 250μg/ml G-418 and 1μg/ml puromycin . TRExBCBL1-Rta ( kindly provided by Prof . J . Jung ) were cultured in RPMI 1640 10% Tet System approved FBS ( Clontech ) and 1% glutamine with 50μg/ml hygromycin . IFIT1 ( L-019616-00-0005 ) , IFIT2 ( L-012582-02-0005 ) , IFIT3 ( L-017691-00-0005 ) and negative control On-target plus Smart Pool siRNAs ( D-001810-03 ) were purchased from Thermo Scientific . Each siRNA was transfected into iSLK cells using Lipofectamine RNAiMAX ( Invitrogen ) according to the manufacturer’s protocol . For KD of all three IFITs , each siRNA was used at 10 nM final concentration and NC siRNA was used at 30 nM . For individual IFIT KD , each siRNA or NC siRNA was used at 10 nM . RT-qPCR or immunoblotting was performed to verify knockdown of the relevant protein . IFIT1 GIPZ shRNA clones ( RHS4531-EG3434 ) were purchased from Dharmacon . Lentiviruses were prepared by transient transfection of 293T cells with a three-plasmid system ( a GIPZ plasmid expressing shRNA against cellular IFIT1; pMD2 . G [envelope plasmid expressing vesicular stomatitis virus glycoprotein]; and psPAX2 [packaging plasmid] ) . Viral supernatant was harvested at 48hr post transfection with 0 . 45μm syringe filter . Lentivirus were concentrated at 10000g in 10% sucrose buffer for 3 . 5hr as described [64] and immediately used to infect iSLK/Bac16 or TREx BCBL1-Rta cells . iSLK/Bac16 cells were infected with each lentivirus at an MOI of 15 and then induced to permit KSHV lytic replication . Cells were harvested 48hr post-induction and Western blotting of IFIT1 was performed . Lentiviruses containing three independent shRNAs were used to infect TRExBCBL1-Rta cells at an MOI of 15 . For shRNA knockdown experiments , TRExBCBL1-Rta cells were infected twice with concentrated lentiviruses within two days . 2 days after the second infection , cells were seeded at 1 × 106/ml and induced with 1μg/ml doxycycline . Cells were harvested or sorted by flow cytometry at 48 post induction . Total cellular RNA was isolated from washed cell pellets using Qiazol and Qiagen miRNeasy columns according to the manufacturer’s protocols . mRNA was purified from 6 μg cellular RNA using Qiagen Oligotex mRNA Minikit ( Qiagen ) . cDNA libraries were prepared using the ABI high Capacity cDNA Reverse Transcription Kit with RNase inhibitor ( Applied Biosystems ) . Real-time quantitative PCR ( qPCR ) was performed with SYBR green PCR Master Mix ( Applied Biosystems ) according to the manufacturer’s protocol . Each sample was analyzed in triplicate with gene specific primers and β-actin was used as the endogenous control . The gene-specific primers were as follows: IFIT1 Q1F: 5’-ggaatacacaacctactagcc-3’; IFIT1 Q1R: 5’-ccaggtcaccagactcctca-3’; IFIT2 Q1F: 5’-gggaaactatgcctgggtc-3’; IFIT2 Q1R: 5’-ccttcgctctttcattttggtttc-3’; IFIT3 Q1F: 5’-tgaggaagggtggacacaactgaa-3’; IFIT3 Q1R: 5’-aggagaattctgggttgttgggct-3’ OAS1 Q1F: 5’-gcgccccaccaagctcaaga-3’ OAS1 Q1R: 5’-gctccctcgctcccaagcat-3’ OAS2 Q1F: acccgaacagttccccctggt-3’ OAS2 Q1R: 5’-acaagggtaccatcggagttgcc-3’ OAS3 Q1F: 5’-tgctgccagcctttgacgcc-3’ OAS3 Q1R: 5’-tcgcccgcattgctgtagctg-3’ OASL Q1F: 5’-gcggagcccatcacggtcac-3’ OASL Q1R: 5’-agcaccaccgcaggccttga-3’ ORF6 Q1F: 5’-ctgccataggagggatgtttg-3’ ORF6 Q1R: 5’-ccatgagcattgctctggct-3’ ORF47 Q1F: 5’-agcctctaccctgccgttgttct-3’; ORF47 Q1R 5’-acgaccgcgactaaaaatgacct-3’; ORF54 Q1F: 5’-gtagccgcatatgccagattgtg-3’ ORF54 Q1R: 5’-ttttgaagcccttgaggatgtgtc-3’ ORF56 Q1F: 5’-cacagattcccgtcaatacaaa-3’; ORF56 Q1R , 5’-gtatcttcagtaggcggcagag-3’; ORF57 Q1-5: 5’-gcagaacaacacggggcgga-3’ ORF57 Q2-3: 5’-gtcgtcgaagcgggggctct-3’ ORF70 Q1F: 5’-gactatacaggccaggggtttgac-3’ ORF70 Q1R: 5’-ggcgggttccacgcacac-3’ K4 Q1F: 5’-gtttgcaatctggggacacg-3’ K4 Q1R: 5’-tggtaaccgagacagcacttg-3’ β-actin Q1F: 5’-tcaagatcattgctcctcctgag-3’ β-actin Q1R: 5’-acatctgctggaaggtggaca-3’ High-throughput deep sequencing of RNA was performed as previously described [29] with some modifications . Briefly , iSLK cells were transfected with a mixture of 10 nM final concentration of each IFIT siRNA or negative control siRNA ( 30nM final concentration ) and were treated with 1 μg/ml doxycycline after 6 hrs . Cells were harvested at 48 hr post induction for RNA isolation . RNA samples from iSLK cells were prepared using Qiagen miRNeasy kits according to the manufacturer’s protocols . cDNA libraries were prepared from poly ( A ) RNA and were sequenced on a HiSeq2000 instrument with 50 cycle single end reads . Sequenced reads obtained from Bac16 KSHV-infected iSLK cells were aligned to the KSHV Bac16 ( GenBank accession no . GQ994935 . 1 ) and Hg19 . Differential gene expression was measured using USeq’s Defined Region Differential Seq application as described previously [29] . Protein samples were analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis ( SDS-PAGE ) and immunoblotted with rabbit polyclonal anti-IFIT1 antibody ( PA3-848 , ThermoScience ) , rabbit polyclonal anti-ORF57 antibody [65] , mouse anti-K8 . 1 monoclonal antibody ( kind gift from Bala Chandran ) , anti-α tubulin polyclonal antibody ( Sigma ) or mouse anti-actin monoclonal antibody ( MA5-15739 , Thermo Science ) and horseradish peroxidase-conjugated anti-rabbit secondary antibody ( NA934V , GE Healthcare ) or anti-mouse secondary antibody ( NA931V , GE Healthcare ) , followed by visualization with a Prometheus ProSignal Peroxide Kit ( Genesee Scientific ) . Image capture and densitometry quantitation of IFIT1 and IFIT3 proteins was performed with a BioRad GelDoc system and Image Lab software ( V5 . 1 ) . Final depletion was calculated as percentage of IFIT KD from NC control and normalized to tubulin . iSLK cells were grown on glass coverslips plated at 120 , 000 cells per well in six-well dishes and treated with 1 μg/ml doxycycline to induce virus lytic replication . Cells were fixed at 48hr , 72hr and 96hr post induction , washed with 1x PBS , fixed and permeabilized with PBS containing 4% paraformaldehyde and 0 . 2% Triton X-100 for 15–20 minutes at room temperature , and then washed two times with 1x PBS followed by incubation with blocking buffer ( 20% goat serum in PBS ) for 30 mins at room temperature . Cells were incubated with anti-IFIT3 polyclonal antibody ( Thermo Science ) or anti-IFIT1 polyclonal antibody ( Thermo Science ) at a dilution of 1:500 , at 37°C for 1 hour . The slides were washed three times with 1x PBS and incubated with Alexa Fluor 594 goat anti-rabbit IgG ( A11072 , Invitrogen ) for 1 hour at 37°C ( in the dark ) . Nuclear staining was performed with 4' , 6- diamidino-2-phenylindole ( DAPI ) ( Invitrogen ) . Images were collected and analyzed with a Zeiss Imager M2 microscope system . To determine cells staining positive for IFIT3 , more than 6 fields in which each field included more than 2000 cells ( totally at least 12000 cells ) were counted with a 20x objective . Statistical testing for comparison of proportions and p values was performed using MedCalc software , which uses the "N-1" Chi-squared test ( MedCalc software , Ostend Belgium ) To induce KSHV lytic gene expression or virus replication , iSLK cells were treated with 1 μg/ml doxycycline . Where KD of IFITs was performed , siRNA transfections were performed 6 hours prior to induction . Cells were harvested at 48 hr post induction for RNA preparation . At 48 hrs post-induction , cell viability was greater than 94% by vital dye staining . For virus production , supernatants of the cells were harvested 5 days post induction , cleared by centrifugation twice , and filtered through a 0 . 80 μm pore size cellulose acetate filter . Serial dilutions of supernatants were used to infect 293T cells . 48 hours after infection , flow cytometry was performed to measure the number of GFP positive cells , each representing a cell infected by a GFP expressing KSHV virion [29] . Each infection was done in triplicate and each infected cell sample was assayed by flow cytometry in technical triplicates . Based on the dilution factor , infectious virus titers in the iSLK cell supernatant were calculated . Pellets of the cells from which supernatant was harvested were processed for DNA isolation using Qiagen DNeasy Blood and Tissue kit . 50 ng of each DNA were used for qPCR using primers specific for ORF59 ( see above ) and SYBR green PCR MasterMix ( ABI ) . RtcB expression plasmid ( kind gift from Dr . Alexei V . Korennykh ) , contains the RtcB gene from E . coli cloned into pGEX-6P and expressed with an N-terminal GST-tag . Protein isolation and purification were performed as previously described [40] with some modifications . Briefly , E . coli BL21-CodonPlus ( DE3 ) carrying the RtcB expression plasmid was grown at 37°C in LB medium containing 50 μg/ml ampicillin until the A600 reached 0 . 6 to 0 . 7 . The culture was chilled down to 20°C and protein expression was induced by IPTG ( 0 . 25mM final concentration ) . Incubation was continued at 20°C for 16 hr with constant shaking . Cells were harvested by centrifugation and stored at −80°C . All subsequent procedures were performed at 4°C . The cell pellet was suspended in 50 ml buffer A ( 20 mM HEPES pH 7 . 5 , 300 mM NaCl , 10% glycerol , 2 mM DTT , 0 . 1 mM EDTA , 1% Triton X-100 ) with proteinase inhibitor ( 0 . 3μg/ml aprotinin , 0 . 5μg/ml leupeptin and 0 . 7μg/ml pepstatin A ) , 10μg/ml DNase and 10μg/ml lysozyme . After mixing for 30 min , cells were sonicated and insoluble material was removed by centrifugation . The clarified lysate was added to a column with 10ml glutathione resin washed with buffer A . The column was rotated for 1h at 4°C to bind GST-RtcB to the resin . Washing was performed with buffer A and the resin was resuspended in low salt buffer A ( 100mM NaCl without Triton X-100 ) . PreScission Protease ( GE Healthcare ) was added and the column was gently rotated overnight at 4°C . The cleaved RtcB was then eluted from the column in 10ml buffer A . Ion exchange ( MonoQ ) purification ( S7 Fig ) followed by S200 gel filtration ( S7 Fig ) was performed with unsalted buffer ( 20 mM HEPES pH 7 . 5 ) , high salt buffer ( 20 mM HEPES pH 7 . 5 , 1M NaCl ) and buffer B ( buffer A without 1% Triton X-100 ) . Purified RtcB was eluted and diluted to 100μM in buffer B with 0 . 5% Triton X-100 , aliquoted and stored at -80°C . We measured several specific cleavage sites generated by RNase L to analyze its activity as described [40] . Briefly , RtcB ligase and RtcB ligation adaptor ( 5’-OH-GAUCGUCGGACTGTAGAACTCTGAAC-3’ ) were added to cellular RNA ( 500ng ) to ligate 2′ , 3′-cyclic phosphate containing RNAs to the adapter . The underlined bases in the ligation primer were RNA and the remainder were DNA . EDTA-quenched ligation reaction ( 1μL ) was used as a template for reverse transcription with Multiscribe reverse transcriptase ( ThermoFisher ) and RT primer ( 5′-TCCCTATCAGTGATAGAGAGTTCA GAGTTCTACAGTCCG-3′ ) . SYBR-green based qPCR was conducted using a universal rev qPCR primer ( 5’-TCCCTATCAGTGATAGAGAG-3’ ) and cleavage site-specific forward primers designed for each RNA target: tRNA His-36 ( 5’-GTTAGTACTCTGCGTTGTGGA-3’ ) , RNY4-27 ( 5’-GATGGTAGTGGGTTATCAGAT-3’ ) and RNY5-30 ( 5’-GTGTTGTGGGTTATTGTTAGA-3’ ) . U6 , which has a naturally occurring 2′ , 3′-cyclic phosphate and an RNase L independent cleavage site , was used as endogenous control . Primers for the U6 site were U6 Q1F: 5’-GCTTCGGCAGCACATATACTA-3’ and U6 Q1R: 5’-CGAATTTGCGTGTCATCCTTG-3’ , qPCR was carried out for 60 cycles using 62°C annealing/extension for 1 min . iSLK/Bac16 cells were treated with 1 μg/ml doxycycline to induce KSHV replication . Cells were harvested at 48 hr post induction for immunoprecipitation . Cells were lysed by freeze-thawing in hypotonic buffer containing 20mM HEPES , pH7 . 3 , 2mM MgCl2 , 10% glycerol , 0 . 2mM EGTA , 1mM DTT , 1x protease inhibitor cocktail ( Sigma ) and RNasin ( Promega ) . All subsequent steps were performed at 4°C . Lysates were clarified by centrifugation and precleared with normal rabbit IgG ( Bethyl ) and protein A-agarose beads , followed by immunoprecipitation with anti-IFIT1 plus anti-IFIT3 Rabbit polyclonal antibody ( ThermoScience ) , or normal rabbit IgG overnight , followed by incubation with protein A-agarose beads . The beads were washed four times in IP washing solution ( 500 mM NaCl , 0 . 25% NP-40 , 0 . 25% Triton X-100 , 0 . 5% CHAPS ) . Coimmunoprecipitated RNA was isolated from the immunoprecipitates using Qiazol with additional glycogen and Qiagen miRNeasy columns according to the manufacturer’s protocols , with an on-column DNase treatment ( Qiagen ) . Immunoprecipitated viral and cellular gene mRNA was quantitated by Real-time Quantitative PCR ( qPCR ) with SYBR green PCR Master Mix ( Applied Biosystems ) according to the manufacturer’s protocol . Each sample was analyzed in triplicate with gene specific primers . The gene-specific primers were as follows: GAPDH Q1F , 5’-agggtcatcatctctgccccctc-3’; GAPDH Q1R , 5’-tgtggtcatgagtccttccacgat-3’ MT-ATP6 Q1F: 5’-gtatgagcgggcgcagtgatt-3’ MT-ATP6 Q1R: 5’-atggggataaggggtgtaggtgtg-3’ U1 Q1F: 5’-ccatgatcacgaaggtggttt-3’ U1 Q1R: 5’-atgcagtcgagtttcccacat-3’ U2 Q1F: 5’-ctcggccttttggctaagat-3’ U2 Q1R: 5’-cgttcctggaggtactgcaa-3’ U5 Q1F: 5’-ctctggtttctcttcagatcgc-3’ U5 Q1R: 5’-ccaaggcaaggctcaaaaaat-3’ U6 Q1F: 5’-gcttcggcagcacatatactaaaattgga-3’ U6 Q1R: 5’-ataggaacgcttcacgaatttgcg-3’ U15 Q1F: 5’-ggtcacgtcctgctcttggtc-3’ U15 Q1R: 5’-atgcctctaaatcgatcaataaat-3’ U16 Q1F: 5’-atgatgtcgtaatttgcgtctt-3’ U16 Q1R: 5’-ctcagtaagaattttcgtcaacc-3’ KSHV gene specific primers are listed above .
|
The innate immune response to infections is triggered by recognition of pathogens as foreign or non-self . Recognition of invading pathogens is carried out by various sensors or pattern recognition receptors ( PRRs ) that detect conserved features of pathogens including lipids , nucleic acids and proteins . PRR activation triggers pathways that ultimately lead to pathogen destruction , including the interferon response . Interferons , in turn induce many interferon-stimulated genes , which inhibit or destroy a wide variety of pathogens , including viruses . IFITs are a family of interferon induced proteins that are thought to recognize RNAs and have antiviral effects primarily on RNA viruses . Kaposi’s sarcoma-associated herpesvirus ( KSHV ) , a DNA virus , is associated with Kaposi’s sarcoma and lymphoid malignancies . In this study we show that IFITs restrict replication of KSHV and does so not only by inhibiting KSHV mRNA abundance but also by enhancing other effectors of the interferon response . This study reveals that the innate immune response can control not only invading viruses but ones that reactivate from latency , that IFITs can inhibit herpesvirus replication and that IFITs may amplify the innate immune response by a feed-forward mechanism .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"antimicrobials",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"gene",
"regulation",
"pathogens",
"drugs",
"messenger",
"rna",
"microbiology",
"antimalarials",
"viruses",
"small",
"nuclear",
"rna",
"dna",
"replication",
"antibiotics",
"dna",
"viruses",
"pharmacology",
"dna",
"herpesviruses",
"doxycycline",
"small",
"interfering",
"rnas",
"proteins",
"medical",
"microbiology",
"gene",
"expression",
"microbial",
"pathogens",
"kaposi's",
"sarcoma-associated",
"herpesvirus",
"viral",
"replication",
"biochemistry",
"rna",
"small",
"nucleolar",
"rna",
"nucleic",
"acids",
"virology",
"viral",
"pathogens",
"genetics",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"interferons",
"non-coding",
"rna",
"organisms"
] |
2019
|
Human IFIT proteins inhibit lytic replication of KSHV: A new feed-forward loop in the innate immune system
|
In learning from trial and error , animals need to relate behavioral decisions to environmental reinforcement even though it may be difficult to assign credit to a particular decision when outcomes are uncertain or subject to delays . When considering the biophysical basis of learning , the credit-assignment problem is compounded because the behavioral decisions themselves result from the spatio-temporal aggregation of many synaptic releases . We present a model of plasticity induction for reinforcement learning in a population of leaky integrate and fire neurons which is based on a cascade of synaptic memory traces . Each synaptic cascade correlates presynaptic input first with postsynaptic events , next with the behavioral decisions and finally with external reinforcement . For operant conditioning , learning succeeds even when reinforcement is delivered with a delay so large that temporal contiguity between decision and pertinent reward is lost due to intervening decisions which are themselves subject to delayed reinforcement . This shows that the model provides a viable mechanism for temporal credit assignment . Further , learning speeds up with increasing population size , so the plasticity cascade simultaneously addresses the spatial problem of assigning credit to synapses in different population neurons . Simulations on other tasks , such as sequential decision making , serve to contrast the performance of the proposed scheme to that of temporal difference-based learning . We argue that , due to their comparative robustness , synaptic plasticity cascades are attractive basic models of reinforcement learning in the brain .
Learning from reinforcement involves widely differing spatial and temporal scales both within the behavioral decision making process itself as well as when relating decisions to outcomes . Since they are adaptive , synapses may be viewed as the elementary decision making entities in the brain . But the presynaptic input of any single synapse will contain only very limited information about the task and , further , the millisecond duration of a synaptic release is much shorter than behaviorally relevant time scales . The behavioral decision results from a spatio-temporal aggregation of synaptic releases which is highly non-linear due to e . g . thresholding in the generation of action potentials . Hence the relationship between any single synaptic release and the behavioral decision is not only tenuous but also non-linear . In relating behavioral decisions to rewarding or unrewarding outcomes , problems arise which are analogous to the ones encountered when relating synaptic releases to decisions . In the “spatial” domain: The state of the world is only partially observable , and hence , what appears to be one and the same decision may sometimes be rewarded and sometimes not . Also , in social interactions , reward may depend on the decisions of other players . In the temporal domain: Whether a decision was appropriate or not may not be immediately obvious and reward may even change with time . Proverbially , short term gain may lead to long term pain ( and vice versa ) . Hence the spatio-temporal credit assignment problem arises: How can a synapse adapt given that reward delivery is delayed and also depends on the releases of many other synapses as well as on external factors ? As one basic mechanism for addressing the temporal problem , theories of reinforcement learning use the eligibility trace , a quantity , decaying exponentially in time , which memorizes the elementary decision up to the time when information about reward becomes available to trigger the persistent adaptive change [1] . Here we point out that a cascade of such synaptic memory traces can in fact provide an integrated solution to the spatio-temporal credit assignment problem by remodulating the presynaptic signal in view of information arising at different stages of the behavioral decision making . Evidence for synaptic eligibility traces comes from experiments on spike timing dependent plasticity ( STDP ) where a synaptic release leads to longterm potentiation ( LTP ) if the neuron emits an action potential shortly thereafter [2] , [3] . Importantly , the length of the LTP-induction time window ( some ) is on the order of the membrane time constant ( ) , i . e . it reflects the time during which the synaptic release has influence on somatic action potential generation . The release itself lasts only for some , so this form of LTP is most easily accounted for by assuming a local synaptic quantity providing , just like an eligibility trace , a memory of the release which decays with time constant . When an action potential is generated , is read-out to determine a quantity which , in the simplest interpretation of the STDP findings , gives the change ( ) of the synaptic strength [4] . Simply equating with , however , may be hasty because many repeated pre/post pairings are required in the STDP-protocol to induce a noticeable change . So it seems more reasonable to view as a second synaptic eligibility trace , keeping a running record of recent pre/post pairings to modulate synaptic strength , perhaps even in a non-linear manner . As has been widely noted [5]–[11] , one can connect the STDP-findings with reinforcement learning by assuming that the transcription of the second eligibility trace into the synaptic change is modulated by neurotransmitters like dopamine which provide feedback about external reward ( Fig . 1A ) . Such plasticity rules address the spatial credit assignment problem for synapses sharing a postsynaptic neuron since captures the relevant correlations between a given synaptic release and the releases of other synapses when they contribute to postsynaptic firing in the neuron . But does not take into account the interaction in decision making between synapses which have different postsynaptic neurons . For temporal credit assignment , the memory length of must correspond to the delay between a synaptic release and the delivery of pertinent reward feedback . This delay consists of the time needed to reach a behavioral decision and the time for this decision to be rewarded . A value on the order of s seems reasonable for , but can easily be much longer , as in a game where multiple decisions are needed to reach a rewarding state . In this case , simply averages pre/post pairing over multiple decisions even if the firing of the particular neuron was important only for some of the decisions . Here we propose extending the eligibility trace cascade by a further trace which takes into account the behavioral decision making process ( Fig . 1B ) . Now the time constant of is simply , since only needs to capture pre/post pairings upto the time when a decision is reached . The decision triggers a transcription of into which is modulated by a feedback signal from the decision making circuitry and a signal derived from the firings of the postsynaptic neuron during the decision period . So while only captures the pre/post correlations , additionally captures the post/decision correlations . The time constant of is , and when reward feedback does become available , the reward together with determines the synaptic change . In Text S1 we show that , for a population of spiking neurons feeding into a decision making circuitry ( Fig . 1C ) , such a synaptic cascade can be mathematically derived by calculating the gradient of the expected reward . The resulting gradient ascent rule , however , has a few biologically undesirable aspects . For instance , it requires that averages pre/post correlations over each decision period . Synapses , however , are unlikely to know when decision periods start and end . For biological realism , we present a modified rule in the main text , where e . g . the averaging over the decision period is replaced by low pass filtering . Learning in a population of spiking neurons using this synaptic plasticity rule is illustrated by simulation results . These show that learning speeds up with increasing population size and that learning speed degrades gracefully when the delay period between decision and reinforcement is increased . In particular , perfect performance is approached even when in the delay period the network has to make further decisions which themselves give rise to delayed reinforcement . Eligibility traces memorize information about the decision making upto the time when reinforcement becomes available . In contrast , temporal difference ( TD ) learning , the other basic approach for temporal credit assignment in reinforcement learning , back-propagates reward to the time of the decision . For this , TD-learning estimates the value of states , or state-decision pairs , where , in the simplest case , a state corresponds to a stimulus . The value itself is the ( discounted ) expected future reward when being in the state , or when making a particular decision in the state . The value can then serve as an immediately available surrogate for the delayed reward signal . During Pavlovian learning , a backward shift in time is observed for the appetitive reaction from the delayed unconditioned stimulus to the conditioned stimulus , and the shift is found as well in the activity of midbrain dopaminergic neurons . The backward shift also occurs in the value estimation error computed by a TD-algorithm modeling the conditioning task , when a state of the algorithm corresponds to the time elapsed since the presentation of the conditioning stimulus [12] . Further to this observation , there has been a surge of interest in modeling dopaminergic activity in terms of TD-learning concepts , as reviewed in [13] . Temporal difference algorithms are based on the assumption that the information available for decision making is rich enough to make the learning problem Markovian . This means that the future is independent of past events , given the current state accessible to the TD-learner . In contrast , eligibility trace based approaches such as our population learning do not require such a completeness of available information . Hence , we present simulation results comparing the performance of the proposed approach to that of TD-learning on tasks , where the Markovian assumption may be violated .
We consider a population of leaky integrate and fire neurons driven by a common presynaptic stimulus and read-out by a decision making circuitry . To facilitate exploration both the population neurons and the decision making are stochastic . As in forced choice tasks , the decision circuitry determines a behavioral choice at the end of stimulus presentation , based on its monitoring of the population activity for the duration of the stimulus . We focus on binary decision making and denote the two possible behavioral choices by . Immediately , or at some later point in time , a behavioral decision may influence whether reward is delivered to the system , but the decision may also impact the environment , i . e . influence the sequence of stimuli presented to the population neurons . Due to the last point , our framework goes beyond operant conditioning and also includes sequential decision tasks . For the decision making circuitry itself , we use a very simple model , assuming that it only considers the number of population neurons which fire in response to the stimulus: For low population activity the likely decision is , but the probability of generating the decision increases with the number of neurons that respond by spiking to the stimulus . Given this decision making circuitry , we present a plasticity rule for the synapses of the population neurons , which enables the system to optimize the received reward . In presenting the plasticity rule we focus on one synapse , with synaptic strength , of one of the population neurons . ( In the simulations , of course , the rule is applied to all synapses of all population neurons . ) Let be the set of spike times representing the presynaptic spike train impinging on the synapse upto time . A presynaptic spike at some time leads to a brief synaptic release with a time constant on the order of a millisecond . The postsynaptic effect of the release will however linger for a while , decaying only with the membrane time constant which is in the range . The first synaptic eligibility trace bridges the gap between the two time scales by low pass filtering ( Fig . 2 , column 1 ) . It evolves as: ( 1 ) Correlations between synaptic and post-synaptic activity are captured by transcribing into a second trace of the form ( 2 ) see Fig . 2 , column 2 . The postsynaptic modulation function depends on the postsynaptic spike times and on the time course of the neuron's membrane potential . Denoting by the set of postsynaptic spike times , the specific form we use for isHere is Dirac's delta-function , and are parameters given in Methods . As has been previously shown [14] , is a useful factor in plasticity rules due to the following properties: Thanks to these properties , plasticity rules where synaptic change is driven by the product of and reward have been widely used in reinforcement learning models [6] , [15]–[17] . Due to , the neuronal quantities modulating plasticity in these rules are not just the pre- and post synaptic firing times but also the membrane potential . This further modulatory factor also arises in models matching STDP-experiments which measure plasticity induction by more than two spikes [18] . In our model , the time constant in Eq . ( 2 ) should be matched to the decision time during which stimuli are presented and we use . Since the match may be imperfect in reality , we denote the actual stimulus duration by the symbol . To describe the stochastic decision making in this period , we introduce the population activity variable which is reset each time one decision is made and subsequently increased when a neuron spikes for the first time in response to the next presented stimulus ( Fig . 2 , column 3 ) . A high ( low ) value of at the end of the decision period biases the next behavioral decision towards ( ) . We do not model the temporal accumulation of population activity leading to explicitly in neural terms , since this could be achieved along the lines previously suggested in [19] . Since the decision circuitry is stochastic , even for a fairly high level of population activity the behavioral decision may be made by chance . In this case , by spiking , a population neuron in fact decreased the likelihood of the behavioral choice which was actually taken , whereas a neuron that stayed silent made the choice more likely . Hence , when the goal is to reinforce a behavioral decision , a sensible strategy is to reinforce a neuronal response when it is aligned with ( firing for , not firing for ) and to impede it when it is not aligned . To this end , the third eligibility trace captures the interactions between single neuron activity , population activity and behavioral decision . It evolves as ( 3 ) where is a feedback signal , based on and , generated by the decision making circuitry and , further , is determined by the postsynaptic activity of the neuron . Mathematically , should reflect how the neuron contributed to the decision and equal according to whether or not the neuron fired in response to the decision stimulus . The feedback signal should consist of pulses generated at the times when a decision is made . The value of should have the same sign as the corresponding decision and be modulated by the population activity which gave rise to the decision . In particular , the magnitude of the pulse is large when is close to the stochastic decision threshold , increasing synaptic plasticity in the cases where the decision making is still very explorative . Since the post-stimulus value of has the same sign as , the term in Eq . ( 3 ) is positive when the neuronal response is aligned with the decision - otherwise it is negative . Because this term remodulates during the transcription and in view of the above characterization of , the eligibility trace has the following property: Since encodes the correlations between the releases of the synapse and the behavioral decision , the final stage of the cascade becomes very simple ( Fig . 2 , column 4 ) . It just remodulates by reward to yield the synaptic change: ( 4 ) Mathematically , the reward function should be made up of pulses at the times when external reinforcement information becomes available , with the height of each pulse proportional to the reward received at that time . The above description uses some mathematical idealizations which biologically are not quite realistic . We envisage that the reinforcement and decision feedback is delivered to the synapses by changes in levels of neurotransmitters such as dopamine , acetylcholine or norepinephrine [20]–[22] . Then , in contrast to the pulses assumed above , the feedback read-out by the synapses should change only quite slowly . In our simulations , this is addressed by low pass filtering the above feedback pulses when obtaining the signals and . Further , we assumed above that in Eq . ( 3 ) encodes whether the neuron fired in response to the decision stimulus . But it seems unrealistic , that a population neuron knows when a stimulus starts and ends . In the simulations we use low pass filtering to compute a version of which just encodes whether the neuron spiked recently , on a time scale given by ( Methods ) . Such a delayed feedback about postsynaptic activity could realistically be provided by calcium related signaling . To study the proposed plasticity rule , we first consider an operant conditioning like task , where for each of the stimuli presented to the network , one of the two possible behavioral decisions is correct . A correct decision is rewarded , whereas an incorrect one is penalized , but in both cases the delivery of reinforcement is delayed for some time . While operant conditioning with delayed reward has been widely considered in the context of temporal discounting [23] , here , we are interested in a quite different issue . We do not wish to assume that little of relevance happens in the delay period between the decision and the corresponding reinforcement since this seems artificial in many real life settings . In the task we consider , during the delay period , other decisions need to be made which are themselves again subject to delayed reinforcement ( Fig . 3A ) . Then temporal contiguity between decision and reward is no longer a proxy for causation . So the issue is not how to trade small immediate reward against a larger but later reward , but how to at all learn the association between decision and reward . In the simulations , a stimulus is represented by a fixed spike pattern made up of 80 Poisson spike trains , each having a duration of and a mean firing rate of 6 Hz . To allow for some variability , on each presentation of the stimulus , the spike times in the pattern are jittered by a zero mean Gaussian with a standard deviation of . This stimulus representation is used throughout the paper . In the present task , we use 10 stimuli and , for each , one of the two possible decisions is randomly assigned as the correct one . Stimuli are presented in random order and right after the decision on one stimulus has been made , the next stimulus is presented . Fig . 3B shows learning curves for tasks where there is a fixed delay between each decision and the delivery of the reinforcement pertinent to that decision . Perfect performance is eventually approached , even for the largest value of considered . For this value , , two other decisions are made in the delay period . Learning time increases in a stepwise manner when extending the delay , with a step occurring each time a further intervening decision has to be made in the delay period ( Fig . 3B inset ) . To demonstrate that the proposed plasticity rule addresses the spatial credit assignment problem as well , we studied learning performance as function of the number of population neurons . The results in Fig . 3C show that learning speeds up with increasing population size . In a larger population there are more synapses and the speedup indicates that the plasticity rule is capable of recruiting the additional synapses to enhance learning . To gauge robustness , we used the same synaptic plasticity parameters for all simulations in Panels B and C . In particular was always set to even though the actual delay in reward delivery is varied substantially in Panel B . To further highlight robustness , Fig . 3D shows the performance for different values of when the actual delay in reward delivery is fixed at . In the above simulations the delay between decision and reward did not change from trial to trial . But the proposed plasticity rule does not rely on this for learning and also works with variable delays . This is shown in Fig . 3E , where a different , randomly chosen , delay was used on each trial . To achieve near perfect performance in the above operant conditioning task , our network had to learn to make close to deterministic decisions . Here we show that , when appropriate , the architecture can also support stochastic decision making . For this we consider a two armed bandit where one of the two targets delivers a fixed reward of when chosen . The second choice target ( which we call intermittent ) will deliver a reward of or depending on whether or not the target is baited . Baiting occurs on a variable interval schedule: Once the reward of has been collected , the target becomes un-baited . It stays un-baited for between to time steps ( randomly chosen ) and is then baited again . Once baited , the target stays in this state until it is chosen . As a consequence , always choosing the intermittent target yields an average reward equal to . This does not improve on choosing the fixed reward target and , hence , a better policy is to pick the intermittent target less frequently . We assume that our network does not have access to the past decisions it has made . Hence on every trial one and the same stimulus is presented to the network ( with the same spike pattern statistics as in the previous subsection ) . The evolution of the average reward collected by the network is shown in Fig . 4A . Due to learning , average reward increases , reaching a value which is within of the reward achievable by the optimal stochastic policy . The probability of choosing the intermittent target decreases from to around as shown in Fig . 4B . This panel also plots the evolution of the value of choosing the intermittent target . The value being the expected reward collected from choosing the intermittent target assuming that the policy is to pick this target with a probability of . Asymptotically approaches a value around . So choosing the intermittent target is much more rewarding on average than choosing the fixed target ( which has a value of ) . Nevertheless , the intermittent target is chosen less frequently than the fixed target . This amounts to a strong deviation from matching or melioration theory [24] which stipulates that choice frequencies adjust up to the point where the value of the two choices becomes the same - this would lead to in the present task . On a task similar to ours , deviations from matching and melioration , favoring a more global optimization of reward , have also been observed in a behavioral experiment with rats [25] . Our plasticity rule , of course , does not explicitly value choices but directly adapts the choice policy to optimize overall reward . This is in contrast to temporal-difference ( TD ) based approaches to learning , where estimating the value of choices ( or , more generally , the value of state-action pairs ) is the key part of the learning procedure . Hence it is of interest to compare the above results to those obtainable with TD-learning . The two most common strategies in TD-learning for making decisions based on the valuation of choices are -greedy and softmax . For -greedy the choice with the highest estimated value is taken with probability , where is typically a small positive parameter . This does not allow for a fine grained control of the level of stochasticity in the decision making , so we will only consider softmax here . For softmax , a decision is made with a probability related to its value as . Here the positive parameter , called inverse temperature , modulates the level of stochasticity in the decision making . TD-theory does not give a prescription for choosing and , hence , we will consider a large range of values for the inverse temperature . The results in panels 4C and 4D plot the asymptotic performance as function of . Panel 4c shows that the average reward achieved by the TD-learner decreases with increasing . So best performance is obtained for , i . e . when the choice valuations estimated during learning are irrelevant . The probability of choosing the intermittent target increases with , Panel 4D . The panel also shows that the estimates of computed by the TD-algorithm are in excellent agreement to the true values of for the policy characterized by . Hence , the TD-learner fails to optimize reward not because the valuation of the decisions is wrong , but it fails because softmax is a poor strategy for transforming valuations into decisions in the present task . The root cause for the failure of TD-learning is that our decision task is not Markovian . Due to the variable interval schedule , the probability that the intermittent target is baited depends on the previous decisions made by the TD-learner . But as in the simulation on population learning , we have assumed that previous decisions are not memorized and the TD-learner is in the same state in each trial . Hence , even given the state accessible to the TD-learner , past events are nevertheless predictive of future ones because the information about the present encoded in the state is incomplete . This violates the Markovian assumption on which TD-learning theory is based . To rectify this , one needs to assume that decisions are made in view of previous decisions and outcomes . Given that the intermittent target can stay un-baited for a maximum of 12 steps , this requires a TD-learner which memorizes decisions and outcomes ( reward/no reward ) for the last 12 time steps . Hence , we simulated a TD-learner with the states needed to represent the task history in sufficient detail to render the decision problem Markovian . We found that the algorithm after learning ( with softmax and ) achieved an average reward of per decision . The algorithm learned to employ sophisticated policies such as not choosing the intermittent target for 8 time steps after it had delivered reward - but polling it frequently thereafter until the intermittent target again delivered reward . Obviously such policies are beyond the scope of the simple memoryless stochastic decision making considered above . We next studied population learning in a sequential decision making task , where reward delivery is contingent on making a sequence of correct decisions . For this , a simple path finding task on a linear track was used ( Fig . 5A ) . We imagine an owner who is tired of having to take his dog for a walk and wants to teach the animal to exercise all by itself . The dog is put in front of the door ( position 1 on the track ) , can move left or right , and may be rewarded on coming home ( position 0 ) . But since the point is to exercise the dog , reward ( ) is only delivered when the dog has reached position 3 at least once while moving on the track . If the dog comes home early without visiting the required position 3 , the learning episode simply ends with neither reward or punishment . The episode ends in the same way if position 5 is ever reached ( the dog should not run away ) . In an initial simulation , we assumed that decisions have to be made based just on the current position on the track . So the stimuli presented to the population just encode this position ( using the same spike pattern statistics as in the previous tasks ) . Given such stimuli , our population model is faced with a non-Markovian decision problem because , the appropriateness of a decision may depend not just on the current stimulus but also on the stimuli which were previously encountered . For instance , whether one should go left or right in position depends on whether position has been visited already . In fact the learning problem is even more dire . When the basis of decision making is just the current position , complete failure will result for any deterministic policy which must lead to one of the following three outcomes: ( i ) direct exit from position 1 to , ( ii ) exit at position , ( iii ) an infinite cycle . This is not to say that nothing can be learned . As the result in the bottom row of Fig . 5A shows , it is possible to increase the odds that an episode will end with reward delivery by adapting a stochastic policy . Initially the network was almost equally likely to go left or right in any position but after learning this has changed . In position for instance left is much more likely than right , whereas , in position , left is just a little bit more likely than right . After learning , the average number of steps per episode is lower than initially ( Fig . 5B , red curve ) . So in terms of average reward per step taken , there is even more improvement through learning than suggested by the blue curve in Fig . 5B . In the simulations we used . This is somewhat longer than the minimal time of 2 . 5 s ( 5 steps of duration ) needed from position to reward delivery . Thanks to working memory , a real dog is of course entirely capable to collect reward by simply running from position to and then back to . So for describing the behavior of an animal with a highly developed nervous system , the above model is woefully inadequate . Nevertheless , it may usefully account for behavior in the presence of working memory impairments . To allow for working memory , in a next set of simulations we switched to stimuli encoding not just the current but also the immediately preceeding position on the track . Of the 80 spike trains in a stimulus presented to the network , 50 were used to encode the current and 30 to encode the preceeding position ( Methods ) . Now , learning with the proposed plasticity rule converges towards perfect performance with the reward per episode approaching and the number of decision steps per episode approaching ( Fig . 5C ) . It is worthwhile noting , that even with a working memory reaching one step back , the decision task is non-Markovian: For instance , knowing that coming from we are now in position does not allow us to tell whether moving left leads to reward . For this we would need to know if we have been in position , say , two steps back . Technically , when remembering the sequence of past positions , the memory depth required to make the decision problem Markovian is infinite because any finite memory can be exhausted by cycling many times between positions and . The non-Markovian nature of the task is highlighted by Fig . 5D , which shows simulation result for TD-learning . The specific algorithm used is SARSA with -greedy decision making ( see [1] and Methods ) . Similarly to Fig . 5C , we assumed that the states upon which the TD-learner bases decisions represents the current and the immediately preceeding position on the track . The solid blue curve in Fig . 5D , computed by averaging performance over multiple runs of the algorithm , demonstrates that TD-learning does not converge towards perfect performance . The dotted blue curve , giving results for a typical single run , shows that in fact TD-learning leads to large irregular oscillations in performance , which are averaged away in the solid curve . While optimal performance is approached initially in the single run , the algorithm is not stable and at some point performance breaks down , initiating a new cycle in the oscillation . To understand the instability in more detail , we denote the states of the TD-learner by notation such as 21 , meaning that coming from the current position is . The TD-learner assigns values to state-decision pairs , which we write as e . g . ( 21 , left ) , by estimating discounted future reward . Now consider the single run of the TD-learner ( dotted blue curve , Fig . 5D ) after some 1500 episodes . The strategy then is close to optimal , so in most episodes when we are in state 21 , i . e . on the inbound leg of the tour , position 3 will have previously been visited . Then left in 21 leads to immediate reward delivery , so the state-action pair ( 21 , left ) has a high value . Next assume that we are on the outbound leg in state 12 . Since the policy is close to optimal , in most episodes the next move is right , in order to visit position 3 . But , due to exploration , the TD-learner will occasionally try the shortcut of going left in state 12 , testing the state-action pair ( 12 , left ) . This leads to state 21 and then most likely to the high value decision left , terminating the episode without reward because the shortcut was taken . But the TD-learner updates the value of the tested state-action pair ( 12 , left ) based not on the failure at the very end of the episode but based on the value of the subsequent state-action pair , in this case ( 21 , left ) . As noted above , the latter pair has high value , so the update increases the value of the shortcut ( 12 , left ) even-though the shortcut resulted in failure ( green curve in Fig . 5D ) . This happens most of the times when the shortcut is tested for exploration , leading to further increases in the green curve , upto the point where the value of ( 12 , left ) is so high that making a shortcut becomes the dominant policy . This causes the observed breakdown in performance . In summary , a central idea in temporal difference learning is to handle non-immediate reward by back-propagating it in time via the valuations of intermediate state-decision pairs . This is mathematically justified in the Markovian case , but may lead to unexpected results for general sequential decision making tasks .
We have presented a model of reinforcement learning in a population of spiking neurons read out by a decision making circuitry where plasticity induction is controlled by a cascade of synaptic memory traces . In each synapse of the population neurons , the presynaptic trace is in stages remodulated by somatic feedback , by feedback about the behavioral decision making and by an external reward signal before being consolidated into a persistent change of the synaptic strength . Simulation results show that this leads to robust learning performance in a variety of reinforcement tasks . Our model builds on , but goes beyond , the classical STDP findings [2] , [3] , [26] . On the neuronal level , we assume that plasticity does not only depend on the timings in a pre- and postsynaptic spike pair but that there is a further modulation by postsynaptic subthreshold activity . Such a modulation also arises when modeling the plasticity findings obtained when the standard STDP-protocol is extended to allow multi spike interactions [18] . For reinforcement learning , plasticity cannot be blind to activity-related downstream information . This matches with experimental observations revealing that the polarity and magnitude of STDP can in fact be regulated by neuromodulators such as dopamine , acetylcholine or noradrenaline which may even revert the sign of the synaptic change [10] , [21] , [22] , e . g . by entering after the mGluR signaling pathways [27]–[29] . Some recent research has further highlighted astrocytes as local communication elements which are capable of modulating synaptic plasticity [30] , [31] . Research on synaptic tagging has revealed the astonishingly large time span during which the consolidation of early-LTP into long lasting synaptic change can be dependent on behavioral reinforcement [32] , [33] . The present work provides a phenomenological model showing how the multi-stage processes observed in the induction of long-term synaptic plasticity can be bound into a functional whole . Previous modeling of population learning has already considered the modulation of plasticity by feedback from the decision circuitry [16] , [34] . However , in these works the cascade was shortcut , with decision and reward feedback interacting directly in the modulation of plasticity . As a consequence the previous plasticity rule was capable of handling delays between decision and reward feedback only when these where very small , namely a fraction of typical stimulus duration . The present rule achieves a far more general solution to the temporal credit assignment problem by using a further stage in the synaptic cascade to decouple decision from reward feedback . Further , the rule is now based directly on optimizing the average reward rate ( Text S1 ) and not just , as previously , a related objective function . This puts the present approach squarely into the field of policy gradient methods [35]–[37] . Within this field , our main contribution is to show how the spatial credit assignment problem of distributing the learning between the population neurons can be solved in a biophysically plausible way . As the results in the section on learning stimulus-response association demonstrate , our plasticity rule leads to a learning performance which scales well to large population sizes ( a more detailed scaling analysis has been given in [34] ) . This is in contrast to the straightforward policy gradient approach of treating the neurons as independent agents which results in a rapid deterioration of learning performance with increasing population size [16] . Crucially in our population model neurons need to cooperate in order to receive reward and hence during learning a difficult spatial credit assignment problem arises . The appropriateness of any single neuron response cannot be determined without taking the responses of the other neurons into account and hence synapses in different neurons need to co-adapt in optimizing reward . This is in contrast to previous work [38] modeling a biofeedback experiment in monkeys [39] where reward delivery was contingent on the firings of a single target neuron . In the model [38] background activity was high , so that reinforcement could be increased by simply strengthening the synapses of the target neuron without any need for coordinated adaptation by the other neurons in the system . Some parameters in our plasticity scheme are related to properties of the learning task . For instance the time constant in the last stage of the cascade represents a guess at the typical delay between decision and reinforcement . Our simulation results indicate that learning is not overly sensitive to the choice of the synaptic parameters ( see e . g . Fig . 3D ) . Nevertheless , learning does of course deteriorate once the mismatch between synaptic and actual task parameters becomes too large . An intriguing possibility for further increasing robustness could be an inhomogeneous population of neurons . After all , a key point in population coding is to provide redundancy [40] , [41] . This is borne out by findings in [16] where , with increasing population size , decision performance improves but the correlation between single neuron performance and decision decreases . Hence it is of interest to study learning when different population neurons have different synaptic parameters . Then the neurons with parameters best matched to the task at hand , are expected to learn best . Thanks to their resulting correlated activity , they should be able to carry the population decision because the contributions from the badly learning mismatched neurons should be uncorrelated and thus tend to cancel . Unfortunately , meaningfully testing whether neuronal variability increases robustness in this manner , requires the simulation of population sizes which are an order of magnitude larger than what is currently within our computational reach . With regard to the temporal credit assignment problem , we think it is important to note that delayed interaction between decision making and reward delivery can arise in diverse manners: Policy gradient methods work in all of the above settings . Of course , missing information can be detrimental to the performance which is achievable at all . But , given this constraint , policy gradient methods will nevertheless optimize the performance . Temporal difference ( TD ) methods , however , by design handle only problems of type iii . In the first two cases TD-learning only applies when the state which serves as basis for the decision making represents the recent task history to the extent that the problem becomes Markovian . Formally , this maps the first two kinds of delays onto the third kind . Representing the recent task history is what working memory is good for - and working memory is well known to enter into decision making as in delayed match to sample tasks . On the other hand , transforming a non-Markovian into a Markovian decision problem can pose daunting demands on the working memory capacity needed to adequately represent the states in the TD-algorithm . With insufficient working memory the algorithm can fail in two distinct ways . The estimates for the value of some state-action pairs may be wrong ( as demonstrated in the sequential decision making task ) , or , even when the estimates are correct , preferentially choosing the available action with highest estimated value may lead to a suboptimal policy ( as in the two armed bandit ) . Policy gradient methods such as our population learning rule seem attractive as basic biological models of reinforcement learning because they work in a very general setting . Arguably , this generality is also a drawback . Precisely because the Markovian property is restrictive , exploiting it in the cases where it does apply , can substantially speed up learning . Hence , it is of interest that policy gradient methods can easily be combined with TD-state valuations in the framework of actor-critic methods . This amounts to simply replacing the direct reward signal in the policy gradient plasticity rule with a signal generated by the TD-valuation circuitry . The TD-signal can either be the estimated value of the current state [42] or the value prediction error [15] . Combining policy gradient with TD-valuations in this way , again brings about the Markovian restriction . Hence , if reinforcement learning is to be both robust and fast , issues of metaplasticity arise: How does brain learn how to learn when ?
The model neurons in our population are escape noise neurons [14] , i . e . leaky integrate and fire neurons where action potentials are generated with an instantaneous firing rate which depends on the membrane potential . Focusing on one of the population neurons , we denote by its input which is a spike pattern made up of spike trains . Each is a list of the input spike times in afferent . We use the symbol to refer to the postsynaptic spike train produced by the neuron , is also a list of spike times . If the neuron , with synaptic vector , produces the output in response to , its membrane potential is determined by ( 5 ) Here is the unit step function and , further , is Dirac's delta function , leading to immediate hyperpolarization after a postsynaptic spike . For the resting potential , denoted above by , we use ( arbitrary units ) . Further , is used for the membrane time constant and for the synaptic time constant . By integrating the differential equation , the membrane potential can be written in spike response form as ( 6 ) The postsynaptic kernel and the reset kernel vanish for . For they are given byNote that the first eligibility trace of synapse can be expressed in terms of the postsynaptic kernel as . Action potential generation is controlled by an instantaneous firing rate which increases with the membrane potential . So , at each point in time , the neuron fires with probability where represents an infinitesimal time window ( we use in the simulations ) . Our firing rate function iswith and . ( In the limit of one would recover a deterministic neuron with a spiking threshold . ) As shown in [14] , the probability density , , that the neuron actually produces the output spike train in response to the stimulus during a decision period lasting from to satisfies: ( 7 ) The derivative of with respect to the strength of synapse is known as characteristic eligibility in reinforcement learning [35] . For our choice of the firing rate function one obtains ( 8 ) where is the first eligibility trace of the synapse ( Eq . 1 ) and the postsynaptic signal of the neuron given right below Eq . ( 2 ) . Note that ( 8 ) is similar to our second eligibility trace , see Eq . ( 2 ) , except that we have replaced the integration over the decision period by low pass filtering with a time constant matched to the stimulus duration . The reason for this is that it seems un-biological to assume that the synapses of the population neurons know when decision periods start and end . We use the superscript , running from to , to index the population neurons . For instance , is the postsynaptic spike train produced by neuron in response to its input spike pattern . As suggested by the notation , the population neurons have different inputs , but their inputs are highly correlated because the neurons are randomly connected to a common input layer which present the stimulus to the network . In particular , we assume that each population neuron synapses onto a site in the input layer with probability , leading to many shared input spike trains between the neurons . The population response is read out by the decision making circuitry based on a spike/no-spike code . For notational convenience we introduce the coding function , with , if the there is no spike in the postsynaptic response , otherwise , if neuron produce at least one spike in response to the stimulus , . In term of this coding function the population activity being read out by the decision making circuitry can be written as:Using this activity reading , the behavioral decision is made probabilistically , the likelihood of producing the decision is given by the logistic function ( 9 ) Note that due to the normalization in the definition of , the magnitude of can be as large as . This is why , decisions based on the activity of a large population can be close to deterministic , despite of the noisy decision making circuitry . We start with the reward feedback , modulating synaptic plasticity in Eq . ( 4 ) . This feedback is encoded by means of a concentration variable , representing ambient levels of a neurotransmitter , e . g . dopamine . In the absence of reward information , the value of approaches a homeostatic level with a time constant . For any point in time when external reward information is available , this reinforcement leads to a change in the production rate of the neurotransmitter . The change is proportional to and lasts for . So up to the point in time when further reinforcement becomes available , the concentration variable evolves as:Here the step function equals if , otherwise the function value is zero . The reward feedback read-out at a synapse is determined by the deviation of the current neurotransmitter level from its homeostatic value and equalsHere the parameter is the positive learning rate which , for notational convenience , we absorb into the reward signal . The decision feedback used in Eq . ( 3 ) is encoded in the concentration of a second neurotransmitter . As for reward feedback , this is achieved by a temporary change in the production rate of the encoding neurotransmitter . For describing , we assume a stimulus that ended at time , evoking the population activity and behavioral decision . As shown in Text S1 , the value of should then be determined by the derivative of with respect to and , in view of Eq . ( 9 ) , this derivative is simply . Hence we usefor the temporal evolution of . Parameter values in the simulations are and . The above equation holds up to time when the subsequent stimulus presentation ends , at which point the decision variables and are replaced by their values for the latter stimulus . The decision feedback is simply For the postsynaptic trace in Eq . ( 3 ) , we assume a concentration variable which reflects the spiking of the neuron . Each time there is a postsynaptic spike , is set to 1; at other times , decays as . The value of should reflect whether or not the neuron spiked in response to the decision stimulus . So , as for the eligibility trace ( see Eq . 2 ) , the relevant time scale is the decision period and this is why the same time constant is used in both cases . The trace is obtained ascomparing to an appropriate threshold . In the simulation we use . For the reasoning behind this choice , consider a stimulus ending at time of duration . The value of at time will accurately reflect whether or not the decision stimulus elicited a postsynaptic spike , if we choose . But since decision feedback is not instantaneous , the value of is mainly read-out at times later than . This is why the smaller value seemed a somewhat better choice . For TD-learning we used the SARSA control algorithm [1] which estimates the values of state-action pairs . At each point in time , the value estimates are updated according toHere and have values between and . The parameter is similar to a learning rate and controls the temporal discounting . The above update is done after every transition from a nonterminal state . If is terminal , then is defined as zero . When in state , the next action is chosen using either -greedy or softmax . In both cases only the values pertinent to the current state enter into the decision making . For memoryless TD-learning in the two armed bandit we used and . A positive discount factor would not qualitatively change the result . For each of runs per chosen value of , we simulated trials . After trials learning had converged and the reported asymptotic quantities are the average over the next trials . For learning with memory we used , and . For the sequential decision making task decision selection used -greedy with . The discount factor was set to and the step-size parameter to . With regard to the failure of TD-learning in the sequential decision making task , we note that there are also eligibility trace based versions , SARSA , of the algorithm with the above version corresponding to . For , the value update takes into account not just the next state-action pair but the value of all subsequent state-action pairs . Importantly , for the special case the subsequent values occurring in the update cancel , and the value update is in effect driven directly by the reward signal [1] . So SARSA is just a complicated way of doing basic Monte Carlo estimation of the values . It hence does not assume that the process is Markovian and SARSA does reliably converge towards optimal performance in our task . For the procedure interpolates between the two extremes and . Consequently the valuation of some state-action pairs ( e . g . the shortcut 12 , left ) will then be wrong but the error will be smaller than for . If action selection is based on softmax the incorrect valuation will nevertheless be detrimental to decision making . However , this need not always be the case for -greedy , due to the thresholding inherent in this decision procedure . In particular , there is a positive critical value for ( which depends mainly on the discount factor ) above which the valuation error will no longer affect the decision making . In this parameter regime , SARSA will reliably learn the optimal policy ( upto the exploration determined by ) . In all the simulations initial values for the synaptic strength were picked from a Gaussian distribution with mean zero and standard deviation equal to 4 , independently for each afferent and each neuron . A learning rate of was used in all simulations , except for the 2-armed bandit task where was used . In the sequential decision making task with working memory , the population is presented stimuli encoding not just the current but also the immediately preceeding position . For this , each location on the track is assigned to a fixed spike pattern made up of 50 spike trains representing the location in the case that it is the current position and , further , to a second spike pattern with 30 spike trains for the case that it is the immediately preceeding position . The stimulus for the network is then obtained by concatenating the 50 spike trains corresponding to the current position with the 30 spike trains for the preceeding position . The curves showing the evolution of performance were obtained by calculating an exponentially weighted moving average in each run and then averaging over multiple runs . For the sequential decision making task reward per episode was considered and the smoothing factor in the exponentially weighted moving average was . In the other task , where performance per trial was considered , the smoothing factor was . For each run a new set of initial synaptic strength and a new set of stimuli was generated . The number of runs was , except in the two armed bandit where we averaged over 40 runs .
|
The key mechanisms supporting memory and learning in the brain rely on changing the strength of synapses which control the transmission of information between neurons . But how are appropriate changes determined when animals learn from trial and error ? Information on success or failure is likely signaled to synapses by neurotransmitters like dopamine . But interpreting this reward signal is difficult because the number of synaptic transmissions occurring during behavioral decision making is huge and each transmission may have contributed differently to the decision , or perhaps not at all . Extrapolating from experimental evidence on synaptic plasticity , we suggest a computational model where each synapse collects information about its contributions to the decision process by means of a cascade of transient memory traces . The final trace then remodulates the reward signal when the persistent change of the synaptic strength is triggered . Simulation results show that with the suggested synaptic plasticity rule a simple neural network can learn even difficult tasks by trial and error , e . g . , when the decision - reward sequence is scrambled due to large delays in reward delivery .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"decision",
"making",
"neural",
"networks",
"neuroscience",
"learning",
"and",
"memory",
"cognitive",
"neuroscience",
"synaptic",
"plasticity",
"population",
"modeling",
"computational",
"neuroscience",
"population",
"biology",
"animal",
"cognition",
"coding",
"mechanisms",
"circuit",
"models",
"developmental",
"neuroscience",
"biology",
"neural",
"circuit",
"formation",
"behavioral",
"neuroscience"
] |
2011
|
Spatio-Temporal Credit Assignment in Neuronal Population Learning
|
We developed a model of cardiac sarcomere contraction to study the calcium-tension relationship in cardiac muscle . Calcium mediates cardiac contraction through its interactions with troponin ( Tn ) and subsequently tropomyosin molecules . Experimental studies have shown that a slight increase in intracellular calcium concentration leads to a rapid increase in sarcomeric tension . Though it is widely accepted that the rapid increase is not possible without the concept of cooperativity , the mechanism is debated . We use the hypothesis that there exists a base level of cooperativity intrinsic to the thin filament that is boosted by mechanical tension , i . e . a high level of mechanical tension in the thin filament impedes the unbinding of calcium from Tn . To test these hypotheses , we developed a computational model in which a set of three parameters and inputs of calcium concentration and sarcomere length result in output tension . Tension as simulated appeared in good agreement with experimentally measured tension . Our results support the hypothesis that high tension in the thin filament impedes Tn deactivation by increasing the energy required to detach calcium from the Tn . Given this hypothesis , the model predicted that the areas with highest tension , i . e . closest to the Z-disk end of the single overlap region , show the largest concentration of active Tn’s .
It has been widely known since the 1883 work of Ringer [1] that calcium ions ( Ca2+ ) regulate muscle contraction . Although the mechanism was unknown at the time , Ringer found that rat hearts did not contract forcefully unless Ca2+ was added to the solution perfusing the muscle . Since Ringer’s discovery the relationship between Ca2+ concentration ( [Ca2+] ) and tension in muscle has been widely studied . Experiments have shown a steep rise in steady state tension with relatively little increase in intracellular [Ca2+] . The desire to understand the steepness of the [Ca2+]—tension relationship has led to the hypothesis that cardiac muscle contraction is highly cooperative [2 , 3] . Cooperativity implies that a single event encourages subsequent similar events . In the case of muscle contraction , a single Ca2+ binding invoking the generation of tension might enhance additional Ca2+ bindings to generate more tension [4] . Various hypotheses about the mechanism of cooperativity in cardiac muscle have been proposed . Before discussing cooperativity , we explain the structures and molecular mechanisms involved in cardiac contraction . The cardiac muscle is striated , meaning that it is composed of a repeating organization of contractile units known as sarcomeres ( Fig 1A ) . Within the sarcomeres thick and thin filaments can interact with one another to generate tension [5 , 6] . The sarcomere is bounded by two opposing Z-disks with the thin filaments protruding from each ( Fig 1A ) . A thin filament contains a double helical arrangement of spherical actin monomers ( Fig 1B ) , each with a myosin binding site on its surface facing the outside of the helix [7] . A thin filament is surrounded by thick filaments , each having myosin heads protruding every 14 . 3 nm that can bind to actin to form a cross-bridge ( XB ) that can generate force [8] . The thick and thin filaments range in length from ~1 . 5 to ~1 . 7 μm and ~1 . 1 to 1 . 3 μm , respectively [9] . The distance between Z-disks is equal to the sarcomere length ( SL ) . Opposing thin filaments can mutually overlap in the so-called double overlap region ( Fig 1A ) . The single overlap region is the section of the thick filament outside of the double overlap region that overlaps with the thin filament . We assume that XB’s can form solely in the single overlap region [10] . As SL increases up to about 2 . 3 μm , the length of the double overlap region decreases , thereby lengthening the single overlap region , thus providing more opportunities for XB formation and subsequently higher force generation ( Fig 1A ) . Ca2+ activates the thin filament , allowing XB’s to form and force to develop ( Fig 1B ) . Seven actin monomers , one tropomyosin ( Tm ) molecule , and one troponin complex ( Tn ) form a repeating structure called a regulatory unit ( RU ) along the thin filament ( Fig 2A ) [11] . Tm molecules are double helical molecules that follow the same helical path as the actin monomers , overlapping one another head to tail under the Tn . At rest , the Tm molecule wraps around the actin monomers in a position that blocks the myosin binding sites , thereby hindering the formation of XB’s and thus contraction . Each Tn is composed of three subunits , the inhibitory subunit ( TnI ) , the Tm binding subunit ( TnT ) that attaches the Tn to Tm , and the Ca2+ binding subunit ( TnC ) [12] ( Fig 2B ) . TnI binds to actin monomers at rest and essentially anchors the Tm molecule in place , blocking the binding sites on actin [13] . When the cell depolarizes , Ca2+ enters and may bind to TnC opening a hydrophobic patch on the N-terminus of TnC . The COOH terminus of TnI exhibits a strong affinity for the NH2 terminus opened on TnC and binds strongly to it , and the inhibitory peptide of TnI releases actin[14] . The Tm strand moves laterally , thus freeing nearby XB binding sites on the thin filament ( Fig 2C ) . We refer to the latter movement of Tm as the activation of the Tn ( Fig 2B ) . It has been shown that , along with Ca2+ binding to TnC , the formation of a strong XB is necessary to fully unblock the binding sites on actin [15] . During relaxation the [Ca2+] in the cell decreases , TnC releases the bound Ca2+ ions , and TnI rebinds to actin moving the Tm back over the XB-binding sites preventing further force generation [16] . Commonly the relation between intracellular [Ca2+] and muscle tension is described by the Hill-relation , having zero active tension at low [Ca2+] and approaching a level of saturation at high [Ca2+] . If a single Ca2+ ion would bind to a single Tn complex without cooperativity , for a physiological muscle contraction , the required increase in [Ca2+] would be unphysiologically large . By comparing the relatively small , physiological change in [Ca2+] with the resulting large change in muscle tension , Ca2+ binding to Tn requires an increase of cooperativity by the so-called Hill coefficient n . In experiments by Dobesh et al . [18] it is found n≈7 . On the basis of conventional chemical principles , such a high level of cooperativity requires full cooperation between a number of binding sites at least equal to the Hill coefficient . In cardiac muscle , each Tn contains three binding sites for Ca2+ . Each of the two intertwined actin strands in the thin filament contains a Tn after every sequence of seven actin monomers . The Tn’s of both actin strands located at equal lengths form repetitive pairs of Tn’s , or in-register Tn’s . Galińska-Rakoczy et al . showed that the C-terminal section of TnI extends over the actin monomer on the adjacent strand reaching to the Tm molecule [17] . Their results suggest that upon Ca2+ binding the Tn may influence the position of the Tm molecule on the surface of the adjacent actin strand as well . Because of this interaction , we consider some degree of cooperativity of Ca2+ binding to in-register Tn’s . Each Tn complex has 3 binding sites for Ca2+ , with one of them having a substantially lower affinity for Ca2+ than the others [19] . So , potentially , each pair of in-register Tn complexes contains 6 binding sites , limiting the intrinsic Hill coefficient to a maximum value of 6 if each of the binding sites exhibits an equal degree of cooperativity . The term ‘intrinsic’ is used to indicate that the binding is considered purely chemical , excluding effects of mechanical tension . In experiments by Sun et al . [20] , in the absence of muscle tension , Ca2+ was bound to Tn with a Hill coefficient of 3 , indicating that the intrinsic cooperativity was about half of the abovementioned maximum . Even in the case of full cooperativity between in-register Tn’s , i . e . with an intrinsic cooperativity coefficient equaling 6 , one more binding site on a Tn in the next RU would be necessary to reach a cooperativity represented by the Hill coefficient of 7 in experiments of Dobesh et al . [18] . However , Tn’s are separated along the thin filament Tn’s by about 35 nm , a distance too large for conventional mechanisms of chemical cooperativity . In many current sarcomere models , the trend is to represent cooperativity along the thin filament as the interaction between neighboring XB’s and RU’s . Razumova and colleagues [21] developed a stiffness-distortion model of the XB and tested three different models of nearest neighbor cooperative interactions ( RU-RU , XB-XB , XB-RU ) . They found that while none of these interactions were able to completely account for the cooperative force development observed in experiments , each mechanism had a large impact on the [Ca2+]-tension relationship . XB-XB interactions had the highest impact on peak force , RU-RU interactions had the greatest impact on the steepness of the [Ca2+]-tension relationship , and XB-RU interactions had the greatest impact on Ca2+ sensitivity . The model as developed by Rice and colleagues [22] was based on the model of Razumova and colleagues [21] including assumed nearest neighbor cooperativity . Campbell and colleagues [23] have developed a model of cooperativity in which adjacent Tm molecules overlap one another at the Tn , thereby encouraging RU activation when the nearest neighbor RU is activated . It is not clear yet on what physical principles such RU-RU , RU-XB and XB-XB interactions would work . It has further been proposed that at longer SL’s the constant volume of the myofiber lattice causes the filaments to squeeze closer together , possibly encouraging binding events by moving a myosin head within reach of a binding site on the thin filament [24 , 25] . Similarly , Daniel et al . propose that the cooperative realignment of binding sites occurs when the thin filament is strained by up to 2 nm [26] , a lateral movement that could possibly move a binding site within reach of a myosin head . Still others believe that the end-to-end interactions of the Tm’s along the thin filament are responsible for the cooperative activation [27] . In the present study we introduce a mechanism based on the mechanochemical interaction of Ca2+ binding to Tn . Binding of Ca2+ to Tn results in conformational changes in the Tn complex . We postulate that mechanical stretch of the thin filament facilitates the change in conformation by reducing the energetic increment required for binding of Ca2+ to Tn . Thus , mechanical tension in the thin filament will shift the equilibrium of Ca2+ binding to Tn towards the bound state . The latter shift in equilibrium is quantified by the use of the general physical principle that the ratio of prevalence of two states depends on the difference in energy between the two states . We think that such a mechanism shows properties of cooperativity even though it differs from a pure chemical mechanism . On the basis of the abovementioned mechanochemical principle , we designed a computational model of cardiac sarcomere mechanics that we call the MechChem model . The MechChem model is intended to predict the tension in a thin filament as a function of [Ca2+] and sarcomere length . We investigated whether the mechanochemical mechanism can boost cooperativity from the intrinsic level with a Hill coefficient of 3 described by Sun et al . [20] up to the level of 7 shown by Dobesh et al . [18] in skinned muscle preparations . The energy to bind Ca2+ to Tn is assumed to decrease linearly with thin filament tension . Currently we focus on the static behavior , implying that thin filament tension was calculated while maintaining constant sarcomere length and [Ca2+] . Under static conditions we considered the chemical reactions involved to be in equilibrium . In investigating the static behavior of cardiac muscle contraction , muscle tension as simulated with the MechChem model is compared with simulations based on conventional chemical binding of Ca2+ to Tn with a high degree of cooperativity , resulting in the typical S-shaped Hill curve . Experimental results to compare with were obtained from skinned muscle preparations , subject to various static levels of [Ca2+] and sarcomere length as reported by Dobesh et al [18] . When fitting both the MechChem model and Hill-type model to experimental findings , a [Ca2+]-tension relation is found for each sarcomere length , requiring the estimation of 3 parameters . Furthermore , it is investigated which parameters values of the MechChem model could remain while changing sarcomere length .
Formation of XB’s is initiated by the binding of Ca2+ to TnC which we will refer to as the binding of Ca2+ to an in-register pair of Tn , referred to in the equations as Tn2 . The related chemical equilibrium reaction is represented by Tn2+n Ca2+⇆Tn2Can ( 1 ) Parameter n indicates the coefficient related to the intrinsic cooperativity of Ca2+ binding to Tn2 . When Ca2+ binds to Tn , Tm moves away from the position in which it blocks the XB-binding sites on the actin monomers . We assume that the unblocking of XB-binding sites , or activation of a Tn , automatically implies XB formation in the environment of that Tn . After formation , the XB exerts a longitudinal force FXB on the thin filament which is guided as tension along the thin filament to the Z-disk ( Fig 3 ) . Thus , toward the Z-disk , the longitudinal tension S ( x ) in the thin filament is increasing by a step FXB at each location where a XB has formed . Variable x represents the distance from the beginning of the single overlap region moving toward the Z-disk ( Fig 3 ) . The maximum value ( xmax ) of x represents the location on the thin filament where the single overlap region ends . The single overlap length depends on SL and the assumed filament lengths as shown in S1 Appendix . Looking from the middle of the sarcomere towards the Z-disk , for tension S ( xj ) just distal to XBj at position xj it holds: S ( xj ) = ∑i=1i=jFXB , i with 1≤j≤jmax ( 2 ) The symbol jmax indicates the number of XB’s attached to the thin filament in the single overlap region . Eq ( 2 ) is graphically elucidated in Fig 3 . Tension S ( x ) increases with each attached XB until distance x exceeds the end of the single overlap region where no XB's can be formed . The total tension of the thin filament equals the tension acting on the Z-disk from that thin filament . The following assumptions are key to the MechChem model implementation presented in this section: Referring to Eq 2 , the principle of the model is explained by the existence of discrete locations where XB’s can form . In living muscle at any given time during contraction , there are countless XB’s . Thus , we decided to represent muscular force generation by the average of many discrete states , allowing us to describe tensions as continuous functions of distance x along the representative thin filament . Focusing on the static conditions only , the involved chemical reactions are considered in equilibrium . According to Eq 1 , we consider binding of Ca2+ to Tn as a chemical binding to an in-register Tn pair , named Tn2 . Equilibrium concentrations of Ca2+ , Tn2 , and Tn2Can are related by: KTnCan= [Tn2][Ca2+]n[Tn2Can] ( 3 ) The symbol KTnCa represents the equilibrium constant , and n is the base cooperativity coefficient representing the intrinsic cooperativity of Ca2+ binding to Tn2 . We assume that when Ca2+ binds to the Tn2 , the nearby XB-binding sites are unblocked . For the proportion of activated Tn’s , P ( x ) , at position x along the thin filament , it holds P ( x ) = [Tn2Can] ( x ) [Tn2] ( x ) +[Tn2Can] ( x ) ( 4 ) Solving Eq ( 3 ) for [Tn2Can] and substituting the result into Eq ( 4 ) renders an expression for P ( x ) as a function of [Ca2+] and KTnCa ( Eq 5 ) . Because in the MechChem model binding of Ca2+ to Tn2 is assumed to depend on mechanical tension in the thin filament , the equilibrium constant KTnCa is considered to depend on the position x along the thin filament . The energy required to detach Ca2+ from the Tn is assumed to increase linearly with the tension S ( x ) in the thin filament . Thus , the equilibrium constant is multiplied by the exponential of the product of tension S ( x ) and a constant Cs , representing the effect of the added affinity for Ca2+ by an increase of tension . The latter relation is based on the general physical principle that in equilibrium the ratio of state concentrations is proportional to the exponential of a constant multiplied by the energy difference between both states . Analogously , the ratio of ion concentrations on both sides of a membrane depends on the voltage difference across the membrane . So , we express the dependence of the equilibrium concentration constant , KTnCa on tension S ( x ) by: KTnCa ( x ) = KTnCa0e−CsS ( x ) ( 6 ) The symbol KTnCa0 represents the equilibrium constant in the absence of tension in the thin filament . The physical dimension of constant Cs is the inverse of tension . Replacing KTnCa ( x ) in Eq ( 5 ) with Eq ( 6 ) yields the tension-dependent expression for P ( x ) in Eq ( 7 ) . We assume that the density of attached XB’s is proportional to P ( x ) . Assuming that all attached XB’s exert the same force during steady state isometric contraction , we find that the XB–induced force density fXB ( x ) acting at location x is proportional to P ( x ) ( Eq 8 ) . Constant Cf represents force density with full Tn activation , having the physical dimension of force per unit length along the thin filament . After elimination of P ( x ) by substitution of Eq 7 into Eq 8 and using the property that the derivative dS ( x ) /dx of tension with respect to x equals force density fXB ( x ) , we find the following differential equation for S ( x ) with boundary condition S ( 0 ) = 0: dS ( x ) dx= Cf1+ en ( −CsS ( x ) −ln ( [Ca2+]KTnCa0 ) ) ( 9 ) Assuming n = 3 according to the findings by Sun et al . [20] for the intrinsic thin filament cooperativity , Eq ( 9 ) contains 3 independent parameters , i . e . , Cf , Cs and KTnCa0 . We have fitted the Hill-type model with the experimental data of Dobesh et al . [18] for comparison , expressing tension SH as a function of [Ca2+] . The three parameters EC50 , nH , and Smax , represent the [Ca2+] at the 50% level of maximum tension , the Hill coefficient , and maximum tension , respectively . Thus we used SH= Smax[Ca2+]nHEC50nH+[Ca2+]nH ( 10 ) Under normal conditions , cardiac muscle cells are enclosed by membranes that regulate the influx of Ca2+ ions from the extracellular space through channels and pumps [28] . Submersing the muscle in detergent causes perforations in the cellular membranes , a procedure known as muscle skinning . Due to these perforations , the membrane channels and pumps no longer regulate intracellular ion concentrations . Thus , it is assumed that when a skinned muscle cell preparation is submerged in a solution containing ions , the intracellular ion concentrations are equal to that of the immersing solution . Thus , the intracellular [Ca2+] can be manipulated , and the muscle will contract and generate tension in response . Data points published by Dobesh and colleagues ( Fig 2A of the original article ) [18] provided the experimentally measured tension values ( Sexp ) at various SL’s and [Ca2+]’s . We tested our model at a range of [Ca2+]’s between 0 . 001 and 10 μM for five SL’s ranging from 1 . 85 to 2 . 25 μm . Single overlap length xmax increases with SL according to the formulation of Rice et al [22] , the equations of which are presented in the S1 Appendix ( Eq A1-A3 ) . The parameters Cs , Cf and KCaTn0 were varied so that the sum of the squared differences between the experimental data by Dobesh and the solution of the differential equation ( Eq 9 ) was minimal . The minimization was performed on each individual curve . The Hill-type model with parameters EC50 , Smax , and nH was fitted to the same experimental data . For the MechChem and Hill-type models , the residual errors were assessed to find systematic differences between model and measurement . Additionally , the root mean squared error ( RMSE ) was calculated for all individual curves ( Eq 11 ) . Smodel and SDobesh represent the tension generated in the model and the tension measured experimentally by Dobesh et al . , respectively . Because the tension values reported by Dobesh et al . referred solely to the active tension generated , the results of the MechChem model are also reported as active tension . Additionally , we calculate the tension solely in the thin filament . Hence , the passive tension component contributed by the extracellular matrix or titin is viewed as a separate component that would be additional to the calculated active tension . The number of points is represented by j . The RMSE compares the results of our model to the data from the experiments of Dobesh et al . [18] . In the MechChem model , tension was obtained by solving Eq 9 numerically in Matlab ( MathWorks , Natick , MA ) with the ode23 solver .
The experimental data of Dobesh and colleagues in skinned cardiac muscle ( Table 2 and Fig 2A of the original article ) [18] is compared to the MechChem ( Fig 4A ) and the Hill ( Fig 4B ) model-generated [Ca2+]-tension relationships in Fig 4 . In Fig 4A best-fit curves are shown , as obtained by the MechChem model after proper adjustment of the three parameters KTnCa0 , Cf and Cs per curve . Similarly , in Fig 4B , the curves are best-fit simulations , obtained by the Hill-type model , after proper adjustment of parameters EC50 , nH and Smax . The lower panels , 4C and 4D , show the fitting errors per curve as a function of the [Ca2+] for the MechChem and Hill-type model , respectively . The black lines indicate the median value of the error . For the MechChem model , there is no clear common pattern , and the error does not significantly differ from zero . For the Hill-type model , the error values show a clear common pattern , showing a consistent underestimation at the beginning of the upslope and the very last data points and a consistent overestimation at the location where the curves bend towards the maximum value . Thus , in contrast with the MechChem model , the shape of the tension curve , as simulated with the Hill-type model , is clearly different from the measured data . The MechChem curves are asymmetric with a relatively sharp upward bend for low [Ca2+] and a moderate bending toward the saturation level . In Table 1 , the triplets of parameter values for the simulated curves are shown per sarcomere length for both the MechChem model and the fits according to the Hill-type model . The best-fit values we obtained for the Hill-type model were similar to those obtained by Dobesh et al . [18] . The differences we report are likely a result of either a different fitting routine or slight errors when the data were extracted from a published figure . On average , the root mean square error RSME for the MechChem model is smaller than that for the Hill-type model . The parameter Cf varied by a maximum of 6 . 5% at different SL’s , whereas KTnCa0 and Cs decreased by 43 . 2% and 70 . 6% , respectively . Interestingly , for the MechChem model , parameter values KTnCa0 and Cs exhibit strong interdependency . Our analyses showed that for a change in KTnCa0 by a factor of a ( a>1 ) , Cs changed according to Eq 12 while exhibiting little influence on the resulting [Ca2+]-tension relationship . According to the abovementioned interdependency , it is possible that either KTnCa0 or Cs changes with SL . We assume that KTnCa0 is the parameter that changes with SL because the necessary increase in KTnCa0 is much less than that in Cs . Additionally , a large body of research has shown that the sensitivity of Tn to bind Ca2+ changes with SL [18 , 29 , 30] . Therefore , we performed a best fit by varying KTnCa0 while keeping Cf and Cs fixed at their average values . The resulting fit is shown in Fig 5 . As shown in Table 1 , the fit is nearly as good as the fit by variation of all three parameters . For the Hill-type model , all parameters vary with sarcomere length , albeit that cooperativity only slightly diminishes with sarcomere length . In the results , obtained with the MechChem model , cooperativity is apparent from the steep upslope of the [Ca2+]-tension relationship ( Fig 4A ) , i . e . the sharp upward bend in the [Ca2+]-tension relationship that begins at [Ca2+] of about 2 . 5 μM . With increasing SL , there is an increase in both maximum tension development and slope in both the model-generated and experimental [Ca2+]-tension relationships . Furthermore , the [Ca2+] required for half maximum tension development ( EC50 ) decreases with longer SL’s indicating the increasing sensitivity of tension development in the thin filament to Ca2+ . The experiments of Dobesh et al . [18] showed a decrease in EC50 from 4 . 28 μM to 3 . 16 μM with SL increasing from 1 . 85 μm to 2 . 25 μm ( as determined by the Hill fit in our analyses ) ( Fig 4B , Table 1 ) . In the MechChem simulations covering the same SL range , EC50 decreased from 4 . 18 μM to 3 . 17 μM when fitting the model to the [Ca2+]-tension curves per SL ( Fig 4A ) . Fig 6A and 6B show the MechChem simulation of tension and degree of activation along the thin filament in the single overlap zone , respectively , at different [Ca2+]’s ( 3 , 4 , 5 μM with SL = 2 . 05 μM ) . These values were chosen to show the results in the middle range of the experimental conditions . The MechChem model predicts that the density of bound XB’s increases with the position x along the thin filament . The plateaus shown in Fig 6B represent the full activation of Tn’s , i . e . all binding sites on that section of the thin filament are exposed for XB binding .
Considering the MechChem simulation with parameters Cf and Cs fixed and KTnCa0 depending on SL ( Fig 5 , Table 1 ) , the model-generated [Ca2+]-tension curves resembled the experimental data of Dobesh et al . [18] . Typically , the log[Ca2+]- tension curve is S-shaped from zero tension at low [Ca2+] to a saturated value at high [Ca2+] . The steepness of the slopes in the experimental data of Dobesh et al . [18] indicates cooperativity with a Hill coefficient around 7 . At low [Ca2+] , the [Ca2+]- tension bends up sharply , giving rise to the steepest part of the curve , indicating a high [Ca2+] sensitivity . With maximum [Ca2+] ( 10 μM ) and SL increasing from 1 . 85 μm to 2 . 25 μm , the simulated peak tension increased by 49% and the measured value by 47% . Simulations also produced an increase in Ca2+ sensitivity , shown by a decrease of KCaTn0 from 16 . 6 down to 11 . 8 μM , implying a leftward shift of the [Ca2+]-tension curves with increasing SL . The shape of the MechChem curves ( Figs 4A and 5 ) covers the experimental data better than the sigmoid Hill-curves ( Fig 4B ) , as shown by the residuals in Fig 4C and 4D , especially for the mid and high [Ca2+] . It should be noted that for low [Ca2+] the residuals of the MechChem curves are somewhat larger , probably because of the steep gradient of that part of the curve . The discrepancy between the results of the MechChem model at low [Ca2+] and the experimental results may be further explained by assuming some dispersion in sarcomere length and in the length of the thin and thick filaments , causing the sharp uprising bend to be smeared out , thus moderating the slope in that part of the curve . Cooperativity is assumed to have two components . The first is the intrinsic cooperativity related to binding of Ca2+ to Tn in absence of mechanical tension . The related intrinsic cooperativity coefficient is set to 3 according to the findings by Sun et al . [20] . Secondly , the mechanochemical component boosts the cooperativity from the intrinsic coefficient of 3 up to the experimentally measured value of about 7 . Sun et al . [20] inhibited tension development in trabeculae with blebbistatin and found that in the absence of tension the Ca-dependent activation of Tn’s along the thin filament remained cooperative with a Hill coefficient nH = 3 . The Hill coefficient indicates the minimum number of binding sites that cooperate . Each Tn contains 3 binding sites for Ca2+ [20] , but they do not all have the same affinity and are not likely to fully cooperate . Galińska-Rakoczy et al . [17] showed that the C-terminal section of TnI touches the Tm on the adjacent actin strand , thus coupling both Tn’s , forming an in-register pair Tn2 . Consequently , for each Tn2 , there are 6 binding sites available for inherent cooperation . Apparently , cooperativity is partial , resulting in n ≈3 . By introduction of the MechChem model of tension-driven cooperativity , we propose that mechanical tension in the thin filament results in the strengthening of the binding between Ca2+ and Tn , thereby hindering deactivation , implying that tension lowers the energy associated with Ca2+ being bound to Tn . Combining intrinsic cooperativity with the mechanochemical mechanism results in the MechChem model , which results display a striking resemblance to the experimentally measured tension as a function of [Ca2+] in skinned muscle preparations ( Figs 4 and 5 ) . Many correlations between change of Ca2+ binding properties and mechanical events have been reported . Rieck et al . [31] showed that even in the absence of tension in the thin filament , the formation of strong XB’s stabilized the open conformation of the Tn . Since XB’s can only form if a nearby Tn is in the unblocked state , that Tn may help to unblock the in-register Tn , thus enhancing the formation of additional XB’s attached to the paired actin strand . Furthermore , Isambert et al . [32] showed that the rigidity of the thin filament decreased when Tm was in the unblocked state as compared to being in the blocked state . This finding shows that there is a mechanical coupling between the conformational changes of Tn and elastic properties of the thin filament . More recently , Desai et al . [33] directly showed that myosin binding was necessary for complete activation of the thin filament upon partial activation due to calcium binding . The MechChem simulated best-fit [Ca2+]-tension curve was determined for intrinsic cooperativity coefficients n = 1 , 2 , 3 and shown in the supplementary material ( S1 Fig ) . The fit is excellent when n = 3 , while for n = 1 or 2 , the curves are not as steep as in the physiological situation . In the experiments by Sun et al . , development of mechanical tension was shown not to have an effect on cooperativity , i . e . the Hill coefficient remained as low as 3 . As they mentioned in their article [20] , they could not exclude the possibility that the applied fluorescent probes attached to the various structures of the Tn complex may change the properties of Tn to some degree . We think that these probes may inhibit the mechanochemical enhancement of Ca2+ affinity by tension . The [Ca2+]-tension relation in skinned muscle has generally been characterized by a modified Hill curve that takes on a symmetric s-shape [34] . The [Ca2+]-tension relation simulated by our model is asymmetric ( Fig 4A and 4C ) ; the initial steep rise in tension generated with additional Ca2+ decreases closer to saturation . Whereas Dobesh et al . fitted symmetric curves , they admitted that the Hill curves consistently overestimated the tension developed in the sarcomeres as the curves rounded toward saturation ( Fig 4B and 4D ) . In an attempt to account for the asymmetry , Dobesh et al . [18] proposed that the experimental data be fit to 2 Hill coefficients that meet at EC50 for each curve . In the Mech Chem model , the [Ca2+]-tension relation is already asymmetric , so the introduction of an additional parameter is not necessary to reshape the curve . Both the Hill-type model and the MechChem model require 3 parameters per curve . By fitting the models to the experimental data for each sarcomere length separately , in the Hill-type model all 3 parameters appeared to depend on SL , albeit that dependency of the Hill coefficient nH appeared weak ( Table 1 ) . In the MechChem model , parameters Cf and Cs did not depend clearly on SL , thus hinting us to keep these parameter values fixed , while only reestimating the equilibrium constant KCaTn0 , expressing Ca2+ affinity in absence of mechanical tension . The resulting fits were nearly as good ( Fig 5 ) , suggesting that Ca2+ affinity apparently depends somehow on SL , while the other parameters were general to all SL’s . Our model comprises two mechanisms of cooperativity that generate the [Ca2+]-tension relationship , i . e . tension in Tm strengthening the bond between Ca2+ and Tn and intrinsic cooperative activation of in-register pairs of Tn . The Hill function is based on common chemical equilibrium and provides the basis for many current models of Ca2+-Tn binding . For example , Rice et al . [22] utilize a modified Hill function to model the transition between the active and inactive states of Tn . The peak intracellular [Ca2+] reaches 1 . 45 μM in the Rice model , a concentration that the Hill function mimics well . The Hill model does not provide a physical explanation for the mechanism of cooperativity in cardiac muscle , but instead utilizes a coefficient that characterizes the steepness of the [Ca2+]-tension relation . Additionally , the Hill model was initially developed to understand the cooperative binding of oxygen to hemoglobin [34] , where it makes physical sense that a first chemical binding of oxygen on a hemoglobin molecule will facilitate subsequent bindings because oxygen-binding sites on hemoglobin are separated by only 2 . 5 to 3 . 5 nm [35] . To reach the high physiologic level of cooperativity , several Tn’s must interact . Because Tn’s are separated by about 35 nm along a single actin strand within the thin filament , it is physically difficult to explain that a chemical binding can influence the binding of a different molecule that far away by conventional chemical principles . It has been shown that the sensitivity of Tn to bind Ca2+ changes with SL [18 , 29] , a characteristic implemented in the MechChem model . There is currently no clear consensus regarding the mechanism behind length dependent Ca2+ sensitivity in cardiac muscle ( for review , [36] ) . One proposed mechanism is that the lattice spacing decreases with longer SL , moving myosin heads closer to the thin filament and rendering the thin filament more sensitive to Ca2+ by enhancing binding [37] . However , the lattice spacing hypothesis has been questioned after it has been shown that muscle length does not necessarily correlate with myofilament spacing [30] . It has also been proposed that phosphorylation of sites on TnI by protein kinase A and protein kinase C alters the Ca2+ sensitivity and could prove significant in the regulation of length dependent activation [14] . Conversely , Lee et al . have shown that the increase of passive tension in titin leads to increased Ca2+ sensitivity [38] . It is possible that the change in KTnCa0 in the MechChem model is due to a combination of the abovementioned mechanisms . As already indicated in the introduction , various models have been developed on the basis of nearest neighbor cooperativity , based on RU-RU , XB-XB or XB-RU interactions . These models generally result in Hill-type [Ca2+]-tension relations that fit accurately to experimental data . However , the curves are slightly , but systematically different from the MechChem curves ( Fig 4 ) . It is not clear yet if these differences are sufficiently strong to make a choice between the two model types . We predict a clear difference to be expected , yet we do not currently have the means to test the model . In a tension bearing sarcomere , the MechChem model predicts that the hindrance to Tn deactivation imposed by high tension in the thin filament causes a higher concentration of activated Tn’s , bound Ca2+ , and XB’s toward the Z-disk end of the single overlap region where tension is highest ( Fig 5B ) . With the nearest neighbor hypothesis , no preference is to be expected on the location of activated Tn’s . However , although not explicitly noted , the MechChem model does take into account some of the nearest neighbor cooperativity mechanisms . The development of tension in the thin filament begets more tension development . Thus , XB’s recruit more XB’s . Additionally , the tension in the thin filament at position x strongly determines whether the Tn is active at point x ( XB-RU cooperativity ) . RU-RU cooperativity is accounted for in the MechChem model through the intrinsic cooperativity coefficient n . Most of the models previously discussed have included only local cooperativity mechanisms , yet there are also models and hypotheses that incorporate cooperative mechanisms acting along the entirety of the thin filament . Brandt and colleagues [39] proposed that the mutual overlap of Tm molecules under Tn causes a simultaneous unblocking or blocking of all Tm molecules along the filament . Conversely , the “cooperative realignment of binding sites” , a model developed by Daniel and colleagues [26] , ignores the effects of Tm but looks instead at the possible impact of strain on the thin filament . Hence , this model can predict cooperativity in tension development but not in activation . Still others attribute the cooperative effect to the constant volume property of the myofiber matrix . The stretch of sarcomeres causes the thick and thin filaments to squeeze closer together increasing the probability of XB binding [24] . Like the cooperative realignment of binding sites , the lattice spacing hypothesis can account for cooperative tension generation but not activation . The model designed by Land & Niederer [40] represented the entire thin filament , a compilation of 26 RU’s . They hypothesized that the state ( position ) of the Tm molecule in each RU has a corresponding free energy determined by the state of neighboring XB’s and RU’s . The energy term is used to compute the probability of a RU being blocked or unblocked . While the simulated results are consistent with available experimental data and insight can be gained through the Land & Niederer model , it is composed of a system of 750 ordinary differential equations . The tension-driven cooperativity model we present consists of a single ordinary differential equation and three key parameters , so we present a highly simplified model that still captures behaviors shown in available experimental data . Our model , like the Land & Niederer [40] model , includes an energy term in the computations of state P . It is a mechanical energy term related to the tension within the thin filament that increases along the thin filament from the start of the single overlap region towards the Z-disk . Our model is different from others because the mechanics of the thin filament ( S ( x ) ) directly impact the related chemistry ( Ca2+ binding to Tn ) whereas most of the other models view the mechanics and chemistry separately . Izakov et al . have hypothesized that the number of strongly bound XB’s along the thin filament affects the binding affinity of Ca2+ to Tn and have implemented this idea in a computational model [41] . Landesberg and Sideman proposed a similar model , but it was a loosely coupled model meaning that Ca2+ was not required to remain bound for the Tn to remain active , but bound XB’s were adequate to do so [42] . Our model differs because we propose that the tension in the thin filament contributes to an increment energy required to unbind Ca2+ from Tn . This energy increases along the thin filament toward the Z-disk as tension increases . In the models discussed above , the global affinity for Ca2+ to bind to Tn will change based on the number of strong XB’s . Within our model , however , the affinities increase with tension in the thin filament , being explained by a linear increase of binding energy with tension in the thin filament . Due to the interaction between mechanics and chemistry within the model , neither mechanics nor chemistry fully account for the activation or subsequent deactivation of the thin filament . While we propose that high tension in the thin filament hinders deactivation , a decrease in intracellular [Ca2+] will trigger relaxation . Due to the buildup of tension within the thin filament as shown in Fig 3 , there is always a loose end closer to the mid-line in which little tension is developed . The hindrance to deactivation imposed by high tension does not exist in these areas , thus promoting the deactivation of the Tn’s closest to the mid-line first . We propose that the areas of highest tension ( closest to the z-disk ) are the latest to deactivate . Our model has provided a potentially experimentally testable hypothesis . Model results suggest directionality in the dispersion of XB’s along the thin filament with a higher concentration of XB’s closer to the Z-disk . Desai and colleagues [33] were recently able to fluorescently label single myosin heads and observe single bindings to the thin filament . They observed that although myosin binding activated the RU , there was no directionality in the binding of XB’s . However , tension was not developed in this model because the myosin heads were not tethered to the thick filament . Additional experimental evidence is needed to test the hypothesis of XB dispersion in the loaded thin filament . The difficulty lies in developing an experimental approach that enables viewing of each individual XB binding in a skinned muscle under tension . The model presented here assumes that when an area on the thin filament becomes unblocked , XB’s are automatically formed . Thus , the fraction of unblocking is proportional to the XB- force developed along the strand . The assumption that XB’s form automatically when RU’s are activated is not physiologically accurate . We expect this to be corrected upon the explicit incorporation of the different steps of the XB cycle . However , the experimentally measured tension in skinned muscle preparations is obtained after the steady state has been reached . Hence , the experimental data is time-independent . The current model results are limited to the [Ca2+]- tension relationship . The introduction of dynamics such as the XB cycle requires additional assumptions to be made and additional unknown parameters added . However , the next logical step for the model is the implementation of the XB cycle that will be studied in the isometric twitch . The MechChem model is a mean field approximation , so it is highly simplified . Specific spatial details such as individual binding sites are not accounted for , but the simplification reduces computational cost considerably . A novel mechanochemical model of tension generation by the sarcomere has been developed based on long range cooperativity imposed by mechanical tension in the thin filament and intrinsic cooperativity resulting from the interaction between the calcium-binding sites on the in-register troponin complexes . Simulated [Ca2+]-tension curves resembled those obtained in steady state isometric muscle experiments . Thus , our results support the hypothesis that high tension in the thin filament impedes deactivation by increasing the energy required to detach calcium from the troponin complex . Furthermore , we found that the tension in the thin filament was relatively low toward the beginning of the single overlap region close to the mid-line of the sarcomere but increased steeply in the overlap region closer to the Z-disk . Model simulations suggest that the concentration of calcium bindings to the troponin complexes and active XB’s are low at the free end of the thin filament and saturated closer to the Z-disk . Future experimental studies are needed to test the latter property , indicating the validity of our hypothesis on the cooperative effect of tension in the thin filament on force generation by the cardiac sarcomere .
|
Cardiac contraction is the culmination of multiple subcellular processes beginning with calcium induced activation of the contractile machinery . Interestingly , small increases in intracellular calcium concentrations lead to disproportionately large increases in tension development within the cardiac muscle , a phenomenon known as ‘cooperative activation’ . Although the concept of cooperative activation is widely accepted , the mechanism is highly debated . Many complex computational models have been developed in an attempt to understand the underlying mechanisms . However , no single mechanism has been able to properly account for the range of experimental data . We propose that the interaction between the mechanics and chemistry in the contractile unit is an essential component of the cooperative activation . We have developed a simple computational model composed of only five parameters and a single ordinary differential equation that describes the cooperative nature of myocardial contraction through mechanochemical interaction . This model can reproduce experimental data relating to cooperative activation and provides a promising tool for future research .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"myofibrils",
"classical",
"mechanics",
"muscle",
"tissue",
"muscle",
"contraction",
"molecular",
"motors",
"actin",
"motors",
"chemical",
"equilibrium",
"motor",
"proteins",
"cardiac",
"muscles",
"research",
"and",
"analysis",
"methods",
"physical",
"chemistry",
"mechanical",
"tension",
"muscle",
"physiology",
"contractile",
"proteins",
"curve",
"fitting",
"musculoskeletal",
"system",
"mathematical",
"functions",
"animal",
"cells",
"proteins",
"troponin",
"mathematical",
"and",
"statistical",
"techniques",
"biological",
"tissue",
"chemistry",
"muscles",
"muscle",
"cells",
"physics",
"biochemistry",
"cytoskeletal",
"proteins",
"sarcomeres",
"cell",
"biology",
"anatomy",
"physiology",
"myosins",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences"
] |
2016
|
Mechano-chemical Interactions in Cardiac Sarcomere Contraction: A Computational Modeling Study
|
The obligate intracellular parasite Toxoplasma gondii exploits cells of the immune system to disseminate . Upon T . gondii-infection , γ–aminobutyric acid ( GABA ) /GABAA receptor signaling triggers a hypermigratory phenotype in dendritic cells ( DCs ) by unknown signal transduction pathways . Here , we demonstrate that calcium ( Ca2+ ) signaling in DCs is indispensable for T . gondii-induced DC hypermotility and transmigration in vitro . We report that activation of GABAA receptors by GABA induces transient Ca2+ entry in DCs . Murine bone marrow-derived DCs preferentially expressed the L-type voltage-dependent Ca2+ channel ( VDCC ) subtype Cav1 . 3 . Silencing of Cav1 . 3 by short hairpin RNA or selective pharmacological antagonism of VDCCs abolished the Toxoplasma-induced hypermigratory phenotype . In a mouse model of toxoplasmosis , VDCC inhibition of adoptively transferred Toxoplasma-infected DCs delayed the appearance of cell-associated parasites in the blood circulation and reduced parasite dissemination to target organs . The present data establish that T . gondii-induced hypermigration of DCs requires signaling via VDCCs and that Ca2+ acts as a second messenger to GABAergic signaling via the VDCC Cav1 . 3 . The findings define a novel motility-related signaling axis in DCs and unveil that interneurons and DCs share common GABAergic motogenic pathways . T . gondii employs GABAergic non-canonical pathways to induce host cell migration and facilitate dissemination .
The obligate intracellular parasite Toxoplasma gondii chronically infects a large portion of the global human population and is capable of infecting any warm-blooded vertebrate [1] . The dissemination of the parasite from the point of entry in the intestinal tract plays a determinant role in the pathogenesis of toxoplasmosis . Although chronic infection is generally considered asymptomatic in otherwise healthy individuals , reactivated infection in the central nervous system ( CNS ) of immune-compromised individuals may be fatal . Congenital toxoplasmosis occurs by transmission to the fetus from the infected mother and can result in serious disabilities or death of the unborn child [2] . Previous studies have demonstrated that active invasion of dendritic cells ( DCs ) by T . gondii tachyzoites rapidly ( within minutes ) induces a hypermigratory phenotype in parasitized DCs [3] . This migratory activation is characterized by cytoskeletal rearrangements and dramatically enhanced cellular locomotion , termed hypermotility [4] , and enhanced transmigratory activity in vitro [5] . These phenotypes have been linked to enhanced dissemination and parasitic loads in mice for different species of apicomplexan parasites [5–7] . The initiation of the hypermigratory phenotype in DCs is related to the discharge of secretory organelles during parasite invasion and does not depend on de novo protein synthesis in the host cell [4] . It is mediated through non-canonical GABAergic signaling pathways , and is independent of MyD88-mediated TLR signaling and chemotaxis [3–5 , 8] . Within the context of the host-parasite interaction , we have recently shown that DCs possess functional GABAA receptors , and the capability to synthesize and secrete γ–aminobutyric acid ( GABA ) [8] . Challenge with T . gondii triggered GABA secretion in the invaded DCs and inhibition of GABAA receptors , GABA synthesis or GABA transport abrogated the T . gondii-induced hypermigratory phenotype [8] . Along these lines , mounting evidence shows that GABA , the main inhibitory neurotransmitter in the vertebrate brain , participates outside the CNS in diverse functions including cell migration , immunomodulation and metastasis [9–11] . GABAA receptors are ionotropic chloride channels whose functions are regulated by cation-chloride co-transporters [12] . Membrane depolarization secondary to GABA receptor activation can elicit opening of voltage-dependent Ca2+ channels ( VDCCs , also termed voltage-gated Ca2+ channels , VGCCs ) that are normally closed at physiologic or resting membrane potential [13–15] . Thus , GABA-mediated Ca2+ influx via VDCCs is a well-established concept in neuronal cells but remains unexplored in immune cells . While various Ca2+ signaling pathways have been implicated in the regulation of multiple DC functions , including activation , maturation and formation of immunological synapses with T cells ( reviewed in [16] ) , knowledge on the role of VDCCs in DCs remains limited [17 , 18] . Here , we show how an obligate intracellular pathogen takes advantage of a hitherto uncharacterized Ca2+ signaling axis in DCs to modulate the migration of parasitized host cells . We demonstrate that the hypermigratory phenotype induced in DCs by T . gondii is predominantly dependent on the L-type VDCC subtype Cav1 . 3 , which is activated by GABAergic signaling upon T . gondii invasion .
Shortly after parasite entry , DCs exhibit a dramatic migratory activation [5 , 8] . Based on the implication of GABAergic signaling and on the rapid onset of the hypermigratory phenotype minutes after parasite invasion [4] , we hypothesized a role for Ca2+ signal transduction for the induction and maintenance of T . gondii-induced migration . When DC motility was assessed in a Ca2+-deprived medium , individual cell track analysis of infected DCs showed a reduction in migrated distances ( Fig 1A ) and a significant downward shift in the distribution of migrated distances ( Fig 1C and 1D ) . In low Ca2+ medium with 1% FBS , the median velocity of T . gondii-infected DCs was significantly reduced , and addition of Ca2+ ( CaCl2 ) at physiological concentration reconstituted hypermotility in infected DCs ( Fig 1B ) . Similarly , base-line motility of non-challenged DCs was reduced upon Ca2+ deprivation and reconstituted by addition of Ca2+ ( Fig 1B and S1 Fig ) . In line with motility assays , the relative transmigration frequencies of infected DCs , and non-infected DCs , across a transwell porous-membrane were significantly reduced in low extracellular Ca2+ and reconstituted upon addition of Ca2+ at physiological concentration ( Fig 1E ) . Next , we assessed the motility of infected DCs in the presence of NiCl2 , which blocks plasma membrane Ca2+ channels . NiCl2 dose-dependently reduced the velocity of hypermotile infected DCs , that reached velocities comparable with non-infected DCs ( Fig 1F ) . Altogether , the present data indicate that the T . gondii-induced hypermigratory phenotype of DCs is dependent on the entry of extracellular Ca2+ through plasma membrane Ca2+ channels . We have previously established that infection by T . gondii induces motility-related GABAergic signaling pathways in DCs [8] . Because hypermotile Toxoplasma-infected DCs exhibited dependency on Ca2+ and the established links between GABA receptor activation and Ca2+ responses in neuronal cellular systems [13 , 14] , we tested whether GABAA receptor activation led to Ca2+ responses in DCs . Perfusion of GABA elicited cytosolic Ca2+ elevations in DCs , visualized by fluorescent Ca2+ indicators ( Fig 2A and S1 Video ) . Stimulation of DCs with GABA led to a simultaneous and transient Ca2+ influx ( Fig 2B and 2C ) in ~ 20% of the tested DC population at a given time point and , for the reference stimulus ATP , in ~ 42% of DCs ( S1 Table ) . Ca2+ transients induced by GABA had relatively similar longevity and relatively lower amplitude than responses to ATP ( Fig 2B and 2C ) , which were in line with ATP responses previously characterized in various types of DCs [19 , 20] . Upon repeated stimulations with GABA and at varying GABA concentrations , consecutive Ca2+ responses were observed in individual cells ( S2 Fig ) . Altogether , the data is in line with the previously recorded GABA-induced membrane potential changes by patch-clamping [8] and demonstrate that GABA stimulation of DCs is followed by influx of Ca2+ and transiently increased cytosolic Ca2+ concentration . Next , we sought to determine if the GABA-induced Ca2+ signaling in Toxoplasma-infected DCs had an impact on hypermotility . First , we determined GABA secretion by infected DCs and the deprivation of GABA upon pre-incubation with GABAergic inhibitors ( SC/SNAP ) using MALDI mass spectrometry analysis of cell supernatants . In supernatants from Toxoplasma-infected DCs , spectra displayed a distinct peak signal ( m/z 104 , 2; Fig 3A ) corresponding to the signal of protonated GABA [M + H]+ chemical grade analytical standard ( S3 Fig ) [21] . Inhibition of GABA synthesis and secretion ( SC , SNAP inhibitors , respectively ) selectively reduced the m/z 104 , 2 peak signal ( Fig 3A ) and abrogated the hypermotility of T . gondii-infected DCs , which was reconstituted by addition of exogenous GABA ( Fig 3B and 3C ) . This provided further specificity to previously reported elevations of GABA secretion in Toxoplasma-infected DCs , as quantified by GABA-ELISA under the same conditions [8] . Next , to test the impact of Ca2+ influx in DC hypermotility under GABA-deprived conditions , a cell membrane Ca2+ channel/ L-type VDCC agonist ( BayK8644 ) was added to the cells . Importantly , the abrogated hypermotility of infected DCs , generated by GABAergic inhibition , was rescued by addition of BayK8644 ( Fig 3B and 3C ) . A moderate but significant increase in cell motility was also observed in naïve DCs in presence of BayK8644 ( Fig 3B and 3C ) . We conclude that , upon GABAergic inhibition , Ca2+ channel ( VDCC ) agonism leading to Ca2+ entry in DCs can reconstitute hypermotility in Toxoplasma-infected DCs . VDCCs are known to respond with Ca2+ permeability to membrane potential changes . Because GABAA receptor activation by GABA elicits membrane potential changes in Toxoplasma-infected DCs [8] and GABA elicited Ca2+ influx ( Fig 2 ) , we investigated the putative involvement of VDCCs in DC hypermotility . L-type VDCC inhibition by nifedipine abolished hypermotility ( Fig 4A ) and significantly reduced transmigration ( Fig 4B ) . In sharp contrast , inhibition of purinergic Ca2+ receptors by PPADS at high concentrations [22] had non-significant effects on hypermotility and transmigration of infected DCs ( Fig 4A and 4B ) , despite that activation of purinergic receptors by ATP caused a significant Ca2+ influx and increased cytosolic Ca2+ levels in DCs ( Fig 2 ) . This indicated that VDCC-related effects governed hypermigration . We therefore explored further the function of VDCCs in relation to GABAergic signaling . We previously reported that inhibition of GABA synthesis ( SC ) and/or transport ( SNAP ) significantly reduced GABA secretion and transmigration of T . gondii-infected DCs [8] . Extending these observations , addition of exogenous GABA rescued the hypermotility of infected DCs under GABAergic inhibition ( Fig 4C and 4D ) . In sharp contrast , VDCC inhibition by nifedipine treatment caused a significant decrease in the motility of infected DCs that was not restored by exogenous GABA ( Fig 4C and 4D ) , indicating implication of L-type VDCCs downstream of GABAergic signaling . At resting membrane potential VDCCs are normally closed and , respond with Ca2+ permeability upon membrane depolarization . To relate the effect of GABAergic signaling to that of membrane depolarization , we treated GABA-deprived infected DCs with the depolarizing agent KCl . Upon blockade of GABA synthesis and secretion , KCl treatment fully restored hypermotility in Toxoplasma-infected DCs ( Fig 4E ) , thus mimicking the effects obtained by addition of exogenous GABA ( Figs 4D and 3C ) . Importantly , hypermotility was not restored by KCl in the presence of the L-type VDCC inhibitor nifedipine ( Fig 4E ) . Taken together with the effects of VDCC agonism ( Fig 3 ) , these data demonstrate a link between L-type VDCCs and the hypermigratory phenotype of T . gondii-infected DCs downstream of GABAergic signaling . In order to determine putative VDCCs mediating the nifedipine-sensitive GABA reconstitution effect , we performed a screen of VDCCs expressed in DCs . RT-PCR analyses indicated transcriptional expression of the L-type VDCC Cav1 . 3 in DCs , similar to brain homogenate ( Fig 5A ) . A real-time quantitative PCR ( qPCR ) screen of VDCCs confirmed a consistent high relative expression of Cav1 . 3 transcripts in 6 mice tested over time , and also less abundant relative expression of Cav2 . 2 ( Fig 5B and 5C ) . Other VDCC types , e . g . Cav1 . 1 , Cav1 . 4 , Cav2 . 1 , Cav3 . 1 , exhibited low , undetectable or inconsistent relative expression ( Fig 5B and 5C ) . In Toxoplasma-challenged DCs , Cav1 . 3 remained the predominantly expressed VDCC type over other types ( Fig 5D and S4A Fig ) and maintained transcriptional expression of Cav1 . 3 in Toxoplasma-infected DCs related to non-challenged DCs was observed during 24 h infection ( S4B Fig ) . Western blot analyses detected polypeptides ( ≈ 250 kDa ) in DCs , corresponding to Cav1 . 3 expression as previously characterized in primary astrocytes [23] , and with similar relative expression in DCs and Toxoplasma-infected DCs ( Fig 5E and S4C Fig ) . Immunocytochemistry using a mAb to a predicted sub-membranous Cav1 . 3 epitope yielded a distinct fluorescence signal in non-infected and in infected permeabilized DCs ( Fig 5F ) . Altogether , we conclude that Cav1 . 3 was the predominantly expressed VDCC in murine bone marrow-derived DCs and that the relative VDCC expression profile varied between mice or varied over time . Upon Toxoplasma-infection , Cav1 . 3 remains the predominant transcriptionally expressed VDCC . To functionally assess the relative contribution of Cav1 . 3 to hypermigration in relation to other putatively expressed VDCCs , we took advantage of a pharmacological antagonist with high specificity for Cav1 . 3 , CPCPT [24] , and a broad inhibitor of L , N and T type VDCCs , benidipine [25] . Both inhibitors similarly abolished the hypermotility of infected DCs ( Fig 5G and S4D Fig ) . While CPCPT significantly reduced transmigration of DCs from different mice , benidipine was a more consistent abrogator of transmigration ( Fig 5H ) . Jointly , these data suggest that VDCCs play a significant role in T . gondii-induced hypermotility of DCs . As Cav1 . 3 appeared to be the most abundantly expressed VDCC , these data suggested that CPCPT and benidipine might act primarily on Cav1 . 3 . To test the functional implication of Cav1 . 3 in T . gondii-induced hypermotility , we employed an RNA interference approach . First , transduction efficacy by the recombinant lentiviral vector was optimized in the murine neuroectodermal cell line NE-4C and in primary DCs ( S5 Fig ) . Cav1 . 3 ( shCav1 . 3 ) and Cav1 . 2 ( shCav1 . 2 ) were successfully targeted in NE-4C cells by this approach ( S6A , S6B and S6C Fig ) . Similarly , in primary DCs , shRNA targeting Cav1 . 3 ( shCav1 . 3 ) , Cav1 . 2 ( shCav1 . 2 ) or control shRNA ( shLuc ) was delivered and the transduced DCs were challenged with T . gondii tachyzoites ( Fig 6A ) . DCs transduced with shCav1 . 3 exhibited significantly reduced Cav1 . 3 mRNA expression , with non-significant effects on Cav1 . 3 mRNA expression by shCav1 . 2 and control shRNA ( Fig 6B ) . Western blotting analyses of DCs and NE-4C cells transduced with shCav1 . 3 showed a reduction in Cav1 . 3 protein expression ( Fig 6C , S6D Fig ) . Because primary DCs may become activated by the lentivirus and activation may impact on motility , we assessed expression of IL-12 mRNA in primary DCs and the NE-4C line . While the expression of IL-12 mRNA was relatively unaffected in NE-4C cells , primary DCs exhibited enhanced expression of IL-12 mRNA upon lentiviral transduction , in a similar fashion for shLuc , shCav1 . 2 and shCav1 . 3 ( S7 Fig ) . We conclude that Cav1 . 3 mRNA and protein expression were selectively reduced in DCs exposed to shCav1 . 3 and that lentiviral transduction generates enhanced IL-12 mRNA expression in primary DCs . To assess the impact of Cav1 . 3 silencing on hypermotility , we first optimized the approach using the murine DC line ( JAWS II ) . JAWS II cells and DCs expressed a similar VDCC profile , with Cav1 . 3 as the most prominently expressed VDCC ( S8A Fig ) and a similar inhibitory profile by calcium blockers on hypermotility was observed ( S8B Fig ) . JAWS II transduced with shCav1 . 3 ( S8C Fig ) exhibited significantly reduced Cav1 . 3 mRNA expression and enhanced IL-12 mRNA expression ( S8D and S8E Fig ) . Importantly , shCav1 . 3-tranduced primary DCs ( Fig 6D ) and JAWS II exhibited reduced motility upon Toxoplasma-challenge . Their velocities reached non-significant differences compared with baseline motility of non-infected DCs ( Fig 6E ) and JAWS II , respectively ( S8F Fig ) . Significant differences in the reduction of motility were observed for shCav1 . 3-tranduced DCs compared with shCav1 . 2- , shLuc- and mock-transduced DCs ( Fig 6E ) . In line with results obtained upon pharmacological L-type VDCC inhibition ( Fig 4D ) , exogenous GABA restored motility in mock-treated GABA-inhibited DCs but failed to restore motility in the shCav1 . 3-transduced cells ( Fig 6F ) . We conclude that selective silencing of Cav1 . 3 abolishes T . gondii-induced hypermotility in DCs . We have previously shown that adoptive transfer of T . gondii-infected DCs in mice leads to rapid dissemination of parasites and to exacerbation of the infection compared to infection with free tachyzoites [5 , 26] , and that GABAergic inhibition blocks this exacerbated dissemination [8] . To assess if VDCC inhibition impacted on parasite loads , benidipine pre-treated infected DCs were adoptively transferred to mice intraperitoneally . When the infections were monitored by in vivo bioluminescence , photonic emissions indicated dissemination of parasites to spleen and mesenteric lymph nodes ( MLN ) ( Fig 7A and 7B ) . Plaquing assays of homogenized whole organs ( non-perfused ) revealed overall reduced mean parasite loads in the spleens of mice challenged with infected DCs ( + ) benidipine compared to mice challenged with infected DCs ( - ) benidipine ( Fig 7C ) . To analyze the contribution of parasites in the blood circulation to the total parasite loads in organs , parasite loads were analyzed after blood perfusion . Perfused spleens exhibited overall reduced parasite loads , and reduced or abolished differences in parasite loads between the benidipine-treated and the non-treated conditions ( Fig 7C ) . This showed that both removal of blood and benidipine treatment had a reducing impact on parasite loads in the spleen . In contrast , blood perfusion yielded more discrete relative reductions of parasite loads in MLNs , which are indirectly linked to the blood circulation via the lymphatic system ( Fig 7D ) . In the perfused mice , parasites were consistently detected in all brains from day 4 versus day 3 in non-perfused mice ( Fig 7E ) , indicating a contribution of parasites that were displaceable by blood perfusion to the total parasite loads in non-perfused mice . Because benidipine-treatment had an impact on splenic parasite loads early during infection and this effect appeared linked to parasites in blood , we analyzed the fate of parasites and DCs within 24 h post-inoculation intraperitoneally . Upon benidipine pre-treatment of infected DCs , significantly reduced parasite numbers were measured in spleen by 24 h ( Fig 8A ) , with a non-significant reduction of parasites in blood and non-significant differences in peritoneum ( Fig 8A ) . Similarly , flow cytometry analyses identified reduced numbers of cell-associated GFP-expressing parasites ( GFP+ ) in the spleen upon benidipine treatment ( Fig 8B and S9A Fig ) , in line with the observed differences by plaquing assays and qPCR ( Fig 8A ) . When adoptively transferred infected DCs were pre-labeled with a cell dye ( CMTMR ) , CMTMR+ GFP+ cells were detected in the spleens and also CMTMR- GFP+ cells ( S9B Fig ) . This indicated direct transport to the spleen by infected DCs and also rapid transfer of parasites to new leukocytes in peritoneum and spleen . Benidipine treatment yielded non-significant effects on DC viability , infection frequencies and parasite viability ( S10 Fig ) . Altogether , the data show that adoptively transferred infected DCs rapidly entered the circulation and that VDCC inhibition led to reduced numbers of parasite-associated cells in spleen during the early phase of infection . VDCC inhibition delayed the appearance of parasites in circulation and , thereby , also the systemic dissemination of T . gondii .
In this study we investigated the molecular signaling mechanisms that govern how T . gondii hijacks the migratory properties of DCs . Building on previous work showing that a hypermigratory phenotype sets in within a few minutes after T . gondii invasion of DCs [4] and depends on GABAergic signaling [8] , we addressed the role of Ca2+ signaling in these processes . Our studies establish that Ca2+ signaling in murine bone marrow-derived DCs is indispensable for T . gondii-induced hypermotility and transmigration in vitro . The observation that the onset of the hypermigratory phenotype was abrogated at sub-physiological extracellular Ca2+ concentrations or by blocking plasma membrane Ca2+ channels underpinned a role for membrane-bound Ca2+ channels . However , Ca2+ mediates signal transduction to multiple cellular pathways . It was therefore crucial to determine its putative interaction with the GABAergic system of DCs . We previously showed that inhibition of GABA synthesis , GABA secretion or GABAA receptor blockade in Toxoplasma-infected DCs abolishes hypermigration [8] . Here , we demonstrate that hypermotility and transmigration are restored in GABA-deprived infected DCs by ( i ) addition of exogenous GABA , ( ii ) by cell membrane depolarization with KCl and ( iii ) by L-type VDCC agonism . Consequently , ( iv ) L-type VDCC blockade hindered reconstitution of hypermotility by GABA and KCl . This pinpointed a role for VDCCs downstream of GABAergic signaling . Further , inhibition of purinergic Ca2+ channels ( P2 receptors ) yielded non-significant effects on hypermotility despite a measurable Ca2+ influx in response to ATP . Also , while Ca2+ -deprivation led to a similar proportional reduction of motility of unchallenged DCs ( baseline motility ) and infected DCs ( hypermotility ) , selective L-type VDCC inhibition abolished hypermigration but had non-significant effects on the baseline motility of DCs . Altogether , this indicated that L-type VDCCs primarily mediated the GABA-evoked motility-related Ca2+ influx and that extracellular Ca2+ influx per se into the cell or increased cytosolic Ca2+ levels per se was not sufficient to induce hypermigration . To our knowledge , the findings demonstrate for the first time that murine DCs express the L-type VDCC subtype Cav1 . 3 , with a functional implication in motility . Cav1 . 3 appeared to be the predominant transcriptionally expressed VDCC in primary DCs , a feature also maintained by the DC line JAWS II . Importantly , silencing of Cav1 . 3 by shRNA or selective pharmacological antagonism of Cav1 . 3 abrogated the hypermigratory phenotype in Toxoplasma-infected DCs , while baseline motility and morphology of DCs remained intact related to mock-treated and non-infected DC . A caveat of lentiviral transduction in primary DCs is that the lentivirus vector may have activation effects on the DCs [27] , yet without reported apparent inhibitory effects on functionality [28] , thereof the requirement of appropriate control experiments . We validated and confirmed our results in two additional cell lines . IL-12 mRNA expression indicated activation by the lentiviral vector primarily in DCs , to a lesser extent in JAWS II and , non-significant effects on the NE-4C line . Silencing of Cav1 . 3 expression abolished the hypermigratory phenotype , in contrast to Cav1 . 2 silencing . This , together with its apparent predominant expression , attributes a primary role in Toxoplasma-induced hypermotility to the VDCC subtype Cav1 . 3 . However , despite that we did not observe compensatory up-regulation of other VDCCs/Cav1 . 2 upon Cav1 . 3 silencing in DCs , the data do not exclude a contributive role for other VDCC subtypes . In fact , the neuronal VDCC family members often display overlapping functions in mediating signal transduction [29] . This may apply to VDCCs in murine DCs too , as relative variations in transcription of several VDCC subtypes were detected in different mice over time , yet conserving a relative predominant expression of the subtype Cav1 . 3 . Altogether , the data at hand defines a role for Cav1 . 3 in Toxoplasma-induced DC hypermotility and establish Ca2+ as a second messenger to GABAergic signaling in DCs . We have previously shown that GABA induces GABAA receptor-activated currents in DCs [8] . Here , we demonstrate that DCs can sense membrane voltage changes caused by depolarization ( KCl or GABA ) and can respond to GABA by a Ca2+ transient . The analogous hypermotility restoration effect of exogenous GABA and depolarization by KCl , together with the opposite effects of the VDCC inhibitors ( benidipine , nifedipine , CPCPT ) and the agonist BayK8644 ( a structural analog of nifedipine with positive inotropic activity ) strongly suggests that GABA mediates membrane depolarization-induced Ca2+ release via VDCCs . Also by analogy to findings in neurons [12] , it is plausible that GABAA receptors and chloride homeostasis are regulated by cation-chloride co-transporters in DCs . Altogether , our findings provide evidence of a direct link between GABA receptor signaling , Cav1 . 3 activation and hypermotility . Although modulated by Toxoplasma infection , functional GABAA receptors appear to be constitutively expressed by murine and human DCs [8] . The effects of GABA on Ca2+ signaling via VDCCs / Cav1 . 3 has not been previously addressed in immune cells [30] . However , Ca2+ channels mediate some of the most rapid biological processes described and VDCC signaling allows for immediate cellular responses to external stimuli [31] . This is in agreement with the features attributed to the hypermigratory phenotype of Toxoplasma-infected DCs [3]: for example , its rapid onset , cytoskeletal remodeling and switch to amoeboid-type of migration within minutes after T . gondii invasion of the host DC in vitro [4] and is also in line with the observed rapid migration of adoptively transferred DCs in vivo [8 , 26] . It has been previously reported that VDCCs may play a role in DC maturation [18] and T cell activation [32] . VDCC-related activity on DCs has been implicated in engulfment of apoptotic bodies , IL-12-production and up-regulation of major histocompatibility complex II [17 , 18] , all of which are important immune functions of DCs . In line with these observations , the hypermigratory phenotype induced by T . gondii appears to rely on receptors and channels expressed by naïve DCs [8] , but additionally requires the active invasion of a T . gondii tachyzoite [4] , which is confirmed here by the observation that exogenous GABA per se is not motogenic on naïve DCs [8] ( while VDCC agonism is ) . Altogether , this also advocates that T . gondii primes the host cell for responsiveness to GABA and is consistent with the idea that GABAergic activation occurs in an autocrine fashion with minimal by-stander effect [8] . Notably , the vast majority of GABA-responding DCs also responded to ATP or to consecutive stimuli with GABA ranging from micromolar to millimolar concentrations , indicating that GABA does not render the DCs refractory to other Ca2+-related stimuli and that intracellular Ca2+ homeostasis is rapidly restored . We cannot exclude the involvement of additional mechanisms for Ca2+ entry in DCs [33] acting sequentially or in parallel . However , their possible contribution to the hypermigratory phenotype should be secondary or posterior to Cav1 . 3 activation , as silencing of Cav1 . 3 in both primary DCs and the DC cell line JAWS II or selective pharmacological inhibition of Cav1 . 3 [24] abrogated T . gondii-induced hypermotility . The posterior involvement of intracellular Ca2+ stores is also likely . Ca2+ also controls a number of critical processes in apicomplexan parasites , including gliding motility , cell invasion and egress [34–36] . It is unlikely that these mechanisms play in the interpretation of our results as inhibitors were added posterior to parasite invasion and non-significant effects were observed on parasite viability , reinvasion after egress or after forced release from treated host cells . On the other hand , our observations suggest that , through activation of the GABAergic system , T . gondii modulates the Ca2+ homeostasis of the infected host cell , albeit transiently and locally . Induction of Ca2+ signaling offers the advantage of bypassing transcriptional regulation in the host cells and thereby accelerating effector functions , i . e . rapid migratory activation of the invaded DC and , thereby , dissemination . We have previously shown that the onset of T . gondii-induced hypermotility precedes chemotactic responses in DCs in vitro and that , after the onset of chemotaxis , GABA/GABAA receptor-mediated hypermotility and CCR7-mediated chemotaxis can cooperatively enhance the migration of infected DCs in vitro [4 , 8] . Thus , Ca2+ entry in DCs , secondary to GABAergic activation , could hypothetically also influence Ca2+-dependent chemotaxis , with propagation of the signal to intracellular Ca2+ stores . Future research needs to determine if Cav1 . 3 is involved in the cytoskeletal rearrangements that accompany the onset of hypermotility , some of which are independent of GABAergic signaling [4] , e . g . the dissolution of adhesion-related podosomes [4] and the switch to amoeboid-like high velocity migration [37] . Our data demonstrate that VDCC inhibition in adoptively transferred infected DCs delays the dissemination of T . gondii tachyzoites in mice . VDCC inhibition reduced the parasite numbers in circulation and in the spleen early after inoculation , likely by delaying the outmigration of infected DCs from the peritoneal cavity [38] . The data advocates that the early presence of parasites in blood is important for setting the parasite loads in mice and that VDCC inhibition delayed this process . In line with this , perfusion experiments showed that the circulating pool of parasites contributes to the total parasite loads in organs and to dissemination during acute infection . Also , the high variability in leukocyte-associated parasitemias between mice 24 h post-inoculation is in contrast with the lower variability of parasite loads in the organs later during infection , and may indicate that parasitemia is intermittent early after infection . Because the spleen is an early site of T . gondii replication during acute infection [39] , this mobilizable pool of parasites ( by blood perfusion ) may be important for the systemic dissemination observed at later time points . Also , DCs and monocytic cells are parasitized early during infection [26 , 38 , 40] and , both leukocyte-associated tachyzoites [41] and extracellular ( free ) tachyzoites are detected in blood later during acute infection ( day 4 ) [41] . Our data show that adoptively transferred infected DCs reach the circulation and spleen rapidly but also that the transfer of replicating tachyzoites to new leukocytes is rapid and can occur in the peritoneal cavity , in line with previous observations [42 , 43] . Toxoplasma tachyzoites replicate in adoptively transferred DCs with lysis of infected DCs occurring within 48 h [5] and VDCC inhibition did not abrogate this process . This , together with the observed absence of parasites in brain parenchyma before day 4 , indicates that it is unlikely that the adoptively transferred DCs transported parasites into the brain parenchyma . Rather , the observed delay in penetration to the parenchyma upon benidipine-treatment may be a consequence of delayed or lower parasitemias . Yet , DCs infiltrate the brain parenchyma during toxoplasmic encephalitis [44] and transportation of parasites to the brain by CD11b+ leukocytes has ben shown [40] . However , more recent findings show that replication of tachyzoites in the endothelium is necessary before passage to the brain parenchyma [41] . Our studies contribute to elucidating the role of infected DCs in circulation and their impact on systemic dissemination , which indirectly impacts on parasitic loads in the brain parenchyma , but do not specifically address the mechanisms of passage of T . gondii tachyzoites across the blood-brain barrier [8] . Jointly , mounting evidences show that Toxoplasma utilizes combined strategies for systemic dissemination [6] , by hijacking leukocytes [5 , 38 , 40] and as free parasites [26 , 41] , and also with significant differences between Toxoplasma genotypes [26 , 45] . Additionally , intracellular localization of tachyzoites in migratory leukocytes may offer a safe intracellular niche for replication and delivery to organs and vasculature . To the best of our knowledge , this constitutes the first report showing that the VDCC signaling axis can be utilized by an intracellular pathogen to modulate host cell migration and potentiate systemic dissemination . Based on the data at hand , we propose a model for the initiation of the hypermigratory phenotype in DCs by T . gondii , mediated by GABAergic signaling and with Ca2+ acting as a second messenger ( Fig 9 ) . Initially , tachyzoite invasion triggers activation of the GABAergic system—GABA synthesis , transport and activation of GABAA receptors . Autocrine secretion of GABA by parasitized DCs leads to a membrane depolarization that activates the VDCC Cav1 . 3 , with entry of Ca2+ as a result . Finally , entry of Ca2+ activates downstream signaling pathways that lead to cytoskeletal rearrangements and hypermotility . Mounting evidence indicates that , rather than being passively transported , intracellular microorganisms induce refined molecular orchestrations to manipulate the signaling pathways that modulate the migration of infected immune cells [6 , 46 , 47] . Continued investigations into how intracellular pathogens manipulate host cell Ca2+ signaling pathways may identify new targets for inhibiting processes associated to pathogenesis .
The Regional Animal Research Ethical Board , Stockholm , Sweden , approved experimental procedures and protocols involving extraction of cells from mice ( N135/15 , N78/16 ) , following proceedings described in EU legislation ( Council Directive 2010/63/EU ) . Mouse bone marrow-derived DCs were generated and typified as previously described [8] . Briefly , cells from bone marrow of 6–10 week old C57BL/6 mice ( Charles River ) were cultivated in RPMI 1640 with 10% fetal bovine serum ( FBS ) , gentamicin ( 20 μg/ml ) , glutamine ( 2 mM ) and HEPES ( 0 . 01 M ) , referred to as complete medium ( CM; all reagents from Life Technologies ) , and supplemented with 10 ng/ml recombinant mouse GM-CSF ( Peprotech ) . Medium was replenished on days 2 and 4 . Loosely adherent cells were harvested on day 6 . The murine DC line JAWS II ( CRL-11904 ) and murine neuroectodermal cell line NE-4C ( CRL-2925 ) were cultured as indicated by the supplier ( American Type Culture Collection ) . Primary astrocytes ( ACs ) were generated from cortices from 1–3 day-old C57BL/6 mice as previously described [48] . Freshly egressed Toxoplasma gondii tachyzoites of the RFP-expressing PRU-RFP [49] or GFP- and luciferase-expressing PTGluc [39] lines , kept on a 2-day passage cycle in murine fibroblast monolayers ( L929 , Sigma-Aldrich ) , were used in assays . γ–aminobutyric acid ( GABA ) , Adenosine triphosphate ( ATP ) , ( S ) -SNAP-5114 ( SNAP ) , semicarbazide ( SC ) , nifedipine , Bay K8644 , ( all from Sigma-Aldrich ) , pyridoxalphosphate-6-azophenyl-2′ , 4′-disulfonic acid ( PPADS ) , ( 4R ) -rel-1 , 4-Dihydro-2 , 6-dimethyl-4- ( 3-nitrophenyl ) -3 , 5-pyridinedicarboxylic acid 3-methyl 5-[ ( 3R ) -1- ( phenylmethyl ) -3-piperidinyl] ester hydrochloride ( benidipine hydrochloride , all from Tocris ) and 1- ( 3-Chlorophenethyl ) -3-cyclopentylpyrimidine-2 , 4 , 6- ( 1H , 3H , 5H ) -trione ( CPCPT , Merck Millipore ) were used at the indicated concentrations . Motility assays were performed as previously described [4] . Briefly , 105 DCs were incubated with freshly egressed tachyzoites ( MOI 3 , 4 h ) . The cells were mixed with collagen I ( 0 . 75 mg/ml , Life Technologies ) and transferred to a chamber slide ( Nalge Nunc Internat . ) or 96-well plate . Imaging was performed for 1 h , 1 frame/min , at 100x magnification ( Zeiss AxioImager ) . Time stacks were stabilized ( Image Stabilizer , ImageJ ) and motility data obtained by manual tracking of cells ( Manual Tracking , ImageJ ) of approximately 50–60 cells per condition . In infected samples , only cells where the RFP and DIC signals co-localized were tracked . Transmigration assays were performed as previously described [8] . Briefly , 106 DCs were incubated with freshly egressed tachyzoites ( MOI 3 , 6 h ) , transferred into transwell filters ( 8 μm pore size; BD ) in duplicate and incubated over night . Transmigrated DCs were quantified using a Neubauer hemocytometer . Ca2+-free medium was prepared from Ca2+-free DMEM , 1% FBS , gentamicin ( 20 μg/ml ) , glutamine ( 2 mM ) , 1 mM EGTA and HEPES ( 0 . 01 M ) , all reagents from Life Technologies . DCs ( 2x105 ) were seeded on 5% 3-aminopropyltriethoxysilane coating glass bottom dish and incubated at 37°C with 5% CO2 for 15 min . DCs were then loaded with 2 μM Fluo-8H/AM ( AAT Bioquest ) in CM at 37°C with 5% CO2 for 15 min , and washed with Krebs-Ringer’s solution ( 150 mM NaCl , 6 mM KCl , 1 . 5 mM CaCl2 , 1 mM MgCl2 , 10 mM HEPES and 10 mM D-glucose ) with 5% FBS . Time-lapse imaging was performed 2 . 5 s/frame , at 37°C with 5% CO2 on 200x magnification ( Zeiss LSM 780 microscope equipped with a definite focus function ) . Cells were perfused with Krebs-Ringer’s solution with 5% FBS via a peristaltic pump ( 0 . 5 ml/min ) , which was also used to deliver pharmacological agents . The signals from individual cells were analyzed with ImageJ ( version 1 . 46r , ROI Multi Measure ) . Each trace was normalized against the minimum value of all time points and a responding cell was defined as a signal exceeding 20% above baseline . To determine the expression of the CaV1 . 3 protein in DCs and ACs , cells were lysed in RIPA buffer ( 150 mM NaCl , 50 mM Tris , 0 . 1% Triton , 0 . 5% deoxycholic acid , 0 . 1% SDS ) with protease and phosphatase inhibitor cocktail ( Thermo Fisher Scientific ) followed by sonication , addition of 4 x laemmli sample buffer and boiling . Proteins were separated by 8% SDS-PAGE , and blotted onto PVDF membrane ( Millipore ) , blocked in 2 . 5% BSA followed by Western blotting with monoclonal anti-CaV1 . 3 C-terminal ( Abcam ) , anti-GAPDH ( Millipore ) and anti-rabbit HRP ( Cell signaling ) . Proteins were revealed by enhanced chemiluminescence ( GE Healthcare ) in a BioRad ChemiDoc XRS+ . DCs ( 105 ) were plated on poly-L-lysine coated coverslips and incubated with T . gondii tachyzoites ( MOI 3 , 4 h ) . Cells were fixed in 4% paraformaldehyde and permeabilized ( 0 . 1% Triton X-100 ) , before incubation with mouse monoclonal anti-CaV1 . 3 biotin ( 1:100 , Abcam ) and streptavidin-Alexa555 ( 1:500 , Molecular Probes ) . Samples were treated with DAPI and imaged by confocal microscopy ( Zeiss LSM780 ) . Total RNA was extracted using TRIzol reagent ( Life Technologies ) . First-strand cDNA was synthesized using Superscript III Reverse Transcriptase ( Life Technologies ) . Real time quantitative PCR ( qPCR ) was performed in triplicates using SYBR green PCR master mix and a 7900HT Fast Real Time PCR system ( Applied Biosystems ) . Products were analyzed with ABI 7900HT Sequence Detection System ( Applied Biosystems ) or Rotor gene ( Corbett ) . 2-ΔCt values are used to calculate the relative expression levels of 9 VDCC subtypes , with TATA box binding protein ( TBP ) as reference gene ( S2 Table ) . For quantification of Cav1 . 3 knock-down , glyceraldehyde 3-phosphate dehydrogenase ( GAPDH ) and actin were used as reference genes ( S2 Table ) . For quantification of T . gondii in tissues , the B1 gene was used ( S2 Table ) . Self-complementary hairpin DNA oligos targeting the CaV1 . 2 ( Cacna1c ) mRNA , CaV1 . 3 ( Cacna1d ) mRNA , and a non-related sequence ( luciferase , Luc ) were chemically synthesized ( DNA Technology , Denmark ) , aligned and ligated in a self-inactivating lentiviral vector ( pLL3 . 7 ) containing a CMV-driven EGFP reporter and a U6 promoter upstream of cloning restriction sites ( HpaI and XhoI ) [50] ( S3 Table ) . Restriction enzyme analysis and direct DNA sequencing confirmed the correct insertion of short hairpin RNA ( shRNA ) sequences . Lentivirus production was done using lipofectamine transfection . Briefly , shCaV1 . 2 , shCaV1 . 3 or shLuc vectors were co-transfected with psPAX2 packaging vector and pCMV-VSVg envelope vector into Lenti-X 293T cells ( Clontech ) and the resulting supernatant was harvested after 60 h . Recovered lentiviral particles were centrifuged to eliminate cell debris , filtered through 0 . 45-mm cellulose acetate filters and concentrated by ultracentrifugation . Titers were determined by infecting Lenti-X 293T cells with serial dilutions of concentrated lentivirus . NE-4C cells , JAWS II cells and DCs ( day 3 ) were transduced by spinoculation at 1000 g for 30 min in presence of hexadimethrine bromide ( Polybrene , 8 μg/ml; Sigma Aldrich ) . Three to 5 days post-transduction , EGFP-expression was verified by epifluorescence microscopy before the cells were used in experiments . Transduction frequency was defined as the number of EGFP-expressing cells related to the total numbers of cells in five representative fields of view . DCs ( 105 ) were incubated with freshly egressed T . gondii PRU tachyzoites ( MOI 3 , 4 h ) . Cells were washed twice and incubated for 16 h in Krebs-Ringer’s solution supplemented with MEM essential and non-essential amino acids ( Life Technologies ) and 20 μg/ml gentamicin , referred to as mod . R . Inhibitors were present before and after the washes . 1 μL of cell supernatants were overlaid with 1μl of matrix ( 2 . 5 mg α-Cyano-4-hydroxycinnamic acid ( HCCA ) dissolved in 50% acetonitrile , 47 . 5% H2O , 2 . 5% TFA ) . Samples were analyzed by MALDI TOF mass spectrometry ( Microflex LT , Bruker Daltronics ) at laser frequency 60 Hz , mass range 0–1000 m/z , delayed ion extraction 100 ns , acceleration voltage 20 kV , lens voltage 6 kV and calibrated using the mass of HCCA matrix ions . Analysis was performed with flexAnalysis ( version 3 . 3 , Bruker Daltronics ) . Cells were collected from blood , peritoneum and spleen and depleted of red blood cells . Cells were then stained for CD11b ( clone M1/70 ) , CD11c ( clone N418 ) , CD19 ( clone 1D3 ) , NK1 . 1 ( clone PK136 ) , CD3 ( clones 145 . 2C11 ) and live/dead marker Viability Dye eFluor 780 ( eBioscience ) or Fixable Yellow Dead Cell Kit ( Invitrogen ) following blocking of Fc receptors ( 24G2 ) . All antibodies were from Biolegend ( San Diego , CA ) . After 30 minutes incubation , the cells were washed extensively and then fixed prior to running on FACCyAN ADP LX 9-colour flow cytometer ( Beckman Coulter , Pasadena , CA ) . Data were analyzed using FlowJo software ( Tree Star Inc , OR ) . Adoptive transfers were performed as previously described [8] . Briefly , DCs were challenged with freshly egressed PTGluc tachyzoites ( 6 h , MOI 3 ) . Extracellular parasites were removed by centrifugation . Following resuspension in RPMI , tachyzoite-infected DCs or freshly egressed tachyzoites were adoptively transferred intraperitoneally into recipient C57BL/6 mice . Total number of colony-forming units ( cfu ) injected into animals was confirmed by plaquing assays . Benidipine ( 40 μM ) was added to DCs for the last 3 h of the 6 h challenge with tachyzoites and replenished ( 40 μM ) prior to injection in mice . When indicated , cells were stained with CMTMR following manufacturer´s instructions ( Invitrogen ) . Eight-10 week old C57BL/6 mice were inoculated i . p . with freshly egressed PTGluc tachyzoites , or with PTGluc-infected DC ± benidipine . 3 mg D-luciferin potassium salt ( Caliper Life Sciences , Hopkinton , MA , USA ) was injected i . p . and mice were anesthetized with 2 . 3% isoflurane prior to BLI . Ten min after injection of D-luciferin , biophotonic images were acquired for 180 s ( medium binning ) with an In Vivo Imaging System ( Spectrum CT , Perkin Elmer ) . For ex vivo imaging , organs are extracted and assessed as above . Analysis of images and assessment of photons emitted from a region of interest ( ROI ) was performed with Live Imaging Software ( version 4 . 2; Caliper Life Sciences ) . Plaquing assays were performed as described [8] . Briefly , organs were extracted and homogenized under conditions that did not affect parasite viability . The number of parasites was determined by plaque formation on fibroblast monolayers . When indicated , tachyzoites where released from infected DCs by repeated passages through a hypodermic needle ( gauge 27 ) , previous to plaquing . Transcardial blood perfusion was performed by injection of 25 ml PBS in the left ventricle after incision of the right atrium . Peritoneal lavage was performed by intraperinoneal perfusion and aspiration of 10 ml PBS using a hypodermic needle . Statistical analyses were performed using R Stats Package version 3 . 0 . 2 ( R Foundation for Statistical Computing , Vienna , Austria ) . Normality was tested by the Shapiro-Wilks test . P-values > 0 . 05 were defined as non-significant .
|
Dendritic cells are considered the gatekeepers of the immune system but can , paradoxically , also function as ‘Trojan horses’ to mediate dissemination of the common intracellular parasite Toxoplasma gondii . Previous work has shown that Toxoplasma hijacks the migratory machinery of dendritic cells by inducing secretion of the neurotransmitter GABA and by activating GABAergic signaling pathways , thereby making infected dendritic cells hypermigratory in vitro and in vivo . Here , we show that the signaling molecule calcium plays a central role for this migratory activation and that signal transduction is preferentially mediated through a subtype of voltage-gated calcium channel ( Cav1 . 3 ) . This study functionally implicates Cav1 . 3 channels in a , hitherto uncharacterized , calcium signaling axis by which dendritic cells are induced to become migratory . The studies show how an obligate intracellular pathogen takes advantage of non-canonical signaling pathways in immune cells to modulate their migratory properties , and thereby facilitate the dissemination of the parasite .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"parasite",
"groups",
"neurochemistry",
"immune",
"physiology",
"body",
"fluids",
"toxoplasma",
"gondii",
"spleen",
"neuroscience",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"parasitology",
"apicomplexa",
"tachyzoites",
"protozoans",
"toxoplasma",
"neurotransmitters",
"cellular",
"structures",
"and",
"organelles",
"head",
"cell",
"membranes",
"gamma-aminobutyric",
"acid",
"biochemistry",
"eukaryota",
"blood",
"anatomy",
"cell",
"biology",
"physiology",
"jaw",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2017
|
Voltage-dependent calcium channel signaling mediates GABAA receptor-induced migratory activation of dendritic cells infected by Toxoplasma gondii
|
The microbes that inhabit particular environments must be able to perform molecular functions that provide them with a competitive advantage to thrive in those environments . As most molecular functions are performed by proteins and are conserved between related proteins , we can expect that organisms successful in a given environmental niche would contain protein families that are specific for functions that are important in that environment . For instance , the human gut is rich in polysaccharides from the diet or secreted by the host , and is dominated by Bacteroides , whose genomes contain highly expanded repertoire of protein families involved in carbohydrate metabolism . To identify other protein families that are specific to this environment , we investigated the distribution of protein families in the currently available human gut genomic and metagenomic data . Using an automated procedure , we identified a group of protein families strongly overrepresented in the human gut . These not only include many families described previously but also , interestingly , a large group of previously unrecognized protein families , which suggests that we still have much to discover about this environment . The identification and analysis of these families could provide us with new information about an environment critical to our health and well being .
Every ecological niche presents specific challenges that face the population of organisms that inhabit them . When analyzing species that thrive in any particular environment , we can expect that certain key functional characteristics would correlate with success and differentiate those species from others that fail in colonizing that environment . This is especially obvious for microbes , and detailed analysis of almost every sequenced microbial genome provides examples of adaptation , mostly in terms of the presence of genes that code for specific functions required for that microbe to succeed in a given environment . However , studying microbes one genome at a time does not generally provide enough data and meaningful statistics to explore fully the relationships between individual gene families and their environments . This has now changed with the advent of metagenomics , which can investigate entire microbial communities associated with single environments . In metagenomics shotgun sequencing , which identifies genes present in a given environment , the associations between gene families and specific environments can be analyzed directly . All such studies carried out so far have identified unique distributions of functional classes of protein families that are strongly correlated with the specific features of the given environment , be it presence of specific nutrients , acidity , high temperature , etc . For instance , Gill et al . have shown that the human gut microbiome is enriched in proteins associated with amino acid and vitamin production [1] . Another study has confirmed these observations and found additional functional groups of proteins overrepresented in the human gut , such as for carbohydrate and lipid transport and metabolism [2] . Similar observations have been made during analysis of the genomes of several human gut–associated microbes , such as Bacteroides fragilis [3] and Bacteriodes thetaiotaomicron [4] . However , these analyses have focused exclusively on already recognized and functionally characterized protein families—all of which were previously identified and characterized by resources such as PFAM [5] , COG [6] or Interpro [7] . As a result , two important groups of protein families were not included in such analyses; namely , families already discovered but not yet characterized , and novel families specific to a newly studied environment but rarely or never found in microbes or in the environments previously studied . . Both sets represent a possible wealth of information about the processes necessary for microbes to survive in the human gut . Their importance for further study was exemplified by a recent metaproteomics study [8] , in which almost 20% of all recognized proteins , including several of the most abundant ones , were classified as “hypothetical proteins” and did not belong to well-characterized protein families . Thousands of such environment-specific protein families have also been identified in other environments , such as the ocean [9] , [10] . In this study , we address this important issue by an ab initio search for protein families in datasets that represent the environment we are studying , and a subsequent abundance/conservation analysis of all protein families , including new examples and those not covered by any functional category . An important issue in interpreting results of such large-scale studies involves widespread inconsistencies in use of the term “protein family” . While the general definition of a protein family as a group of proteins that evolved from a common ancestor seems very clear , in practical applications , this term can mean anything from a group of very close homologs to an extensive , very divergent group of proteins that shared a common ancestor billions of years ago , but have now evolved into a multitude of sub-families with different functions . Automated procedures for indentifying protein families typically indentify closely related families composed of highly similar proteins , which , upon further analysis , could be included in an already known family or combined with others to form a larger family . Therefore , estimates of the numbers of new protein families provided in large-scale automated project are typically too high . In the context of this paper , we address this problem with detailed analysis of some of the families found in the automated analysis . The human gut is a very specific environment , rich in diverse nutrients , but also full of challenges for its microbial inhabitants . Because of its richness , the microbes inhabiting human gut form one of the densest microbial communities on Earth , reaching 1011 cells per gram [11] . Species that inhabit that environment have to be able to extract energy from diverse and rapidly changing sources , reflecting the diverse human diet that can vary significantly in content and quantity over time in both daily and seasonal cycles . Species forming the human gut microbiome also need to survive encounters with the human immune system and to coexist with other microbes . Sets of specific microbial proteins must carry out the essential tasks of recognizing new nutrients , transporting them into the cytosol and metabolizing them , neutralizing or suppressing human immunity , and signaling to other bacteria and host cells . The presence of genes coding for such proteins in a genome would provide a distinct competitive advantage to a human gut symbiont or commensal microbe . In this paper , we seek to identify such environmentally specific protein families , focusing on the human gut as a target environment . Because of the obvious importance of this environment for human health , several groups have performed large-scale , random , shotgun sequencing experiments on representative samples providing a direct view of the gene content of this environment [1]–[2] . At the same time , a major sequencing effort , the NIH Human Microbiome Project ( HMP ) , is specifically targeting genomes of human gut microbes [12] as identified , for instance , by 16S rRNA studies . Genomic sequencing provides information for individual species but , with a coordinated effort to sequence the genomes of hundreds of microbes from a single environment , the resultant data can also be translated into an overall gene content . Thus , two sets of independent data can be obtained that describe the gene content of the same environment . Both approaches have their advantages and shortcomings: metagenomic shotgun sequencing provides a relatively unbiased , but small sample of genes that can be found in a given environment . On the other hand , genomic sequencing provides a full set of proteins from a genome , but its success depends on our ability to culture specific species and , thus , might leave large groups of microbes without any representation . Arguably , both of these approaches provide only a very crude approximation of the actual gene content of an environment . However , as we will show , data from both methods present a surprisingly coherent view of the gene content of the human gut , at least on the level of protein families , which encourages us that the data are robust enough for a survey analysis , such as presented here . We hypothesize that genes coding for proteins that are necessary and beneficial for survival of microbes in the human gut environment will be found abundantly both in the genomes of the species found in that environment and in metagenomic data sampling of the same environment . Hence , we can verify observations made on one set of data by using the other as a reference . At the same time , since an extensive study of the human gut environment and its microbiome was only started very recently , protein family databases and annotation resources , which typically work with significant time lag in recognizing novel protein families , simply haven't had enough time to include data for new families found only in this environment . In this manuscript , by automated clustering in metagenomics samples from the human gut we identify about 1 , 800 novel protein families and curate and analyze in detail about 180 of them . Some of these families have been confirmed and characterized by structural studies , since the PSI large-scale Structural Genomics Centers have used a preliminary version of our analysis to select some of the most abundant protein families in the human gut as targets for structural determination [13] . We also present a comprehensive analysis of the distribution of protein families in the human gut environment , including both those previously known , as well as the new families identified in this study .
While many of the ORFs identified in metagenomics shotgun sequencing projects can be classified into already known and defined protein families , many—often over 50% ( see Figure 1 ) —cannot . About 6% are singletons ( sometimes called ORFans ) [14] , i . e . , proteins that don't have any homologs in current protein databases . Nevertheless , most of the unclassified proteins do form families of varying sizes and such new families may play very important roles in specific environments , but , by default , were omitted from all previous analyses . In our study , we aim to get a complete picture of protein family distributions in the new environment . To this end , we optimized a previously introduced [10] clustering technique ( see the Methods section for details ) and used it on the set of over 600 , 000 ORFs from two large human gut metagenomics projects [1] , [2] . We identified almost 1 , 800 protein families fulfilling our size criteria , of which 926 could be matched to PfamB , the uncurated section of the PFAM database , while the other 835 were found de novo in the metagenomic data . We now describe results of various types of analyses applied to these data , including manual curation and experimental verification . In Figure 2 , we compare the distribution of sizes of the new protein families identified here to that of PfamA families that were represented in the metagenomics samples , as sorted by the approximate number of members present in the metagenomic dataset . Both sets have similar size distributions , with PFAM families being somewhat larger . It is interesting to note that only about 2 , 300 ( from over 10 , 000 ) PFAM families pass the size threshold ( i . e . have ten or more members in the gut-related genomes and metagenomic samples ) to be included in this histogram . In the next step , we study coverage of the metagenomics datasets , as well as both reference genome sets ( HGR and HGU ) by the expanded set of families that includes characterized domains from the PFAM database ( PfamA ) [5] , as well as the families newly found in this work ( see the previous section ) . The level of coverage of HGR and HGU genomes by PfamA families is 51% and 52% , respectively . However , the level of coverage drops dramatically to 39% for metagenomic samples . Clearly , while both HGR and metagenomics samples represent the same environment , the metagenomic datasets contain a larger portion of previously uncharacterized genes , most likely from genomes of as-yet-uncharacterized species . Adding new families identified in this work increases coverage of the metagenomic dataset by approximately 8 . 9% and increases coverage of reference genome sets by 8 . 4% and 3 . 5% for HGR and HGU genomes , respectively . However , in all sets , a large percentage , 40–45% of all ORFs , still cannot be assigned to either an already known or a new family . This group of ORFs can be broadly divided into two groups: a majority ( ∼88% of the unclassified proteins , i . e . 45% of the total ) are proteins that form small families ( <10 members ) , which were not included in the analysis because of the size thresholds used in this work . These “microfamilies” may be an important source of information , but the computational complexity of applying detailed analysis to each of these possible families must await future research . It is very likely that these microfamilies will expand to full-sized families with the addition of new metagenomics datasets , or will be found to be included in already defined families as the sensitivity of their profile description improves with addition of further homologs . The remaining 12% of the unclassified proteins , i . e . 6% of the total , have no BLAST matches internal to the human gut metagenomics samples and , thus , cannot be grouped into clusters of metagenomic sequences . Truly unique protein sequences may be specific to uncharacterized , rare organisms , but it is also possible that they represent failures of sequencing technologies , bad ORF calls , etc . The validity of ORF calls can be monitored; in the analysis of the GOS metagenomics samples , the number of similar sequences has been shown to be strongly correlated with the validity of an ORF call [9] , other criteria can be used as well [10] . Once a complete set of protein families is identified , the next step is to determine the extent to which these families are specific to our target environment ( the human distal gut ) . To this end , we calculate an “essentiality coefficient” ( Es ) for every family ( see Methods section for a formal definition of Es , as well as for definitions of other measures of environmental specificity of protein families ) . An essentiality coefficient equal to 1 means that at least one member of a given family was found in the genome of each of the human gut–associated microbes , but no members were found in any of the reference set of genomes—thus , this family is considered as essential for the gut environment . An Es close to 0 indicates lack of preference , and an Es close to −1 indicates an “anti-preference . ” Figure 3 presents the distribution of essentiality coefficients for protein families from the PfamA database and for new protein families found in this work . PfamA protein families show an almost symmetric distribution of preference and anti-preference for genomes of human gut–related microbes . At the same time , the new families found in this work are very specific for human gut–related microbes . This outcome is , of course , expected as these families were identified by clustering from the metagenomics datasets with the aim of identifying environment-specific families . Interestingly , some families that were not specific for the human gut environment were notably found by clustering metagenomics ORFs ( lower-right region of graphs in Figure 3b ) . These are the protein families , found by clustering the metagenomic datasets that turn out to be more frequently conserved in random genomes not connected to that environment . One example is the family HGC00614 , composed of 18 proteins found in the metagenomic data . Upon constructing an appropriate HMM , we found that this family is a likely new family in the PFAM PLP_aminotran ( CL0061 ) clan , with many homologs across multiple species . It is also worth noting , that some families found in metagenomic data have not been found in any fully sequenced genomes of microbes from the same environment , clearly showing that complete genome sequencing still hasn't fully explored the diversity of genes present in this environment . Several different measures can be proposed to compare distributions of a protein family between two datasets . For instance , the comparative overrepresentation ( Ov ) in a specific dataset details the number of members a family has in one dataset as compared to another reference set . Another metric is the expansion ( Ex ) of a protein family when the relative counts of protein families are compared but , rather than normalizing by the total number of proteins in the genomic set , counts are normalized by the number of genomes that contain at least one match . This metric highlights families that may not have the largest counts , but when found , have multiple copies in the same genome . Yet another measure is the essentiality coefficient ( Es ) used above in Figure 3 , which compares the percentage of genomes in each group that contain at least one member of a family . So far , we have only used the latter specificity measure ( Es ) . In the following analysis , we will use and compare all three measures as each captures some of the intuitive notion of specificity . Each measure corresponds to a different biological mechanism of “specificity” . Having multiple paralogs of proteins from families with high Es , but low Ov or Ex , clearly does not provide an advantage to a microbe , therefore protein families that score well with Es likely perform highly specific , but essential functions . On the other hand , large number of members of overrepresented or expanded families provides such advantage , but may represent only one of many possibilities of solving a given problem; hence , they are not present in all microbes in a given environment . For instance , metabolic enzymes would likely belong to the latter category , while defense and host signaling proteins would likely belong to the former . As discussed extensively in the papers that study the human gut microbiome directly through metagenomics sequencing [1] , [2] or indirectly through genome sequencing of specific microbes representative of this environment [3] , [4] , certain protein families involved in specific types of function were observed to be strongly expanded in the human gut microbiome compared to families found in “average” microbes . However , these studies did not cover protein families of unknown function , and focused only on one measure of specificity that is related to our overrepresentation measure Ov , in Table 1 . In our analyses , we use and compare three different specificity measures: Ov in Table 1 , Ex in Table 2 , and Es in Table 3 . Our research also focused on complete family coverage , including families of known and unknown function , as well as new families specific for the gut environment . Novel families were ranked by the three different ranking methods , with the top 10 hits listed; Ov in Table 4 , Ex in Table 5 , and Es in Table 6 . The full list of 180 annotated protein families is detailed in Table S1 , in the supplemental material . Domains of unknown function ( DUF ) dominate the overrepresented group with four such families in the top 10 when sorted by overrepresentation ( Ov ) , but the DUFs are also present in other forms of ranking . The presence of so many weakly characterized protein families in all specificity categories clearly illustrates the inadequacy of our knowledge about this important environment . Similarly , all previous analyses focused mostly on metabolic proteins and interpreted the specificity of the human gut environment predominantly in the view of its unique metabolic content . We show here that protein families involved in regulatory and DNA exchange functions are also strongly present among the most overrepresented families . It is possible and , indeed , very likely that , by using more sensitive sequence analysis tools , many of the families identified here would be eventually grouped into larger entities , such as clans in PFAM [5] ( or superfamilies in other protein classification systems ) , that represent more distant evolutionary relationships . However , for the purpose of this analysis , we will focus on the family level as practically defined by major community resources , such as PFAM [5] or Interpro [7] . Upon further analysis of the families identified in an automated , ab initio clustering of protein sets we realized that many may not fit such definitions . For instance , proteins that form distant branches of already existing families may form well-defined clusters in the automated analysis , but careful optimization of HMMs that define old versus new families would be necessary to decide if they would form a new family or if they could be included in the old family by readjusting its definition . For instance , we found several potential families that belong to the SusC and SusD mega-families . SusC and SusD are part of the sus ( starch utilization system ) operon in B . thetaiotamicron , an archetype of polysaccharide utilization loci found in multiple copies in all Bacteroides and related species [15] , [16] . Both families are extremely divergent; only a small number of their members are covered by PFAM HMMs that define Ton_B–like and SusD families , respectively . The complex evolution of the SusD protein family is the subject of a separate paper [17] . Families that define new domains in proteins with already recognized PFAM domains form the second group . Again , without detailed analysis , it would be difficult to decide if such families should be defined as new or covered by readjustment of the boundaries of already defined families . We used several filters to identify and remove the group of new families that would be most likely to overlap with already existing PFAM families ( see Methods ) , undoubtedly eliminating some genuine , novel families . Next , we analyzed the remaining ones by hand to identify those that are most likely to conform to the “PFAM standard” , i . e . , families that represent functional domains that do not overlap with protein families described in the PFAM database . At this point , the hand-curated set of PFAM-quality families exceeds 180 and would undoubtedly expand further as the curation and analysis continue . We provide the current list of curated families as a Table S3 in the Supplemental Materials . Tables 4–6 present the top families from this group in three different “specificity” categories . ( An analogous table for the Pfam families was presented in the previous section . ) As mentioned earlier , a preliminary version of this analysis was used to select structure determination targets for the four large NIH Protein Structure Initiative production centers in two “target drafts” [13] in mid- and late 2008 . As of May 2009 , representatives of almost 800 of the 1 , 761 protein families identified here had been successfully expressed and purified in vitro , supporting the conjecture that the new families represent real proteins and not “shadow ORFs” or other sequencing artifacts . The last column in Tables 4–6 provides information about the status of the representative of a given family that is most advanced in the PSI production pipeline . The structures of representatives of several protein families described here have been successfully solved , and their coordinates deposited into the Protein Data Base . For instance , Thermotoga maritima proteins TM1486 ( 1VPV ) and TM841 ( 1MGP ) represent DegV ( PF02645 ) , a large family of proteins , shown by structure analysis to be involved in fatty acid binding . The Lactobacillus acidophilus NCFM protein LBA1001 , PDB entry 3EDO , incorrectly described in the literature as a TRP repressor , has 142 homologs in metagenomic datasets , and at genomic levels of conservation goes from 12% of species in the HGU sample to 84% in the HGR set . A third example of a protein prevalent in the human gut environment is the protein family represented by PDB entry 2PC1 . This acetyltransferase/GNAT family protein has 47 metagenomic homologs and is present in 73% of HGR species , while it is rare in the HGU list ( 5% of species ) . Other protein families determined to be important to the human gut environment and found independently by this study include PfamA family PF08842 ( DUF1812 ) , represented by PDB entry 3GF8 . Only 3 homologs are found in genomes of free-living bacteria ( HGU set ) , compared to 47 in the of human gut-related microbes . The proteins matching this family in the HGU genomes were found to be hypothetical proteins in Porphyromonas gingivitis , an human oral pathogen , which was included in the HGU set because of its specific definition ( see Methods ) , but should probably be reclassified to the HGR set . The Protein Structure Initiative has also solved several proteins from family PB002962 ( PDB entries 3DB7 and 3DUE ) . This family was found in eight of the thirteen of the metagenomic samples , with a total of 34 homologs and present in only 1 . 6% of HGU genomes as compared to 21 . 5% of HGR genomes .
The gastrointestinal tract is extremely important for overall human health . Numerous diseases , from digestive disorders and immune diseases to numerous types of cancer , notably involve the GI system . At the same time , the human GI system , and especially , the distal gut , is a surprisingly complex and little understood environment , inhabited by a complicated bacterial community that carry enzymes for processing byproducts and downstream products of metabolism in the stomach and proximal gut . Rich in nutrients , the gut harbors one of the densest microbial populations known . These microbes and their metabolism play a critical role both in health and in diseases of the GI system . While the culturable microbes living in the human gut have been studied for decades ( for instance , E . coli ) , the development of new technologies and the concept of metagenomics provided a decisive , paradigm-changing shift in studies of this environment , in which the diversity and the communal nature of the human gut microbiome could be uncovered . We thus now have access to several synergistic , but independent , lines of investigation into the surprisingly unknown world of microbes inhabiting human cavities . Here , we investigated what types and number of novel , previously uncharacterized , protein families can be found in this environment . In our analysis , we have shown that many protein families , most completely uncharacterized , show strong specificity for this environment . Undoubtedly , the functions of these proteins play an important role in the maintenance and operation of the human gut microbiome . Approximate function predictions based on distant homology recognition identified many proteins that are involved not only in metabolism , but also in signaling , regulation , and phage activity , and are obviously very important in such dense bacterial communities . We have identified not only a few thousand known protein families as strongly overrepresented in the human gut environment , but also , many potentially new protein families . Many of these assignments have now been confirmed by structural determination by the PSI centers , and many of their functions have been predicted due to fold recognition techniques . However , many yet uncharacterized or completely novel families have been shown to be specific to the human gut environment . This observation , in turn , suggests that many unknown and uncharacterized processes are yet to be discovered in this environment . Apart from these interesting insights about this specific environment , our observations suggest this approach is applicable to analyzing other environments . Historically , genomic analysis has focused on individual species , but it is important to remember that an organism does not exist in a vacuum . Organisms evolved their specific traits in the context of their environment . By sampling the gene pools in a given environment , we can learn about the protein families that are key for survival in those environments . The methods presented here should aid in organizing and streamlining such analyses .
Our analysis is derived from several different sources: metagenomic sequencing , 16S rRNA sampling , fully sequenced cultured genomes from NCBI , and draft genomes published by the Human Gut Microbiome Initiative ( HGMI ) [12] . Each of these data sources is publicly available . We used a human gut metagenomic dataset derived from the Kurokawa [2] study . This dataset contains 350 , 000 assembled contigs from 13 individuals , both male and female , with ages ranging from 3 months to 45 years . Although these genomic data come from 13 separate individuals , we have treated them as a single set to improve the odds of finding human gut–related proteins . Preparation of the sequence metagenomic data begins with Open Reading Frame ( ORF ) prediction done by Metagene [18] . Metagene analysis produced a set of 665 , 559 ORFs . From this initial set , incomplete ORFs that ran off the edge of the sequence read were removed . A total of 303 , 314 complete ORFs were left . This set was then used to identify protein families ( see the section Clustering and identification of uncharacterized and new families ) . The HGMI sequenced genomes provide an ideal reference set of human gut–related microbial genomes . In addition to the human gut–related reference genomes ( HGR ) , we also needed a set of genomes not related to the human gut environment for comparative analysis . The set of selected fully sequenced genomes was derived from the collection of bacterial genomes available from NCBI . As of July 2008 , this library included 765 bacterial genomes . We utilized data from 16S rRNA sampling to eliminate genomes linked to the human gut environment by targeted metagenomic sampling . The 16S rRNA data was derived from two sources: Greengenes [19] and David Relmann's published human gut sample 16S RNA set [20] . Using data available in the Green Genes , we searched for 16S rRNA sequences associated with keywords “human” and one of the following: “fecal , ” “faecal , ” “colon , ” “intestine , ” “stool , ” “rectum , ” “cecum , ” “feces , ” “intestinal , ” “colitis , ” “stomach , ” or “gut . ” This search produced a set of 38 , 839 16S rRNA sequences . This set was added to the 11 , 831 sequences from the Relman dataset . Using a broad Operational Taxonomic Unit ( OTU ) of 90% sequence identity , we ran BLAST against the set of NCBI bacterial genomes and selected 493 species not linked to the human gut microbiome ( i . e . those which did not match any 16S RNA sequences from species related to human gut ) . We refer to this latter set as the Human Gut–Unrelated ( HGU ) set . To create the set of Human Gut–Related genomes , we started with 45 genomes from the HGMI project , each currently in the draft stage . In addition to that base set , we added 20 finalized NCBI bacterial genomes tagged with matching 16S rRNA sequences that were manually confirmed by examining NCBI genome project annotations . This provided us with a set of 65 genomes referred to as the Human Gut–Related ( HGR ) set . Detailed information about both sets is available in the Table S2 in Supplemental Materials . One of the important aspects of analysis of metagenomic sequences is the identification of novel sequences . These sequences with no known homolog in existing sequence databases are referred to as orphan sequences . In the study by Kurokawa et al . [2] , orphan analysis was carried out by taking over 600 , 000 predicted ORFs and looking for genes previously seen with BlastP with a threshold of 1 . 0e-5 against a custom , extended , non-redundant ( NR ) , sequence database . Of the original set , 162 , 647 genes were determined to be orphan sequences . This set was combined with 503 , 115 other orphan genes from other metagenomic environments . The total set of orphans was calculated by producing an all-to-all BlastP [21] comparison . Connections were drawn between proteins with alignments that had a Blast score of 60 or greater and were marked as a match and the connection graph was then clustered with TribeMCL [22] . The main difference between our analysis and that of Kurokawa et al . is that they augment their human gut metagenomic ORF orphan set with orphans from other metagenomic environments . We believe the main benefit of metagenomic sequencing is that protein families related to specific environments can be targeted . These environmentally specific signals may have been lost by adding sequences from other environments . In our study focused on identification of novel and uncharacterized protein families we used the procedure described below ( outline of the procedure is also give in a separate table ( T1 ) in the supplement materials ) . We used the set of metagenomic sequences prepared as described in the Data Preparation section earlier . It includes 303 , 314 complete ORFs . In the first step of the analysis we removed metagenomic sequences that belong to families annotated in PfamA database . This was done by masking all fragments which were aligned with HMMs representing PfamA families ( we used hhmscan from HMMER package [23] ) . Subsequently we identified uncharacterized and putative novel families in the remaining ( i . e . unmasked ) sequences . Our method to produce automatically derived human gut–related uncharacterized sequence clusters yielded 1 , 761 putative protein families . By analyzing the families , we found several common flaws and characteristics that created less-than-optimal automatic family descriptions . Initially , by taking the target HMMs and rescanning the metagenomic dataset , we found that 112 models produced no hits within the cutoff value . To eliminate bad data or uninteresting results , we set up a series of criteria to filter possible families . One of the most common problems was identification of families that had not been fully detected by the existing PfamA models but , in fact , were branches of annotated PfamA families . After applying FFAS to detect similarities between families found by the clustering technique and PfamA families , we removed families with a good probability of being linked to existing PfamA families . This initial filter reduced the set of families to 1 , 250 . The next step was to remove possible clusters of incomplete proteins . We found many cases of sequence fragment clusters that were , in fact , associated with PfamA families , but not described by the HMM representing a family . In these cases , the PfamA model only describes the most conserved subsection of the family alignment and by clustering , we collected sequence fragments left outside the HMM . To filter these clusters , we accepted only families that could be aligned with at least 75% of at least one protein found in a known genome . This filter reduced the number of families to 486 . We also filtered using the length of the HHM representing the proposed family to remove sequences unlikely to represent full protein domains . After removing families represented by HMMs shorter than 100 elements , the number of families dropped to 317 . Another problem was the presence of HMMs that produced fewer hits in the metagenomic sequences than the original set of sequences used to create them . By removing clusters with fewer then 10 metagenomic hits , the number of protein families dropped to 291 . All 291 candidates for new families were subsequently manually curated by inspecting Psi-Blast and FFAS results and the following cases were removed from the list: Manual curation eliminated 111 families reducing the number of families to 183 . Full list of the curated families is provided in the Table S3 in the Supplemental Materials . We proposed three parameters to evaluate overrepresentation of protein families in the human gut microbiome: comparative overrepresentation ( Ov ) , expansion ( Ex ) , and essentiality ( Es ) defined by the following formulae:where: Numerical values of each of overrepresentation , expansion , and essentiality were calculated for all new families identified by our analysis and also , separately for all families from the PfamA database ( see Supplementary Materials ) . The top-ranking PfamA families and new families are shown in Tables 1–3 and 4–6 , respectively .
|
Metagenomics provides a unique opportunity to sample the gene content of microbial communities adapted to specific environments and for the study of the correlations between the presence or absence of gene families that occur in organisms within that environment . Such studies provide detailed information about the adaptation of microbes to a given environment and , indirectly , provide clues about the most important molecular processes that are specific for that environment . Having performed such an analysis for the community of the human distal gut , we report many new protein families and identify many others that are highly specific for this particular environment . The function of most of these proteins is unknown , which illustrates the extent of our ignorance about the organisms within this environment that are so important for human health and well being .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"computational",
"biology/macromolecular",
"sequence",
"analysis",
"computational",
"biology/metagenomics",
"computational",
"biology/ecosystem",
"modeling",
"biochemistry/bioinformatics"
] |
2010
|
Expansion of the Protein Repertoire in Newly Explored Environments: Human Gut Microbiome Specific Protein Families
|
Some isolates of Yersinia pseudotuberculosis produce the cytotoxic necrotizing factor ( CNFY ) , but the functional consequences of this toxin for host-pathogen interactions during the infection are unknown . In the present study we show that CNFY has a strong influence on virulence . We demonstrate that the CNFY toxin is thermo-regulated and highly expressed in all colonized lymphatic tissues and organs of orally infected mice . Most strikingly , we found that a cnfY knock-out variant of a naturally toxin-expressing Y . pseudotuberculosis isolate is strongly impaired in its ability to disseminate into the mesenteric lymph nodes , liver and spleen , and has fully lost its lethality . The CNFY toxin contributes significantly to the induction of acute inflammatory responses and to the formation of necrotic areas in infected tissues . The analysis of the host immune response demonstrated that presence of CNFY leads to a strong reduction of professional phagocytes and natural killer cells in particular in the spleen , whereas loss of the toxin allows efficient tissue infiltration of these immune cells and rapid killing of the pathogen . Addition of purified CNFY triggers formation of actin-rich membrane ruffles and filopodia , which correlates with the activation of the Rho GTPases , RhoA , Rac1 and Cdc42 . The analysis of type III effector delivery into epithelial and immune cells in vitro and during the course of the infection further demonstrated that CNFY enhances the Yop translocation process and supports a role for the toxin in the suppression of the antibacterial host response . In summary , we highlight the importance of CNFY for pathogenicity by showing that this toxin modulates inflammatory responses , protects the bacteria from attacks of innate immune effectors and enhances the severity of a Yersinia infection .
Enteropathogenic Yersinia species such as Y . enterocolitica and Y . pseudotuberculosis initially infect the terminal ileum and colonize the Peyer's patches ( PPs ) within several hours of infections . Bacteria are subsequently transported to the mesenteric lymph nodes ( MLNs ) and can also spread systemically to reach liver and spleen via the bloodstream . The infections typically result in enteritis , enterocolitis and mesenteric lymphadenitis where the infected tissues show formation of microabscesses or granuloma-like lesions with central necrosis [1] . Enteropathogenic yersiniae have been shown to secrete exotoxins and/or inject effector proteins by specialized secretion machineries to manipulate host cell functions , including cytoskeletal rearrangements , to prevent immune responses and to establish a successful infection . They encode a type III secretion system ( T3SS ) on a 70 kb virulence-associated plasmid ( pYV ) that is essential for their defense against the host immune system [2]–[4] . The Yersinia T3SS has been shown to form a syringe-like apparatus with a thin needle-like surface exposed projection [5] . It is used to insert a translocation channel ( composed of YopB and YopD ) within the host membrane to inject the effector proteins YopE , YopH , YopJ/YopP , YopK/YopQ , YopM , YopO/YpkA , and YopT into the cell's cytoplasm . Yops target different cell signaling molecules and processes , in particular cytokine production and actin dynamics , often resulting in the inhibition of phagocytosis [6] . YopH is a tyrosine phosphatase that dephosphorylates proteins of the focal adhesion complex [7]–[10] . The effectors YopE , YopT and YopO/YpkA manipulate the regulation of Rho GTPases , which control the formation of lamellipodia , filopodia and stress fibers [2] , [11] . YopJ/YopP promotes cell death of macrophages by inactivating the counterregulators of the Toll-like receptor 4-triggered apoptotic pathway , the mitogen-activated protein kinase kinases ( MEKs ) and the inhibitor κB kinase β ( IKK β ) [12]–[16] . YopM forms a complex with RSK and PRK kinase isoforms , traffics to the nucleus , and is important for Yersinia to persist in liver and spleen with a contextual decrease of several proinflammatory cytokines , including IL-1β , IL-12 , IL-18 , interferon γ , and TNF-α , and depletion of NK cells [17]–[21] . The effector YopK/YopQ seems to play a role in orchestrating the translocation of effector proteins by modulating the ratio of the pore-forming proteins YopB and YopD . This appears to prevent unintended Yop delivery and neutrophil death , which would enhance the inflammatory response possibly favoring the host [22]–[24] . Insertion of the YopB/D translocation channel allows Yop delivery while maintain the host cell membrane intact . The YopB/D complex results in activation of Rho GTPases , actin polymerization and pore-formation . However , pore formation is usually prevented by the GTPase-downregulating function of YopE and YopT . Yet , expression of constitutively active forms of Rac1 and RhoA leads to a loss of membrane integrity and results in increased pore formation even when YopE and YopT are expressed [25] . In addition , signaling pathways triggered by high affinity-binding of the main Yersinia adhesins YadA and InvA to β1 integrin receptors and YopB/D signaling were shown to induce activity of Rho GTPases and actin polymerization which are crucial for efficient translocation of the Yop effectors [26] . Another Yersinia factor shown to activate the small GTPase RhoA is the cytotoxic necrotizing factor-Y ( CNFY ) [27] , [28] . CNFY is prevalent in some Y . pseudotuberculosis isolates , e . g . the widely used Y . pseudotuberculosis strain YPIII . All these strains belong to the serogroup III , but other isolates of this serogroup do not express CNFY and contain deletions within the corresponding cnfY gene [27] . On the amino acid level , CNFY is highly similar ( >68% ) to the CNF toxins found mainly in E . coli strains isolated from patients and domestic animals with extraintestinal infections ( CNF1-3 ) [29] , [30] . CNF1 is the best-characterized toxin of this class of bacterial toxins and is transferred to host cells through outer membrane vesicles ( OMVs ) [31]–[34] . The CNF1 toxin is a single-chain A-B toxin with an N-terminal delivery domain including subdomains for receptor binding , pore formation and proteolytic cleavage , and a C-terminal deamidase domain [35] , [36] . Internalization of the toxin into target cells occurs through receptor-mediated endocytosis , which appears to be independent of clathrin and lipid rafts ( sphingolipid/cholesterol rich microdomains ) [37] , [38] . After uptake , the 55 kDa C-terminal deamidase domain is autocatalytically cleaved off in the late endosome , and delivered into the cytoplasm in a pH-dependent manner [39] . CNF1 deamidates Gln-61/-63 of RhoA , Rac1 and Cdc42 to Glu-61/-63 resulting in Rho GTPases with a blocked GTP hydrolase activity . Deamidated Rho GTPases induces polymerization of F-actin at focal contacts , increase cell-matrix adhesion , and promote formation of stress fibers , lamellipodia and filopodia , which led to the classification as ‘constitutively active’ [40]–[45] . Cytoskeletal rearrangements attributed to CNF lead to multinucleated cells due to inhibited cytokinesis with ongoing cell cycle progression [46] . Additionally , CNF1 has been reported to ( i ) induce phagocytosis in epithelial cells and reduce CR3-mediated phagocytosis in monocytes [47] , [48] , ( ii ) promote bacterial cell entry [49] , ( iii ) decrease the barrier function of intestinal tight junctions [40] , [50] , ( iv ) decrease transmigration of polymorphonuclear leukocytes across a T84 monolayer [51] , and ( v ) induce apoptosis of bladder cells [52] . The overall amino acid sequence of CNFY of Y . pseudotuberculosis is very similar to CNF1 . However , CNFY is not recognized by neutralizing antibodies against CNF1 [27] . Moreover , CNFY seems to bind to different cell receptors and preferentially deamidates RhoA ( over Rac1 and Cdc42 ) in cultured epithelial cells [28] , [38] . Although CNFY and certain Yop effectors alter the cytoskeleton by affecting the activity of the Rho GTPases , little is known about the interplay , cooperation and joint role of these toxins in the pathogenic lifestyle of Y . pseudotuberculosis . Here , we provide evidence that CNFY is an important virulence factor of Y . pseudotuberculosis YPIII . CNFY is shown to enhance Yop protein delivery , which is crucial for pathogenicity . Furthermore , the toxin was found to induce inflammatory responses and increase the severity of a Yersinia infection .
Since many Y . pseudotuberculosis isolates as well as Y . pestis contain deletions within the cnfY gene [27] , we first tested whether the intact cnfY toxin gene in the Y . pseudotuberculosis wild-type strain YPIII is expressed and induced under virulence-relevant growth conditions . A cnfY-lacZ transcriptional fusion was only slightly expressed when Y . pseudotuberculosis was grown at 25°C , but its expression was strongly induced at 37°C and reached its maximum during stationary phase ( Fig . S1A ) . High cnfY transcription was generally observed in complex media , in particular BHI , whereas only low expression levels were detected in all tested minimal media ( Fig . S1B , data not shown ) . In summary , cnfY is predominantly expressed at 37°C in a nutrient rich environment , resembling conditions found in the mammalian intestinal tract . This result prompted us to test expression of the toxin during infection . BALB/c mice were orally infected with 2×108 bacteria of the Y . pseudotuberculosis wild-type strain YPIII expressing a cnfY-luxCDABE fusion , and the bioluminescent signal was monitored in the mice for six days using an in vivo imaging system . Only very low luciferase activity was measured in the bacterial culture before infection ( data not shown ) and in the intestinal tract directly after oral ingestion ( 1 h , Fig . 1 ) . However , a very strong bioluminescent signal of the cnfY-luxCDABE fusion was detectable during the entire following course of the infection . The most intensive signals were detected two days post infection in the intestine and associated lymphoid tissues ( Fig . 1 ) . No light emission was monitored in mice infected with bacteria carrying the promoterless luxCDABE operon in the identical expression system ( data not shown ) . In order to study cnfY expression in the individual infected tissues , we used a set of established fluorescent fusion vectors for in vivo expression analysis . To do so , Y . pseudotuberculosis YPIII harboring a plasmid-encoded constitutive PgapA::dsred2 reporter construct and a compatible PcnfY::gfpmut3 . 1 fusion was used to infect BALB/c mice . Five days post infection , the small intestine , caecum , colon , PPs , MLNs , spleen and liver were isolated and cryosections were prepared . The bacteria in the tissues were visualized by monitoring dsRed2 , and then tested for PcnfY::gfpmut3 . 1 . As shown in Fig . 2 , the PcnfY::gfpmut3 . 1 fusion was expressed in all tested organs . In summary , a temperature shift to 37°C , but most likely no tissue-specific signals are required to induce toxin expression in infected tissues . Absence of a functional toxin gene in other Y . pseudotuberculosis clinical isolates , may suggest that CNFY only adds another potential virulence factor to the variety of effector proteins and toxins that are produced by this pathogen . However , high expression of cnfY during the entire course of an infection also indicates that presence of this toxin may enhance the pathogenicity of Y . pseudotuberculosis . To first assess the impact of CNFY on pathogenesis , the potential of the Y . pseudotuberculosis wild-type strain YPIII and the isogenic cnfY-deficient strain to cause lethal infections was compared . BALB/c mice were orally infected with 2×109 bacteria of the cnfY mutant ( YP147 ) and the wild-type strain ( YPIII ) harboring the empty vector ( pJNS11 ) or a cnfY-encoding plasmid ( pJNS10 ) . Survival and weight of the mice were monitored over two weeks and date of death was recorded ( Fig . 3 , S2 ) . Mice infected with YPIII showed signs of the infection , e . g . weight loss , piloerection and lethargy , and succumbed to infection between day four and day six . Strikingly , none of the mice infected with YP147 developed severe disease symptoms and all mice were still alive 14 days post infection . Monitoring of body weight demonstrated that also mice infected with the cnfY knock-out strain YP147 showed a slight reduction in weight , but they recovered quickly and regained weight ( Fig . S2 ) . Presence of the cnfY-encoding low-copy number plasmids reverted the avirulent phenotype of the cnfY mutant and reduced the average day of death of the wild-type strain YPIII by one day , most likely due to the overexpression of the toxin . The Y . pseudotuberculosis YPIII isolate , unlike other Y . pseudotuberculosis strains , is unable to replicate in murine macrophages due to a defective allele of phoP [53] . To exclude that CNFY influence on virulence is only visible in a phoP-deficient derivative with an overall lower pathogenicity , the inability to grow in macrophages was complemented by an exchange of the allele against the phoP ORF from Y . pseudotuberculosis IP32953 . However , when mice were challenged with 2×109 CFU of the equivalent phoP+ strains , 100% of the mice infected with the CNFY-positive strain died during the observation period , while 80% of the mice infected with the isogenic cnfY-deficient strain survived and regained weight ( Fig . S3 ) . To gain a deeper insight into the differences in the infection process of CNFY-positive and -negative strains , we determined the number of bacteria that colonized the small intestine , caecum , PPs , MLNs , liver and spleen of BALB/c mice at different time points after oral infection with 2×108 bacteria ( Fig . 4 ) . Comparable amounts of wild-type ( YPIII ) and the mutant strains ( YP147 ) were recovered from PPs and caecum during infection , and only a very small increase of bacterial counts was observed with the cnfY mutant in the small intestine at days 5–7 post infection ( Fig . 4 ) . However , significantly reduced numbers of YP147 were recovered from MLNs and spleen ( Fig . 4 ) . The number of cnfY-positive and -negative bacteria in these organs was almost identical up to day three post infection , but the cnfY mutant was eliminated very rapidly later during the infection . At day seven , none or only few mutants were recovered from MLNs and spleen , whereas 108–109 bacteria of the wild-type strain were recovered per gram of both organs . The effect was less pronounced in the liver , but the strongly reduced number of mutant bacteria relative to the wild-type bacteria six and seven days after infection clearly indicated that the presence of CNFY is also advantageous for the colonization of the liver ( Fig . 4 ) . This demonstrated that loss of CNFY , resulting in avirulence of Y . pseudotuberculosis YPIII , is reflected by a fast elimination of the bacteria from MLNs , liver and spleen . Within the first week after infection with wild-type strain YPIII the size of the spleen and liver decreased two-fold , whereby changes of the organ size were first visible at day three post infection ( Fig . S4A , B ) . In contrast , infection with the isogenic cnfY mutant strain YP147 had no effect on the size of the liver and induced a considerable increase of the size of the spleen . In addition , mice infected with wild-type strain YPIII had significantly shorter intestines ( 30% ) at day six and seven post infection than mice infected with the cnfY mutant ( Fig . S4C ) . The shortening of the intestine is a sign of marked intestinal inflammation . This indicated that CNFY not only affects colonization of systemic organs , but has also a strong influence on the host's inflammatory response against the bacterial infection . Histopathological examination of the infected host tissues demonstrated marked differences of the overall inflammatory reaction , which was stronger in YPIII-infected animals , especially in the small intestine and spleen compared to YP147-infected mice . In the intestine , inflammation was most prominent in the ileum and caecum in both groups . However , in YPIII-infected mice inflammation was diffuse affecting the entire ileum at day six ( Fig . 5A , upper panel , 5B middle panel ) . In YP147-infected mice inflammation was locally restricted to multifocal lesions characterized by the presence of inflammatory cells from the muscular layer up to the epithelial cells ( Fig . 5A , lower panel , 5B right panel ) . In these areas inflammation led to epithelial cell hyperplasia ( increased proliferation ) resulting in an increase of the villi length . However , this lesion is only locally restricted and adjacent tissue remains unaltered . In addition , inflammation was more generalized in YPIII- compared to YP147-infected organs . In mice infected with YP147 , no bacterial foci ( diffuse patches of bacteria ) could be detected microscopically in hematoxylin and eosin ( H & E ) stained sections of the spleen at day six post infection , whereas in the majority of YPIII-infected mice bacterial foci were visible in the histological sections ( Fig . 5C ) . YPIII infections were accompanied by a more severe inflammation of the spleen , where presence of the bacteria resulted in necrotizing spleenitis leading to splenic atrophy with marked depletion of the white pulp . YPIII caused multifocal necrosis in spleen , whereas in YP147-infected spleens , only mild hyperplasia of the white pulp and increased erythropoiesis were found ( Fig . 5C ) . Taken together , CNFY has a significant influence on the number of microcolonies in the tissues and leads to a more severe and widespread inflammation in the small intestine , liver and spleen . Because of the strong influence of CNFY on the colonization of bacteria in MLNs , spleen and liver , it was hypothesized that the toxin might counteract host immune defenses . To test this hypothesis , we infected BALB/c mice with 2×108 bacteria of the wild-type or the cnfY mutant strain , and immune cell composition in the spleen was analyzed by multi-color flow cytometry three days and six days post infection . The spleen was chosen since here the most pronounced CNFY-triggered pathological effects had been observed . Cell suspensions of isolated tissues were prepared and cells were stained with fluorescently labeled antibodies to distinguish neutrophils from macrophages/monocytes , dendritic cells ( DCs ) , natural killer ( NK ) cells , B cells , and T cells ( Fig . S5 ) . All alterations of immune cell populations seen at day six ( data not shown ) were already visible at day three post infection , when the bacterial load is still similar and the overall health status of YPIII-infected mice is only slightly and not severely reduced as at day six . A very pronounced variation of the immune cell population between the YPIII- and YP147-infected mice was observed ( Fig . 6 ) . All types of immune cells were significantly decreased in the spleen three days after infection with YPIII when the spleen started to shrink , but the most severe changes were observed with cells of the innate immune system . In particular , numbers of macrophages , monocytes and NK cells were significantly reduced; whereas reduction of neutrophils and conventional DCs was less pronounced . In contrast , no reduction of immune cells was detectable in spleens of YP147-infected mice ( Fig . 6 ) . In contrast , a significant higher influx of neutrophils and macrophages/monocytes was observed , which is consistent with the rapid clearance of mutant bacteria from the spleen upon triggering of the immune response . To determine whether CNFY affects the steady-state level of certain cell populations , the population percentage was also compared and further confirmed a significant expansion of neutrophils and macrophages/monocytes in YP147-infected spleens ( Fig . S6 ) . These and the histopathological data strongly suggest that the CNFY toxin reduces influx and/or causes rapid cell death of invading immune cells in the spleen . Our infection experiments clearly demonstrated that absence of the CNFY toxin renders the bacteria completely avirulent , resulting in the clearance of the bacteria in MLNs , liver and spleen . A similar attenuation in mouse models of oral infection was observed ( i ) when the virulence plasmid , encoding the T3SS and the Yop effectors is cured from Y . pseudotuberculosis YPIII , ( ii ) when multiple yop genes were deleted or ( iii ) when the regulator LcrF that controls expression of the T3SS/Yops is absent [54] , [55] . Moreover , a significant influx of neutrophils was observed in the spleen of mice infected with a yopM mutant strain of Y . pestis , while the numbers of neutrophils decreased during infection with the parental strain [20] , [21] . In addition , YopJ translocation has been shown to promote cell death of professional phagocytes [13] , [15] . This suggested that the CNFY toxin is important for the efficient injection of the Yop effectors into host cells during the infection process . In fact , recent work by Mejia et al . [26] demonstrated that efficient translocation of the Yop effectors requires Rho activation – a process that has been shown to be stimulated by the CNFY toxin [28] , [38] . To address whether CNFY-mediated activation of Rho GTPases influences Yop-translocation into professional phagocytes , we first tested the influence of recombinant CNFY toxin on non-activated and PMA-activated macrophages , thus mimicking its effect on unstimulated and stimulated macrophages during infection . Intoxification of murine macrophages ( J774A . 1 ) led to activation of all three Rho GTPases , RhoA , Cdc42 and Rac1 ( Fig . 7A ) . CNFY further induced a marked increase in cell size with some giant multinucleated cells ( Fig . 7B ) . These CNFY effects occurred independently of macrophage stimulation with PMA . This indicates that CNFY controls actin dynamics in macrophages through deamidation of Rho GTPases . Since host actin polymerization by Rho activation plays a role in Yop translocation by Y . pseudotuberculosis [26] we also tested the influence of CNFY on Yop delivery . To do so , we generated Y . pseudotuberculosis strains expressing a YopE-β-lactamase reporter fusion [56] , namely YP173 ( YPIII-ETEM ) , YP174 ( YP101ΔsycS-ETEM ) , and YP217 ( YP147ΔcnfY-ETEM ) , and used these strains to infect host cells treated with the dye CCF4-AM . CCF4-AM consists of coumarin and fluorescein conjugated by a lactam ring and is modified by cellular esterases , whereby the dye becomes green fluorescent and is trapped inside the cell . If the β-lactam ring is cleaved by β-lactamase the dye changes its fluorescence from green to blue [57] , [58] . The green to blue conversion allows identification of host cells in which the YopE-β-lactamase fusion protein has been successfully injected . We first used this fluorescence-based system to monitor translocation of the chimeric protein into HEp-2 cells , and determined the number of green and blue fluorescent cells by fluorescence microscopy and flow cytometry . Efficient translocation of YopE-β-lactamase into epithelial cells was observed upon infection with YP173 ( YPIII-ETEM ) , but not with the secretion-deficient control strain YP174 ( YP101ΔyscS-ETEM ) ( Fig . 8A , B ) . YopE-β-lactamase translocation by the cnfY-deficient strain YP217 ( YP147ΔcnfY-ETEM ) was significantly reduced compared to YP173 ( YPIII-ETEM ) , whereas preincubation of the host cells with CNFY increased translocation of the fusion protein ( Fig . 8A , B ) , indicating that CNFY enhances effector delivery . Since Y . pseudotuberculosis predominantly injects the Yops into professional phagocytes in vivo [59] , we also tested CNFY influence on YopE-β-lactamase translocation into murine macrophages , and found that pretreatment with CNFY also boosts Yop delivery into these phagocytes ( Fig . 8C ) . Stimulation of Rac1 through YadA and invasin-bound β1-integrins was shown to be essential for Yersinia uptake into epithelial cells [11] , but neither internalization nor activation of Rac1 was required for Yop translocation by Y . pseudotuberculosis into HeLa cells [26] . This suggested that CNFY-mediated stimulation of Yop delivery into macrophages might preferentially be caused by activation of RhoA . To validate this assumption , we pretreated macrophages with the Clostridium botulinum C3 toxin , an ADP-ribosylating protein that specifically inhibits RhoA , B and C , or with toxin B from variant Clostridium difficile serotype F strain 1470 ( TcdBF ) , which specifically inhibits Rac but not RhoA/B/C [60] , [61] . Treatment with the toxins induced actin cytoskeleton rearrangements and cell morphology changes , but had no effect on the viability of the macrophages and the number of associated bacteria ( data not shown ) . As shown in Fig . 8C , the RhoA/B/C inhibitor reduced the percentage of blue macrophages significantly , whereas the Rac inhibitor had no influence on YopE-β-lactamase translocation . These findings indicated that the CNFY toxin enhances Yop delivery into murine macrophages , and in particular activation of RhoA seems to play a role in the processes that stimulate Yop translocation into these professional phagocytes . It has been reported that translocated effector YopE of Y . pseudotuberculosis YPIII is a GTPase-activating protein ( GAP ) for Rac1 and RhoA and this function appears important to regulate Yop translocation and modulate host defenses crucial for virulence [62]–[65] . This raised the question how YopE and CNFY contribute to RhoA-GTP and Rac1-GTP levels and Yop translocation . To address this , we analyzed RhoA and Rac1 activation and Yop translocation in the presence and absence of YopE in untreated or CNFY-pretreated murine macrophages . As shown in Fig . S7 , only low amounts of active Rac1 and RhoA could be detected in uninfected macrophages . Addition of the wild-type strain YPIII pregrown at 37°C to mimic the situation prior to host cell contact induced activation of RhoA and Rac1 . Absence of YopE resulted in a small additional increase in RhoA-GTP , but had no or only a slight influence on Rac1-GTP levels . Furthermore , it had no or only a very small stimulatory effect on the translocation of YopD and YopH without or after pretreatment of the macrophages with CNFY ( Fig . S7 ) . This indicates that under these conditions intracellular YopE is unable to efficiently counteract CNFY-mediated RhoA/Rac1 activation and reduce Yop translocation into murine macrophages . We next analyzed whether the CNFY toxin affects Yop translocation into host cells in the original tissue environment . MLNs were harvested from uninfected mice and filtered to disrupt the tissue architecture and generate single-cell suspensions . Single cell suspensions were infected with a multiplicity of infection ( MOI ) of 10 , incubated with CCF4-AM , and then analyzed by flow cytometry . As shown in Fig . S8 , significantly higher numbers of blue cells with translocated YopE-β-lactamase were measured after infection with YPIII , indicating that Yop delivery into host cells can be enhanced by the toxin through activation of Rho GTPases . It has been previously reported that Y . pseudotuberculosis selectively targets Yops to professional phagocytes in the PPs , MLNs and spleen during the oral route of infection [59] . To analyze whether the CNFY toxin also affects YopE-β-lactamase delivery in the course of an infection , we orally infected mice with 2×109 bacteria YP173 and the isogenic cnfY mutant strain YP217 . The T3SS-deficient yscS mutant , encoding the YopE-β-lactamase , and YPIII without the fusion were used as negative controls . At day three post infection mice were sacrificed , the PPs , MLNs , and spleen were harvested , and the translocation of Yops into various immune cell subsets was analyzed by flow cytometry ( Fig . S9 ) . Following infection with the YopE-β-lactamase expressing wild-type strain 4 . 5% of all living cells within PPs were affected by Yop translocation . In contrast , only 1 . 5% of all living cells in the PPs contained the fusion protein after infection with the cnfY-deficient strain ( Fig . S10 ) . Yop translocation efficiency was still significantly reduced in tissues infected with the cnfY-deficient strain when the percentage of translocated blue cells was normalized to the bacterial load of the tissue/organ ( Fig . 9 ) . This excludes that lower bacterial numbers account for this effect , but it also assumes that bacteria are infecting different cell types at the same MOI , which is unknown . Yop delivery was also significantly lower in the absence of the CNFY toxin in MLNs and spleen in which the total number of targeted cells was reduced compared to PPs ( Fig . 9 , S10 ) . We further determined whether CNFY-mediated stimulation of Yop translocation affected specific immune cells more frequently than others . Translocation of YopE-β-lactamase into each immune cell type was compared in MLNs and spleen from mice infected with YP173 ( YPIII-ETEM ) or the cnfY mutant derivative YP217 ( YP147-ETEM ) ( Fig . 9B , S10 ) . In general , all immune cells analyzed were targeted by Y . pseudotuberculosis . However , Yop-injected neutrophils were significantly enriched in the MLNs and the spleen , indicating that this cell population is preferentially targeted in the tissues . In addition , DCs , NK cells and macrophages were well represented in the blue population , while B and T cells remained underrepresented ( Fig . 9B , S10 ) . This is in full agreement with previous studies demonstrating that translocated YopH of Y . pseudotuberculosis strain IP2666 is enriched in neutrophils , macrophages and DCs in MLNs and spleen [59] . We further found that the apparent enhanced targeting to professional phagocytes , in particular neutrophils , macrophages and DCs and to a smaller extent also translocation into B and T cells was reduced in the absence of CNFY in the MLNs and spleen three days post infection ( Fig . 9B , S10 ) . Yop translocation into NK cells was also somewhat reduced in the spleen . Since Y . pseudotuberculosis induces host cell death [66] , which may be reflected in the strong reduction of professional phagocytes in the spleen ( Fig . 6 ) , the actual amount of Yop translocation in this organ is probably underestimated . Taken together , these results demonstrate that the CNFY toxin plays a critical role during the infection , facilitating targeting of Yops to host immune cells , in particular professional phagocytes .
Many bacterial toxins and translocated effector proteins target Rho GTPases , which control crucial eukaryotic signal transduction pathways involved in the organization of the cell cytoskeleton , cell cycle progression , genetic information processing , and host defense processes to promote invasion , survival and replication of pathogens within their hosts [29] , [67] . In this study we investigated the Rho-activating cytotoxic necrotizing factor CNFY of Yersinia . Although much progress has been made unravelling the molecular mechanism of this toxin , the functional consequences for host-pathogen interaction and pathogenesis were largely unknown . Using a murine model for gastrointestinal tract infections we provide evidence that this Rho-activating protein is crucial for virulence of the naturally toxin-expressing Y . pseudotuberculosis strain YPIII . The importance of CNFY for pathogenesis was first established by the analysis of the expression and the role of the toxin during the infection of mice . We show that cnfY is strongly expressed in all infected tissues during pathogenesis in mice , and is crucial for virulence , in particular for the dissemination of the bacteria into the MLNs , spleen and liver . Histological analysis and immune cell composition of the infected tissues suggest that CNFY contributes significantly to the acute characteristics of the inflammatory response and host tissue damage during infection . Histo-pathologic evaluation underlines the finding that CNFY induces apoptosis , as focal necrosis was not seen in YP147-infected animals . Cell death leads to atrophy of the spleen in YPIII-infected mice . Moreover , a restriction of the inflammation to small foci could be observed in the intestine of YP147-infected animals , whereas the entire ileum was affected by a diffuse inflammation in YPIII-infected animals , explaining the shortening of the intestine . Hyperplasia of the white pulp seen in YP147-infected mice displays the immune response triggered by the infection . The infection is restricted to small foci in the intestine and is reversible , whereas the infection in YPIII infected animals is generalized and most probably leads to death by endotoxiemia . This inflammatory necrotizing phenotype is reminiscent of earlier studies analyzing the effect of CNF1 of E . coli using subcutaneous injections as well as animal models of urinary tract and prostatitis infection [31] , [68] , [69] . Infections of the gastrointestinal tract by enteropathogenic Yersiniae lead to a biphasic inflammatory process in which bacterial adhesion and transmigration through the intestinal epithelial layer triggers an initial antibacterial defense response with little inflammation , e . g . expression of IL-8 by epithelial cells , which is followed by an acute infiltration and activation of neutrophils , cytokine production and tissue necrosis [70] . First recognition of Y . pseudotuberculosis occurs through contact of the bacterial LPS with TLR4 on naïve host macrophages and this leads to proinflammatory cytokine production through activation of MAPK and NF-kB . However , translocation of YopJ inhibits activation of MAPK and NF-κB and induces an apoptotic signaling pathway including activation of initiator caspase-8 , and the executioner caspase-3 , -7 , and -9 [70] , [71] . Apoptotic macrophages are eliminated and this process also triggers production of anti-inflammatory cytokines such as IL-10 and TGF-β [72] , [73] . However , induction of apoptosis is probably not always fully immunologically silent , e . g . phagocytosis of apoptotic cells by other phagocytes , can prime other immune responses such as activation of CD8+ T cells [71] . During the course of the infection , the number of activated macrophages increases whereas the number of naïve macrophages declines . In activated macrophages Yersinia causes cell death by inflammatory pyroptosis . This occurs through activation of a multiprotein complex , called the inflammasome , which forms a platform for the autoprocessing and activation of the cysteine protease caspase-1 . Activation of caspase-1 results in the secretion of the inflammatory cytokines such as IL-1α , IL1-β , and IL-18 , and triggers cell death [66] . LPS , the T3SS and the translocated YopJ protein of Yersinia were shown to induce caspase-1 activation and pyroptosis [71] , [74] . Induction of pyroptosis ( inflammatory death ) in activated macrophages corresponds to later stages of the infection with Y . pseudotuberculosis , where inflammation and necrosis is evident from histopathology . Based on our results it is very likely that CNFY supports Yersinia-induced pyroptosis of activated phagocytes in the spleen during later stages of the infection . CNFY was shown to manipulate the number of immune cells and induce inflammatory responses . The number of macrophages , monocytes and neutrophils decreased strongly ( 4- to 15-fold ) in all lymphoid organs three days post infection . In contrast , infection with the cnfY mutant resulted in no reduction , but rather in an increase of phagocytes when compared to uninfected control mice , and the overall inflammation of the infected tissues was considerably reduced . Similar to CNF1 of E . coli [34] , it is possible that CNFY is transported by OMVs , which act as long-range toxin delivery vectors , and is then able to reduce chemotaxis and influx of professional phagocytes by constitutive Rho GTPase activation . We further demonstrate that CNFY enhances Yop delivery into phagocytes during infection . This strongly suggests that increased translocation of YopJ could stimulate cell death in the spleen . Consistent with this assumption , YopJ and CNFY promote systemic dissemination following oral infection . Work by Monack et al . [75] showed that a yopJ mutant is deficient for spread from the PPs to other lymphoid tissues ( MLNs , spleen ) , similar to the cnfY mutant investigated in this study . Moreover , wild-type Yersinia induce apoptosis of macrophages from infected spleens [75] , implying that mainly YopJ is used to eliminate immune cells in the spleen to dampen the immune response against Yersinia during infection . YopJ was also shown to subvert the NOD2/RICK/TAK1 pathway , activate caspase-1 and induce IL-1β secretion within PPs , which is associated with increased barrier permeability [76] . This suggests that CNFY action also enhances YopJ-dependent intestinal barrier disruption and promotes the dissemination of Yersinia by exploiting the mucosal inflammatory response . In addition , CNFY seems to contribute to depletion of NK cells in the spleen . It was observed that Y . pestis but not an isogenic yopM deficient mutant caused a significant global decrease in NK cell numbers [21] , indicating that NK cell depletion is enhanced by CNFY-mediated activation of YopM translocation . Here , we observed that RhoA , Rac1 and Cdc42 are activated in CNFY-intoxicated macrophages , which is reflected by the high content of actin cables/stress fibres , the formation of lamellipodia and filopodia , pronounced cell spreading and inhibited cytokinesis . Previous work demonstrated that CNFY predominantly activates RhoA in epithelial cells [27] , [28] . However , a very recently published study also reported CNFY-mediated activation of Rac1 and Cdc42 in HeLa cells [77] . Use of different toxin concentrations , incubation times and cell types ( human epithelial cells versus murine macrophages ) in which the CNF toxins may display a different selectivity and different efficiencies of cell toxifications are likely to account for these variations . In fact , CNFY-mediated Rho GTPase activation pattern varies during intoxication whereby RhoA activation is generally more pronounced than Rac1 and Cdc42 2–3 h after toxin addition ( J . Schweer , unpublished results , [77] ) . This suggests that at very early time points and/or under low toxin concentrations predominantly RhoA might be activated . A previous report demonstrates that Y . pseudotuberculosis selectively modulates RhoA activity ( e . g . by signals triggered by the YopB/D translocon and/or from engagement of β1-integrin receptors ) to induce cellular changes that control T3SS pore formation and effector translocation [26] . Here , we strengthen this observation , as CNFY-mediated stimulation of Yop delivery of Y . pseudotuberculosis was sensitive to the Rho inhibitor C3-transferase of C . botulinum , but insensitive to Rac1 inhibition by TcdBF toxin of C . difficile . In contrast , new experiments addressing the influence of CNFY on Yop translocation of Y . enterocolitica demonstrated that CNFY also stimulates effector delivery by this pathogen , although this process seemed entirely dependent on Rac and not on Rho GTPases [77] . Different YadA/InvA-promoted signalling events , differences in Yop protein abundance ( e . g . RhoA-inactivating YopT is absent in YPIII ) and differences in the regulation of Yop delivery by Rho GTPases between the different species may be responsible for this discrepancy . Some effector proteins , in particular YopE , were shown to inhibit Yop delivery by inactivation of RhoA and Rac1 most likely as part of an intra-cellular control mechanism which measures and adapts the amount of protein translocated by Yersinia during infection . This is reflected by elevated levels of Yop effector translocation into epithelial cells by yopE-deficient strains [11] , [63]–[65] , [78] . Our analysis demonstrated that absence of YopE caused no or only a small increase in Rac1/RhoA activation and Yop translocation during infection of murine macrophages with Y . pseudotuberculosis YPIII with or without treatment with CNFY . This indicates that intracellular YopE is not able to counteract CNFY in these phagocytes . Recently published work showing that none of the Rho inhibiting effectors ( YopE , YopT and YopO ) could reduce the effect of CNFY on Yop translocation by Y . enterocolitica into human epithelial cells supports this observation [77] . However , we cannot exclude the possibility that other conditions ( e . g . conditions which enhance ( i ) deamidation and subsequent ubiquitin-dependent degradation of the modified Rho GTPases or ( ii ) YopE translocation and activity ) allow counterregulation . Very recently , it has also been reported that NOD1 , a pattern recognition receptor that senses cytosolic microbial products similar to NOD2 , monitors the activation state of all three Rho GTPases . Activation of Rho GTPases triggered the NOD1 signalling cascade with consequent RIP2-mediated induction of NF-κB-dependent inflammatory responses [79] . NOD1 activation was triggered by activation of Rac1 and Cdc42 by the Salmonella effector SopE . In line with this , all three Rho GTPases , Rac1 , RhoA and Cdc42 were previously shown to activate the NF-κB pathway [80] and particular Rac1 has been reported to contribute to NF-κB activation by CNF1 of E . coli by clustering the NF-κB inhibitor IκBa and components of the IκBα E3-ubiquitin ligase into membrane ruffles [79] . Based on our current knowledge we envision a model in which the CNFY toxin exerts its function in a multi-step process ( Fig . 10 ) . The first step corresponds to the uptake of the CNFY toxin by infiltrating innate immune cells ( e . g . neutrophils , macrophages , DCs ) in the early phase of the infection process . This triggers activation of the Rho GTPases , in particular RhoA , in the phagocytes . Induced actin polymerization resulting from Rho GTPase activation enhances Yop delivery into host cells to counteract innate and adaptive immune responses . As a consequence , invading immune cells are inhibited and undergo apoptosis leading to uncontrolled proliferation of the pathogens . Higher CNFY toxin concentrations by replicating pathogens potentiate activation of RhoA , Rac1 and Cdc42 which triggers inflammatory responses e . g . via the NOD1-RIP2 signalling cascade . In addition , interaction of Yersinia with increased numbers of activated macrophages causes cell death by inflammatory pyroptosis leading to strong inflammation and necrosis of the organs during later phases of the infection process . In summary , our data identify CNFY as an important Rho GTPase-activating toxin which is instrumental for Yersinia to amplify crucial virulence factor functions which determines the success of the infection and the severity of the associated disease .
All animal work was performed in strict accordance with the German Recommendations of the Society for Laboratory Animal Science ( GV-SOLAS ) and the European Health Recommendations of the Federation of Laboratory Animal Science Associations ( FELASA ) . The animal protocol was approved by the Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit: animal licensing committee permission no . 33 . 9 . 42502-04-055/09 . Animals were handled with appropriate care and welfare , and all efforts were made to minimize suffering . The strains used in this study are listed in Table 1 . Overnight cultures of E . coli were routinely grown at 37°C , Yersinia strains were grown at 25°C or 37°C in LB ( Luria-Bertani ) broth . The antibiotics used for bacterial selection were as follows: carbenicillin 100 µg/ml , chloramphenicol 30 µg/ml , kanamycin 50 µg/ml , and gentamicin 50 µg/ml . For infection experiments , bacteria were grown at 25°C or 37°C , washed and diluted in PBS prior to infection . For invasion assays and mouse infections , bacteria were grown to stationary phase , washed and resuspended in PBS . For the in vitro Yop delivery assay , bacteria were grown in LB medium at 37°C . Human HEp-2 cells were cultured in RPMI 1640 media with an alternative to L-glutamine with increased stability ( Invitrogen ) supplemented with 7 . 5% newborn calf serum ( Sigma Aldrich ) . Murine J774A . 1 macrophages were cultured in the same medium supplemented with 5% fetal calf serum ( PAA ) . All cell lines were cultivated at 37°C in the presence of 5% CO2 . All DNA manipulations , PCR , restriction digestions , ligations and transformations were performed using standard techniques as described previously [81] , [82] . Plasmids used in this study are listed in Table 1 . Plasmid pJNS01 ( cnfYhis6 ) was constructed by amplification of cnfY ( ypk_2615 ) from genomic DNA of Y . pseudotuberculosis YPIII with primers II795/II795 and integrated into the XhoI/NheI sites of pET28a ( Novagen ) . For construction of the luxCDABE , gfpmut3 . 1 and lacZ reporter gene fusions encoded on plasmids pJNS2-4 , the promoter region of cnfY ( primer pair II896/II898 ) was amplified and ligated into the BamHI/SalI sites of pFU54 , pFU58 and pFU68 . Plasmid pJNS10 , containing the cnfY promotor region with the cnfY gene , was constructed by the insertion of a PCR fragment amplified with primers II896 and IV16 from chromosomal DNA of Y . pseudotuberculosis YPIII into the BamHI/NotI sites of pFU234 . All clones were confirmed by sequencing and restriction . For generation of the mutant strain YP147 , a cnfY::KanR mutation encoded on plasmid pJNS05 was constructed . To do so , the kanamycin resistence gene was amplified using the kan primers ( I661/I662 ) and plasmid pACYC177 as template . Next , Y . pseudotuberculosis YPIII genomic DNA was used as a template to amplify 500-bp regions flanking the target gene cnfY . The upstream fragment was amplified with the primer pair III710/715 of which the reverse primer contained 20 nt at the 5′-end which are homologous to the start of the kanamycin resistance gene . The downstream fragment was amplified with primer pair III712/III714 of which the forward primer contained additional 20 nt at the 3′-end which are homologous to the end of the kanamycin resistance gene . A PCR reaction was performed with the forward primer and the reverse primer using the upstream and downstream PCR products of the target gene and the kanamycin gene fragment as templates and cloned into pAKH3 . The resulting plasmid pJNS05 was integrated into the cnfY locus of YPIII via conjugation as described [83] . Chromosomal integration of the fragments was selected by plating on LB supplemented with kanamycin . Excision of the plasmid including the defective cnfY allele of YPIII was obtained by plating of the strain on 10% sucrose and generated strain YP147 was analyzed by PCR and DNA sequencing . For the construction of strain YP56 ( ΔphoPQ ) , a phoPQ::KanR PCR fragment was generated . For this purpose , the kanamycin resistance cassette was created using the primers 360/361 and pACYC177 as template . The primers contain homology regions ( 20 nt ) to the upstream or downstream region of the phoPQ gene . A fragment including sequences of the phoPQ upstream region was generated by PCR using the primers 538/539 , a fragment including sequences of the phoPQ downstream region was amplified using the primers 540/541 . Primer 539 and 540 contain 20 nt of the kanamycin resistance cassette . A PCR fragment consisting of these three fragments was amplified using the primers 538/541 . The product was transformed into Y . pseudotuberculosis YPIII pKOBEG-sacB and a phoPQ::KanR mutant ( YP56 ) was generated and selected for as described [84] . For the construction of YP149 , the phoPQ gene of strain IP32953 was amplified by PCR with primer III926 and III927 , creating SacI and XhoI restriction sites . The fragment was cloned into plasmid pDM4 , generating vector pVP1 . Integration of the plasmid was obtained through conjugation of strain S17λpir pVP1 with the YP56 as described [83] . Excision of the plasmid including the defective phoPQ allele was obtained by plating of the strain on 10% sucrose . PCR and DNA sequencing proved presence of the intact phoPQ allele . To generate strain YP188 the cnfY gene was destroyed as described above for YP147 ( ΔcnfY ) . Strains YP173 , YP174 and YP217 were constructed by chromosomal integration of the YopE-β-lactamase ( ETEM ) fusion plasmid pSR47s-E-TEM1 into the yopE locus . Integration was obtained through conjugation of E . coli K-12 strain S17λpir pSR47s-E-TEM1 with the Y . pseudotuberculosis strains YPIII , YP101 ( ΔyscS ) and YP147 ( ΔcnfY ) as described [83] . For generation of the mutant strain YP275 , a yopE::KanR mutation encoded on plasmid pJNS13 was constructed as described above for YP147 . Primer V553/V554 and V555/V556 were used to amplify 500-bp regions flanking the target gene yopE . Subsequently , plasmid pJNS13 was integrated into the yopE locus of YPIII via conjugation as described [83] . β-galactosidase activity was determined of three independent cultures of bacteria harboring the cnfY-lacZ fusion as described [83] . The activities were calculated as follows: β-galactosidase activity OD420 • 6 . 75 OD600−1 • Δt ( min ) −1 • vol ( ml ) −1 . For overexpression of CNFY E . coli strain BL21λDE3 was transformed with the cnfY expression plasmid pJNS01 and grown at 37°C in LB medium to an OD600 of 0 , 6 . Subsequently , Plac-driven expression was induced upon addition of 250 µM IPTG and grown at 17°C overnight . CNFY-His6 production was tested by westernblot analysis using an antibody directed against the His-tag ( Qiagen ) . For purification of CNFY , cells were harvested , resuspended in 50 mM NaH2PO4 , pH 8 . 0 , 300 mM NaCl , 10 mM imidazole and lysed with a French press ( 120 . 000 psi ) . The soluble CNFY-His6 extract was separated from insoluble cell material by centrifugation at 25 . 000 g . The CNFY-His6 protein was then purified by affinity chromatography on Ni-NTA agarose ( Qiagen ) . The column was washed with three column volumes of 50 mM NaH2PO4 , pH 8 . 0 , 300 mM NaCl , 20 mM imidazole and eluted with 50 mM NaH2PO4 , pH 8 . 0 , 300 mM NaCl containing 250 mM imidazole . In order to study the influence of the recombinant CNFY protein on actin cytoskeleton rearrangements , 1×105 J774A . 1 cells were incubated with purified CNFY toxin ( 10 nM ) or PBS for 24 h . Subsequently , cells were fixed with 4% paraformaldehyde ( in PBS ) for 10 min at room temperature , washed with PBS and permealized with 0 , 1% Triton X-100 in PBS for 5 min . The actin cytoskeleton was stained with Phalloidin-FITC ( 0 , 5 µg/ml PBS; Invitrogen ) for 15 min at room temperature . Cells were washed in PBS , and the nuclei were stained with DAPI ( 1 µg/ml in TBST ) for 5 min at room temperature . Cells were visualized using a fluorescence microscope ( Axiovert II with Axiocam HR , Zeiss , Germany ) and the AxioVision program ( Zeiss , Germany ) . Activation of RhoA was tested using the Rho activation assay kit 17-294 ( Millipore , Billerica , MA , USA ) and activation of Rac1 and Cdc42 was determined with the Rho/Rac/Cdc42 Activation Assay Combo Kit ( Cell Biolabs , San Diego , CA , USA ) . Approximately 1×106 cells of the macrophage cell line J774A . 1 were starved for at least 20 h in RPMI 1640 without FCS , and incubated with PBS or 25 nM ( 3 µg/ml ) recombinant CNFY for 2–3 h . To test the influence of YopE on Rac1 and RhoA activation , macrophages were subsequently infected for 20 min with wild-type strain YPIII or the isogenic yopE mutant YP275 with an MOI of 100 . Cells were lysed and activation of small Rho GTPases was tested and visualized by western blotting according to the manufacturer's protocol . The Yop delivery assay was performed as described previously [85] . 5×104 ( for fluorescence microscopy ) or 1×106 ( for flow cytometry ) HEp-2 or J774A . 1 cells were incubated with recombinant CNFY ( 25 nM/3 µg/ml ) , exoenzyme C3 transferase from C . botulinum ( CT04 , Cytoskeleton . Inc ) ( 0 . 5 µg/ml , 1 µg/ml ) , C . difficile toxin TcdBF ( 85 ng/ml , 250 ng/ml ) [60] , [61] , or the same amount of PBS before the cells were infected with bacteria with a MOI of 10 . After 1 h cells were washed and dyed with CCF4-AM according to the manufacturer's protocol using the LiveBLAzer-FRET B/G Loading Kit from Life Technologies . Yop translocation was visualized by a fluorescence microscope ( Axiovert II with Axiocam HR , Zeiss , Germany ) using the AxioVision program ( Zeiss , Germany ) or detected with an LSR Fortessa cell analyzer ( BD Bioscience ) . Acquired data of flow cytometry were then analyzed with FlowJo software ( Treestar ) . To compare Yop translocation of YPIII and YP275 ( ΔyopE ) , both strains were pregrown at 37°C and added with an MOI of 100 to approximately 1×106 cells of murine macrophages incubated with PBS or 25 nM recombinant CNFY for 3 h . One hour post infection , cells were washed with PBS , resuspended in SDS sample buffer and separated on 12% SDS polyacrylamide gels . Proteins were blotted onto a membrane and intracellular Yops were visualized with an antiserum directed against all secreted Yops ( α-Yop ) . For the in vitro analysis of Yop delivery into primary cells , mesenteric lymph nodes from uninfected 6- to 8-week-old BALB/c mice were removed . To generate single-cell suspensions the cells were pressed through a 70 µm cell strainer . Harvested cells were counted using an Accuri C6 flow cytometer ( BD Bioscience ) . Bacteria were cultured overnight at 25°C in LB medium , inoculated 1∶20 in fresh LB and grown for 3 h at 37°C prior to infection . Cells were infected for 1 h at 37°C with the wild-type ( YP173 ) , the cnfY ( YP217 ) or the yscS ( YP174 ) mutant encoding the YopE-β-lactamase fusion – ETEM with an MOI of 10 . Infected cells were washed twice with RPMI 1640 medium supplemented with 20 mM HEPES ( pH 7 . 0 ) , 0 . 4% BSA and 50 µg/ml gentamicin to kill bacteria . Subsequently , 2×106 cells were labeled with LiveBLAzer-FRET B/G Loading Kit from Life Technologies . After staining for 1 h at room temperature , cells were prepared for flow cytometry and analyzed as described above . BALB/c female mice aged between 6- and 8-week-old were purchased from Janvier ( Saint Berthevin Cedex , France ) and housed under specific pathogen-free conditions according to FELASA recommendations in the animal facility of the Helmholtz Centre for Infection Research , Braunschweig . For the survival assays , mice were infected orally with approximately 2×109 bacteria of each strain . The infected mice were monitored for 14 days every day to determine the survival rate , the body weight and health status . Bacteria used for organ burden experiments were grown over night in LB medium at 25°C , washed and resuspended in PBS . Groups of 7–10 animals were orally infected with approximately 2×108 bacteria of Y . pseudotuberculosis strains YPIII and YP147 ( ΔcnfY ) using a gavage needle . At specific time points after infection , mice were euthanized by CO2 asphyxiation . PPs , small intestine , caecum , MLNs , liver and spleen were isolated . The ileum was rinsed with sterile PBS and incubated with 100 µg/ml gentamicin in order to kill bacteria on the luminal surface . After 30 min , gentamicin was removed by washing with PBS . Subsequently , all organs were weighted and homogenized in PBS at 30 . 000 rpm for 30 sec using a Polytron PT 2100 homogenizer ( Kinematica , Switzerland ) . To determine the bacterial load of the organs serial dilutions of the homogenates were plated on LB plates with and without antibiotics . The colony forming units ( cfu ) were counted and are given as cfu per g organ/tissue . BALB/c mice were infected intragastrically with 2×109 bacteria of strain YPIII-ETEM ( YP173 ) and the isogenic cnfY mutant YP147-ETEM ( YP217 ) , wild-type YPIII and YP101-ETEM ( YP174 ) . Infection was allowed to proceed for three days . Subsequently , the infected lymphatic tissues ( PPs , MLNs , and spleen ) were isolated and single cell suspensions were generated in PBS by pressing the cells through a cell strainer ( 70 µm , Falcon ) . To eliminate erythrocytes , spleen cells were incubated for 3 min in lysis buffer ( 7 . 8 mM NH4Cl , 10 mM KHCO3 , 100 µM EDTA ) . All cells were resuspended in PBS containing 0 . 2% BSA and total cell number was determined using an Accuri C6 flow cytometer ( BD Bioscience ) . For flow cytometry analysis 1×106 cells were transferred per tube and FcγR was blocked using CD16/CD32 ( BioXCell; anti-mouse CD16/CD32 ) antibody for 15 min at 4°C . Immune cells were first stained for 15 min at 4°C using a biotin-conjugated antibody against CD19 . Subsequently , other cellular surface marker for innate immune cells or T cell panel were stained for 20 min at 4°C in FACS buffer ( PBS+0 . 2% BSA ) using the following antibodies: SA-PerCP-Cy5 . 5 , CD11c-APC , CD11c-PE-Cy7 , Gr1-A750 , CD3-PE , CD4-APC-Cy7 , CD3-PE-Cy7 , NKp46-PE and CD25-APC . Samples were washed twice in FACS buffer and labeled with 1 µg/ml CCF4-AM using the LiveBLAzer-FRET B/G Loading Kit ( Life Technologies ) for 1 hour at 20°C in the presence of 1 . 5 mM probenecid ( Sigma ) and 50 µg/ml gentamicin . Cell subsets were defined as following: B cells ( CD19+ CD3− ) , T cells ( CD19− CD3+ ) , NK cells ( CD19− CD3− NKp46+ ) , neutrophils ( CD19− CD3− CD49b− Ly6G+ CD11b+ ) , macrophages/monocytes ( CD19− CD3− CD49b− Ly6G− CD11b+ ) , and DCs ( CD19− CD3− CD49b− Ly-6G− B220− F4/80− CD11c+ ) . Cells were analyzed in a LSR Fortessa cell analyzer ( BD Bioscience ) . Acquired data were analyzed with FlowJo software ( Treestar ) . Cells from tissues that were not treated with CFF4-AM and/or antibodies were used as negative controls . To characterize the host immune response induced upon infection with the wild-type strain YPIII or the isogenic cnfY mutant strain YP147 , mice were orally infected with approximately 2×108 bacteria of Y . pseudotuberculosis strains YPIII or YP147 ( Δ cnfY ) . Three and six days after infection , PPs , MLNs and spleen were isolated . Single cell suspensions were obtained by mechanical disruption of the organs through a cell strainer . To eliminate erythrocytes , spleen cells were also incubated for 3 min in erythrolysis buffer ( 7 . 8 mM NH4Cl , 10 mM KHCO3 , 100 µM EDTA ) . All cells were resuspended in FACS buffer and total cell number was determined using an Accuri C6 flow cytometer ( BD Bioscience ) . Amounts of 1–2×106 cells were transferred per tube . To exclude dead cells from the analysis live/dead staining ( Invitrogen; Live/dead fixable blue dead cell stain kit , UV excitation ) was performed for 30 min . FcγR and IgG were blocked by 15 min incubation with CD16/CD32 ( BioXCell; anti-mouse CD16/CD32 ) and ratIgG ( Jackson ImmunoResearch; ChromPure Rat IgG , whole molecule ) antibodies . Cellular surface markers for either lymphoid or myeloid panel were stained for 15 min at 4°C in PBS-BSA ( 0 . 2% ) using the following antibodies: CD3-APC , CD4-PerCP-Cy5 . 5 , CD8-eFluor450 , CD335-PerCP-Cy5 . 5 , CD11b-PacificBlue , CD19-Biotin , CD45R-PerCP-Cy5 . 5 , F4/80-PE and CD11c-APCeFluor780 from BD Bioscience , and CD19-FITC , CD49b-Biotin , and Ly-6C-APC from BioLegend . All antibodies were titrated for optimal staining conditions . Biotin-conjugated antibodies were incubated with streptavidin for 15 min at 4°C . After staining cells were fixed with the Foxp3 staining buffer set from eBioscience . Cells were then washed twice and resuspended in 200 µl FACS buffer . Samples were loaded into an LSR Fortessa cell analyzer ( BD Bioscience ) . Acquired data were analyzed with FlowJo software ( Treestar ) . YPIII harboring a PcnfY::gfpmut3 . 1 fusion ( pJNS03 ) and a PgapA::dsRed2 expression construct ( pFU228 ) were grown in LB medium at 25°C overnight . Mice were infected orally with 2×108 bacteria . After five days mice were sacrificed by CO2 asphyxiation . For cryosections , the small intestine , colon , caecum , MLNs , spleen and liver were frozen in Tissue-Tek OCT freezing medium ( Sakura Finetek ) on dry ice . Sections of 8–10 mm were prepared using a Zeiss cryostat Hyrax C50 and mounted on SuperFrost Plus slides ( Thermo Scientific ) . Air-dried sections were fixed for 10 min in ice-cold acetone and washed twice with PBS . For visualization of the nuclei in the fixed tissue , samples were stained with 49 , 6- diamidino-2-phenylindole ( DAPI , Sigma ) for 3–5 min , air-dried and mounted with 80% glycerol in PBS . Localization of yersiniae in the infected tissues and expression of the PcnfY::gfpmut3 . 1 fusion of these bacteria were visualized by a fluorescence microscope ( Axiovert II with Axiocam HR , Zeiss , Germany ) using the AxioVision program ( Zeiss , Germany ) . To detect the cnfY gene expression during the infection process Y . pseudotuberculosis wild-type strain YPIII harboring the PcnfY::luxCDABE fusion vector pJNS02 or the empty vector pFU54 were grown in LB medium at 25°C overnight . About 2×108 luminescent bacteria were used for oral infection . For in vivo imaging mice were anesthesized with isoflurane and the bacterial infection was followed daily using the IVIS Lumina system ( Xenogen ) . To ensure maintenance of the plasmids during the course of infection , the bacteria were isolated from the small intestine , colon , caecum , MLNs , spleen and liver and tested for the presence of the plasmid . MLNs , spleen , liver , small intestine , caecum and colon were analyzed histopathologically of three mice per group . According to standard histology procedures , organs were fixed in 4% neutrally buffered formaldehyde for 24 to 48 h , embedded in paraffin and 3 µm sections were stained with hematoxylin-eosin ( H & E ) . Sections were evaluated by light-microscopy blinded to the experimental groups .
|
Various toxins and effector proteins of bacterial pathogens have been found to manipulate eukaryotic cell machineries to promote persistence and proliferation within their hosts . Many of these virulence factors target small Rho GTPases , but their role in pathogenesis is often unknown . Here , we addressed the expression and functional consequences of the CNFY toxin found in some isolates of Y . pseudotuberculosis . We found that CNFY besides modulating the cell cytoskeleton by activation of the GTPases RhoA , Rac1 and Cdc42 , contributes to increased inflammation and tissue damage . Moreover , CNFY increases the ability of Yersinia to prevent the attack of the immune system , by enhancing the delivery of antiphagocytic and cytotoxic effectors into professional phagocytes . Our findings provide the first insights into the multi-functional action and severe consequences of the CNFY toxin on the inflammatory response and disease-associated tissue damage during the natural course of the infection .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Material",
"and",
"Methods"
] |
[] |
2013
|
The Cytotoxic Necrotizing Factor of Yersinia pseudotuberculosis (CNFY) Enhances Inflammation and Yop Delivery during Infection by Activation of Rho GTPases
|
Pathogenic treponemes related to Treponema pallidum are both human ( causing syphilis , yaws , bejel ) and animal pathogens ( infections of primates , venereal spirochetosis in rabbits ) . A set of 11 treponemal genome sequences including those of five Treponema pallidum ssp . pallidum ( TPA ) strains ( Nichols , DAL-1 , Mexico A , SS14 , Chicago ) , four T . p . ssp . pertenue ( TPE ) strains ( CDC-2 , Gauthier , Samoa D , Fribourg-Blanc ) , one T . p . ssp . endemicum ( TEN ) strain ( Bosnia A ) and one strain ( Cuniculi A ) of Treponema paraluisleporidarum ecovar Cuniculus ( TPeC ) were tested for the presence of positively selected genes . A total of 1068 orthologous genes annotated in all 11 genomes were tested for the presence of positively selected genes using both site and branch-site models with CODEML ( PAML package ) . Subsequent analyses with sequences obtained from 62 treponemal draft genomes were used for the identification of positively selected amino acid positions . Synthetic biotinylated peptides were designed to cover positively selected protein regions and these peptides were tested for reactivity with the patient's syphilis sera . Altogether , 22 positively selected genes were identified in the TP genomes and TPA sets of positively selected genes differed from TPE genes . While genetic variability among TPA strains was predominantly present in a number of genetic loci , genetic variability within TPE and TEN strains was distributed more equally along the chromosome . Several syphilitic sera were shown to react with some peptides derived from the protein sequences evolving under positive selection . The syphilis- , yaws- , and bejel-causing strains differed relative to sets of positively selected genes . Most of the positively selected chromosomal loci were identified among the TPA treponemes . The local accumulation of genetic variability suggests that the diversification of TPA strains took place predominantly in a limited number of genomic regions compared to the more dispersed genetic diversity differentiating TPE and TEN strains . The identification of positively selected sites in tpr genes and genes encoding outer membrane proteins suggests their role during infection of human and animal hosts . The driving force for adaptive evolution at these loci thus appears to be the host immune response as supported by observed reactivity of syphilitic sera with some peptides derived from protein sequences showing adaptive evolution .
Adaptive evolution including positive selection plays crucial roles in the evolution of bacterial human pathogens and both have been well documented on a genome-wide scale in a number of bacterial genera including Escherichia , Helicobacter , Neisseria , Listeria , Salmonella , Streptococcus , Campylobacter , and Actinobacillus [1–8] . Pathogenic treponemes are both human and animal pathogens . Human pathogens include Treponema pallidum ssp . pallidum ( TPA ) , the causative agent of syphilis , T . p . ssp . pertenue ( TPE , the causative agent of yaws ) , and T . p . ssp . endemicum ( TEN , the causative agent of bejel ) while animal pathogens include TPE causing non-human primate infections [9–12] , and T . paraluisleporidarum ecovar Cuniculus ( TPeC; formerly denoted as Treponema paraluiscuniculi ) , the causative agent of venereal spirochetosis in rabbits , and T . paraluisleporidarum ecovar Lepus , which infects hares [13–15] . Although the recent work of Edmondson et al . [16] reported successful long-term cultivation of T . pallidum in a tissue culture system , most of the data on treponemal genetics comes from the whole genome sequencing studies [17 , 18] . The above-listed pathogens are monomorphic , i . e . , highly similar at the genetic level and all the genomes of these pathogenic treponemes characterized to date share a genetic identity of 98% or higher [17 , 18] . The group of human pathogens ( TPA , TPE , TEN ) are even more related sharing a genetic identity over 99 . 7% [17 , 19 , 20] . Currently , 14 complete treponemal genomes have been published including six TPA genomes , six TPE genomes , one TEN genome , and one TPeC genome [12 , 19–29] . In addition , 23 , 25 , 8 , and 6 draft whole genome sequences of treponemal strains or isolates were published recently by Arora et al . [30] , Pinto et al . [31] , Sun et al . [32] , and Marks et al . [33] , respectively . The genomes of pathogenic treponemes related to T . pallidum contain no prophages or insertion sequence-elements [21 , 34] , or plasmids [35] . Therefore , recombination is expected to be quite infrequent among these treponemes due to a lack of mobile genetic elements [21] . However , traces of both intragenomic DNA recombinations via gene conversion [36–38] and intergenomic homologous recombination after DNA horizontal gene transfer have been described [20 , 25] . In addition , traces of positive selection have been detected in previously published papers including TPA and TPE comparisons [19] , comparisons within TPA strains [39] , and detailed intrastrain analysis [40] . However , no comprehensive analysis of positively selected loci in the genomes of pathogenic treponemes has been performed to date . Despite the fact that treponemes related to Treponema pallidum are monomorphic bacteria with extremely low level of genetic diversity [18] , it has been shown that human immunity does not protect against different subspecies and not even against different syphilis strains [41] . Therefore , divergent genes encoding differences in proteomes of individual treponemal strains and subspecies are likely of importance for development of syphilis vaccine . In this communication , the whole genome sequences of 11 treponemal strains were systematically analyzed for the presence of positive selection . The identified genes were further reanalyzed relative to all sequences available in 62 draft genomes published to date . The causative agents of syphilis , yaws , and bejel differed in sets of positively selected genes . Moreover , several synthetic peptides covering positively selected protein regions were found to interact with syphilitic sera .
A set of 11 treponemal genomic sequences was examined in this study and included genomes of five TPA strains ( Nichols , DAL-1 , Mexico A , SS14 , Chicago B ) , four TPE strains ( CDC-2 , Gauthier , Samoa D , Fribourg-Blanc ) , one TEN strain ( Bosnia A ) , and one strain of TPeC ( Cuniculi A ) . An overview of the complete genome sequences used is shown in Table 1 . Subsequent analysis of selected genes was performed on additional 62 draft genomes ( Arora et al . [30]; Pinto et al . [31]; Sun et al . [32]; GenBank genome TPA sequences [UW074B , UW189B , UW228B , UW254B , UW391B] and TEN Iraq B [CP032303 . 1] ) . The overall algorithm is shown in Fig 1 . The TPE Samoa D was used as a reference genome , and all 1068 orthologous genes were extracted from the other 10 complete genomes using given annotation coordinates . The orthologous sequences from the complete genomes were aligned at the codon level using Matlab R2013a software and the Bioinformatics Toolbox . Only genes where at least three nonsynonymous mutations at different sites occurred were further analyzed with respect to the presence of positive selection . A BLAST search was used to determine orthologous sequences in the draft genomes . These sequences were aligned in Matlab at the nucleotide level because large number of ambiguous sites precluded proper automatic ORF localization . Then the sequences from the draft genomes were scanned for nucleotide differences . After filtering sites with unknown nucleotides , insertions , and deletions , only orthologs with more than five nucleotide differences at different sites were analyzed further . These orthologs were aligned with the corresponding genes from the complete genomes to determine the ORF and identify the number of nonsynonymous mutations . At the same time , the TPeC orthologs were excluded in this step due to frequent sequential diversity and due to the lack of pathogenicity of TPeC to humans . The removal of the TPeC orthologs did not change the number of detected positively selected genes . Compared to an analysis of whole genomes , no new locus with more than three nonsynonymous mutations at different sites was identified during the analysis of draft genomes . For each analyzed gene from the complete genomes , a maximum likelihood phylogenetic tree with 50 bootstrap replicates was constructed using MEGA 6 [48] . Different evolution models ( Kimura 2-parameter [49] , Tamura 3-parameter [50] , and Tamura-Nei [51] ) were applied to each gene . The trees of each gene were compared by calculating Robinson-Foulds distances [52] using R software ( packages phytools and phangorn ) [53] . The comparison showed that the choice of the evolution model did not significantly change the topology of the tree; the Tamura-Nei evolution model was chosen for the construction of all phylogenetic trees . A calculation of mutational rate ratio ω between two gene sequences was the basis for the positive selection analysis . The ω was calculated as a ratio of nonsynonymous to synonymous mutational rates . The ratio indicates negative purifying selection ( 0 < ω < 1 ) , neutral evolution ( ω = 1 ) , and positive selection ( ω > 1 ) [54] . A set of selected genes from complete genomes was tested relative to positive selection using the maximum likelihood method using the CODEML of the PAML software package [55] . PAML version 4 [56] and its user interface PAMLX [57] were used in our study . For each analyzed gene , its maximum likelihood phylogenetic tree was used as an input tree . The CODEML offers several different codon evolutionary models , and the statistical likelihood ratio test ( LRT ) was used to compare the codon evolutionary model to the null model . The Bayes empirical Bayes method ( BEB ) [58] was then used to evaluate the posterior probability of sites considered to have been positively selected . The CODEML models could produce different results ( i . e . , a list of sites under positive selection ) since they calculate different parameter estimates . Site models allow ω to vary in each site ( codon ) within the gene . Statistical testing was required for sites with ω > 1 . Two pairs of models were predominantly used since their LRTs have low false-positive rates . M1a ( nearly neutral evolution ) was compared to M2a ( positive selection ) [58 , 59] and M7 ( beta ) was compared to M8 ( beta & ω ) [60] . Our preliminary testing found that the two model pairs gave the same or very similar results . Therefore we chose to use the M7-M8 model pair . The M7 model is a null model that allows 10 classes of sites with a ω beta-distribution within the interval 0 ≤ ω ≤ 1 . Sites with ω > 1 are not allowed . The alternative M8 model adds an eleventh class of sites with ω > 1 . Each site was tested to determine the class to which it belongs . The LRT compares twice the log-likelihood difference 2Δl = 2 ( l1-l0 ) between the M7 model ( log likelihood value l0 ) and the M8 model ( log likelihood value l1 ) to the χ2 distribution [61] . If the twice log-likelihood difference is above a critical χ2 value , then the null model is rejected , and the positive selection is statistically significant . A considerable disadvantage of the site models is that ω was calculated as an average over all codons of the site . Therefore , the site models are not suitable for the data where ω also varies between lineages . In contrast , the branch-site models search for positive selection in sites and pre-specifies lineages where different rates of ω may occur [62] . Sequences of lineages are a priori divided into a group of foreground lineages where positive selection may occur and group of background lineages where only purifying selection or neutral evolution occurs . We used branch-site model A , which allows four classes of sites and different setups of foreground lineages to be tested depending on the gene phylogeny . In branch-site model A , all lineages under purifying selection with a low value of ω0 belong to site class 0 . Weak purifying selection and neutral evolution with ω1 near to value 1 are allowed in site class 1 . In site class 2a , a proportion of class 0 sites in foreground lineages is under positive selection with ω2 > 1 . Similarly , site class 2b is a proportion of class 1 sites under positive selection with ω2 > 1 . The null model for LRT has ω2 = 1 . Critical values of LRT ( 2Δl ) are 2 . 71 at 5% and 5 . 41 at 1% [63] . The posterior probabilities of suggested sites under positive selection were calculated using the BEB method . The average pairwise p-distances ( APD ) and average number of mutations ( transitions and transversions ) , calculated using MEGA-X [64] , were used to evaluate genetic diversity . A pairwise deletion of sites with gaps/missing data was used . The Fisher exact statistical test was used to assess the significance of the changes between average numbers of mutations . Synthetic biotinylated peptides , covering protein regions containing positively selected residues , were designed . Peptide synthesis was performed by JPT Peptide Technologies ( Berlin , Germany ) on a 50–200 nmol scale . The lyophilized peptides were resuspended in TBS buffer ( 25 mM Tris , 150 mM NaCl , pH = 7 . 2 ) at 1 mM concentrations and were stored at −20°C . Prior to further use , synthetic peptides were diluted 1000x in TBS buffer . Streptavidin-coated 96 well plates ( Pierce Streptavidin Coated High Binding Capacity White 96-Well Plates; Thermo Scientific , Rockford , USA ) were washed three times with 200 μl of washing buffer ( TBS buffer containing Tween 20 ( 0 . 05% ) and Bovine Serum Albumin , BSA ( 0 . 1% ) ( Sigma-Aldrich , Prague , Czech Republic ) ; then 100 μl of diluted peptide was added to each well and incubated for 30 min at room temperature as recommended by the manufacturer ( Thermo Scientific ) with mild shaking . Subsequently , each well was washed three times with washing buffer ( 200 μl in each step ) ; then 100 μl of blocking buffer SuperBlock Blocking Buffer in TBS ( Thermo Scientific ) were added and incubated for 30 min at room temperature with mild shaking . Each well was washed three times with washing buffer ( 200 μl in each step ) and 100 μl of diluted sera ( 1:500 in washing buffer ) were added and incubated for 30 min at room temperature with mild shaking . Each well was washed three times with washing buffer ( 200 μl in each step ) and 100 μl of diluted secondary antibody conjugated with horseradish peroxidase were subsequently added ( 1:2000 in washing buffer; Goat Anti-Human IgG/IgA/IgM Horseradish Peroxidase Conjugate; Life Technologies , Carlsbad , USA ) and incubated for 30 min at room temperature with mild shaking . Each well was then washed three times with washing buffer ( 200 μl in each step ) ; then 100 μl of chemiluminescent detection solution ( Super Signal ELISA Pico Chemiluminescent Substrate; Thermo Scientific ) was added . Luminescence was measured on a TriStar2 LB 942 luminometer with a Modular Multimode Microplate Reader ( Berthold Technologies , Bad Wildbad , Germany ) . Each experiment was performed at least three times . A signal was considered positive when it was higher than the average of the three lowest values for each serum plus five standard deviations of the average value . The human sera were collected from adult patients diagnosed with syphilis at the Department of Dermatology , 1st Faculty of Medicine , Charles University , Prague , Czech Republic . Sera from child patients diagnosed with Lyme disease were obtained from the Department of Children's Infectious Diseases , Faculty of Medicine and University Hospital , Masaryk University , Brno , Czech Republic . All clinical samples were obtained after patients or parents of involved children signed an informed consent . The design of the study was approved by the ethics committee of the Faculty of Medicine , Masaryk University . All human sera were collected under established guidelines .
A comparison of 47 orthologous gene sequences from the complete genomes where at least three nonsynonymous mutations at different sites occurred was used to identify positively selected genes using the site and branch-site models of the CODEML in PAML package [55] . The completely sequenced genomes are listed in Table 1 and include 11 genomes . In addition , 25 draft TPA genomes [31] , 23 draft TPA and TPE genomes [30] , 8 TPA genomes [32] , 5 TPA genomes from GenBank ( UW074B , UW189B , UW228B , UW254B , UW391B ) , and one TEN ( Iraq B ) were also analyzed . The overall algorithm is shown in Fig 1 . In all cases of complete treponemal genomes , the genome structure was identical or very similar allowing straightforward identification of gene orthologs . However , in many cases , draft genome sequences were either incomplete or contained many ambiguous bases precluding their use in analyses . This resulted in a variable number of sequences used for identification of positively selected sites within individual loci . Altogether , 22 positively selected genes were identified in the TP genomes using site model analysis in PAML ( Table 2 ) . The number of positively selected amino acid sites varied from 1 to 65 , with a median value of 8 . 5 . A list of positively selected protein sites identified using PAML software ( site and branch-site models ) as well as PAML-identified positively selected protein sites within treponemal subspecies are shown in S1 Table . Average pairwise p-distances ( APD ) were calculated for each of the 22 genes from 10 complete genomes ( Cuniculi A genome was removed from analysis ) and from the draft genomes . The APD value for each gene was compared with APDw10 = 0 . 000525 of the 10 complete genomes without 54 variable loci ( listed in S2 Table ) which can be considered as a background level of polymorphism . All 22 genes evolving under adaptive evolution had elevated nucleotide substitution density . Out of these 22 genes , 14 genes were previously reported as recombinant ( Table 2 ) . These genes are listed in Table 3 . Functionally , these genes comprised the tpr genes ( tprC , D , G , I , J , L ) , outer membrane proteins ( TP0133 , TP0136 , TP0548 , TP0856 , TP0858 , TP0865 ) , and genes encoding the outer membrane biogenesis protein ( TP0326 [BamA] ) and methyl-accepting chemotaxis protein ( TP0488 [Mcp-2] ) . Putative recombination loci were most frequently identified in TEN strain ( n = 9 ) while in other treponemes there were fewer predicted recombinant loci ( TPA , n = 7; TPeC , n = 1; TPE , n = 1 ) . When recombinant sequences were removed from analyses , positive selection among 14 these loci ( Table 3 ) was found mostly within TPA strains or isolates ( n = 9 ) , within TPE ( n = 4 ) , between TEN and TPA/TPE sequences ( n = 2 ) , between TPA and TPE strains ( n = 1 ) and between TPA and TPE/TEN sequences ( n = 1 ) . A list of the 8 positively selected genes is shown in Table 4 . These genes include tprF gene , genes encoding outer membrane proteins ( TP0515 , TP0733 , TP0859 ) , subtilisin-like proteins ( TP0314 , TP0462 ) , enzymes ( TP0619 ) , and hypothetical protein ( TP0126b ) . Evidence of positive selection from analyses performed within the strains belonging to different treponemal subspecies and analyses from the PAML branch-site model revealed that positive selection was found mostly between TPA and TPE strains ( n = 6 ) , and within TPA strains or isolates ( n = 2 ) . In one case ( TP0859 ) , positive selection was found both between TPA and TPE strains ( n = 1 ) , and between TEN and TPA/TPE ( n = 1 ) . Identified positively selected genes in various treponemal species and subspecies are shown in Table 5 and Fig 2 . While Table 5 lists all identified recombinant and positively selected genes in TPA , TPE , and TEN groups of strains or isolates , Fig 2 shows only recombinant or positively selected genes that were identified within a particular subspecies . To test whether genetic diversity in treponemal genomes was distributed evenly along the chromosome , an average pairwise p-distance ( APD ) and average number of mutations ( ANM , transitions and transversions ) were calculated ( Table 6 ) for concatenation of 54 genes ( S2 Table ) representing a total length of 86 . 8 kbp ( 7 . 6% of the genome length ) . These 54 genes included the tpr genes , genes with traces of possible recombination events ( according to Gray et al . [38] , and Čejková et al . [19] ) , genes showing inter-strain variability between TPE and TPA strains and their paralogs ( according to Čejková et al . [19] ) , genes showing intra-strain variability [40] , and previously identified positively selected genes ( according to Čejková et al . [19] ) ( S2 Table ) . The TPA strains contained greater genetic diversity within these genes compared to TPE strains ( APD ( TPA ) = 420 . 0*10−5 versus APD ( TPE ) = 384 . 2*10−5 ) . Note that the average pairwise p-distances within TPA strains ( compared to TPE/TEN strains ) are lower when complete genomes were analyzed but higher within selected 54 loci ( Table 6 ) . In addition , genetic distance within TPA strains ( compared to TPE/TEN strains ) is markedly lower in complete genomes without selected 54 genes compared to whole genome analyses ( p = 0 . 0008 ) . Synthetic biotinylated peptides were designed to cover protein regions where positively selected amino acid residues were detected . These peptides were tested for reactivity with the patient's syphilis sera ( Table 7 ) . As a control , serum from a patient with Lyme disease was used . A positive result was obtained for one to seven peptides ( out of 8 tested ) depending on the serum used . At the same time , serum obtained from Lyme disease patient failed to recognize peptides derived from treponemal proteins but did react with peptides derived from borrelial protein ErpA .
In this study , 22 genes showing traces of positive selection were identified among the TPA , TPE , and TEN genomes . Within this group of genes , recombination was previously reported in 14 genes . Genes with previously detected recombination events were often found to contain positively selected amino acid residues both among the recombinant and the non-recombinant sequences , which indicates that both recombination and positive selection are different mechanisms of treponemal adaptive molecular evolution . Adaptive evolution is common to many bacterial pathogens and can usually be found in genes important to the interaction between the host and the pathogen , i . e . , where new protein variants are of selective advantage for the survival of the pathogenic strain . The immune pressure from the host favors , in microbial genes encoding proteins exposed on the surface of the pathogen , emerging genetic variants , which due to immune evasion , get positively selected . In Escherichia coli , positive selection is limited to a few dozen genes [3] while in several other genomes , including Streptococcus [1] and Campylobacter [2] , traces of positive selection have been found in more than half of the core genome . The extent to which adaptive evolution of different bacterial pathogens differs depends on several bacterial features including bacterial mutation rate , frequency of genetic recombination and horizontal gene transfer , and genome size . Genetic recombination occurs more frequently in Neisseria [8] and Helicobacter [5] compared to several bacterial pathogens such as Escherichia [3] , Salmonella [4] , and Listeria [6] . Moreover , compared to E . coli , Helicobacter pylori has about a 100-times higher mutation rate due to the lack of a highly efficient DNA repair system [72] . Treponemes related to T . pallidum represent bacterial pathogens with small genomes , with an extreme paucity of outer membrane proteins [73] , high genetic similarity , and a relatively low mutation rate [28] . Moreover , there are no known mechanisms of horizontal gene transfer in syphilis , bejel , and yaws treponemes [18] . These features of pathogenic treponemes are consistent with a relatively small number of positively selected genetic loci , which consists of just 22 genes ( 2 . 1% of all protein-encoding genes ) . Moreover , this situation also partly reflects the fact that the number of determined treponemal genomes is quite low due to the difficulties in long-term cultivation of treponemes [16]; the sequenced genomes currently available do not reveal the entire genetic variability present among human pathogenic treponemes . A recently developed MLST typing of TPA treponemes [74 , 75] revealed a number of genetic variants of the TP0705 gene and the vast majority of these variants resulted in amino acid replacements , which suggests positive evolution at this locus . This locus was not identified in this study because of the limited variability present in the currently available genome sequences . It is therefore expected that the list of positively selected/recombinant treponemal loci will grow larger as the number of additional genomes accumulate . Among the genes identified in this study , a substantial number of genes were shown to have evidence of recombination event . Several recent studies revealed that genetic recombination in pathogenic treponemes is not only limited to intra-genomic homologous recombination and gene conversion events involving rDNA loci , tpr genes and their paralogs , and the TP0856 and TP0858 genes [36 , 38 , 68] . In addition to intra-genome recombinations , two genes , TP0326 and TP0488 , in the TPA Mexico A genome , were found as a result of recombination with exogenous DNA , likely as a result of DNA uptake and chromosome incorporation during coinfection with a different treponemal subspecies [25] . Similar recombinations were detected in a TPA isolate from South Africa [66] . In the work of Grange et al . [76] , TEN strain 11qj within the TP0548 locus and a nucleotide sequence almost identical to TPE strains [67 , 77 , 78] indicating that the TP0548 locus in other TEN strains is also a result of an interstrain recombination event . In the genome of TPA Sea84-1 , the TP0621 locus revealed sequences identical to TPE [23] . Moreover , whole-genome sequencing of TEN Bosnia A revealed several genomic loci similar to TPA strains [20] . All these findings suggest that genetic recombinations of exogenous DNA into the chromosomal loci of treponemes are rare but detectable events . In addition , these findings suggest that the corresponding recombination could be of selective advantage during host infection given the fact that the recombination of foreign DNA without available horizontal gene transfer mechanisms is quite infrequent . In the work of Arora et al . [30] , the authors find the genes with predicted putative recombinant regions ( e . g . , TP0136 , TP0462 , TP0548 , TP0733 , TP0894-898 ) using a phylogenetic incongruence method , ClonalFrameML , and Gubbins , overlapping with genes identified in our study . However , several predicted recombinant genes comprising TP0179 , TP0313 , TP0315 , TP0967 , TP0968 [30] , were not identified in this study reflecting the fact that not all recombination events result in detectable positive selection signal . Out of the 22 genes showing adaptive evolution identified in this study , nine genes ( 40 . 9% ) were identified in the set of 71 genes showing intra-strain heterogeneity ( reviewed in Šmajs et al . [18] ) . Since most of the observed intra-strain heterogeneity resulted in non-synonymous amino acid changes , identified intra-strain heterogeneity should be considered as ongoing adaptive evolution and , as such , it is not surprising that both sets of positively selected/recombinant genes and genes showing intra-strain heterogeneity overlapped considerably . Functionally , the genes showing adaptive evolution identified in this study were the tpr genes ( tprC , D , F , G , I , J , L ) , outer membrane proteins ( TP0133 , TP0136 , TP0515 , TP0548 , TP0733 , TP0859 , TP0865 ) and genes encoding outer membrane biogenesis protein ( BamA ) , lipoproteins ( TP0462 , TP0856 , TP0858 ) , methyl-accepting chemotaxis protein ( TP0488 ) , and three other proteins ( TP0126a , TP0314 , TP0619 ) . The driving force for adaptive evolution at these loci thus appears to be the host immune response . The observed reactivity of syphilitic sera with some peptides derived from protein sequences evolving under positive evolution supports this prediction . This study has several limitations . First , the number of analyzed genomes was quite limited , especially regarding the TEN genomes . In addition , in recently published draft genomes of TPA , the candidate genes are often only partially sequenced . With a much larger set of treponemal sequences , one can expect a higher number of positively selected genes . The second major limitation was the disproportionality in the number of analyzed genes per species/subspecies , which reflects genome availability . Another limitation is related to the fact that the analyzed complete genomes were obtained from treponemal strains propagated in rabbits and could therefore reflect adaptation of treponemes to this host . However , the analysis of draft genome sequences in this study obtained directly from clinical material suggests that the observed traces of positive selections are present also during infection of humans . Moreover , the identified positively selected positions may represent recent mutations that were not yet removed by negative selection . In this study , a detailed analysis of traces of positive selection in 3 T . pallidum subspecies including ssp . pallidum ( TPA ) , ssp . pertenue ( TPE ) , and ssp . endemicum ( TEN ) enabled us to classify most of the identified positively selected genes to a particular subspecies when analyses were performed separately within strains and isolates of the same subspecies or when a PAML branch-site model was used to identify lineages with positively selected loci . The majority of positively selected genes were identified within the TPA and TPE genomes , likely as a result of the highest number of available sequences for these subspecies . However , TPA sets of positively selected genes differed from TPE genes . Among TPA , members of the paralogous tpr family ( tprCGJ ) and the TP0136 paralogous gene family ( TP0136 , TP0462 ) prevailed , while among TPE , a paralogous gene family containing TP0856 , TP0858 , TP0859 , TP0865 , showed adaptive evolution . Interestingly , the genes belonging to the latter family ( TP0548 , TP0856 , TP0858 , TP0859 ) were found to be recombinant among TEN genomes . These findings suggest that genomic loci showing signs of adaptive evolution could differ between TPA and TPE/TEN strains/isolates . This finding supports the observed and consistent genetic differences between treponemal subspecies TPA , TPE , and TEN , and shows that the ways TPA and TPE strains interact with a host during infection is different . Although some authors suggest that the subspecies classification is a case of opportunity and not the consequence of genetic and biological differences [79 , 80] , our findings support the latter explanation .
|
In the genus Treponema there are several human and animal pathogens that include the causative agent of syphilis ( Treponema pallidum ssp . pallidum; TPA ) , the causative agent of yaws ( T . p . ssp . pertenue; TPE ) , and the causative agent of endemic syphilis ( T . p . ssp . endemicum; TEN ) . T . paraluisleporidarum causes venereal spirochetosis in rabbits . We used whole genome sequences of 11 treponemal strains together with additional 62 draft genomic data to identify genes evolving under positive selection . The identified genes evolving under positive selection partly overlapped with the genes previously reported as recombinant and were found to be different in treponemal subspecies . Since both genetic recombination and positive selection could allow a survival of pathogenic bacteria despite the human immune response , identification of such genes could predict the major antigens recognized by the human immune system and also identify the most suitable components for development of an anti-treponemal vaccine .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"urology",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"tropical",
"diseases",
"microbiology",
"membrane",
"proteins",
"treponematoses",
"bacterial",
"diseases",
"outer",
"membrane",
"proteins",
"sexually",
"transmitted",
"diseases",
"genome",
"analysis",
"neglected",
"tropical",
"diseases",
"evolutionary",
"adaptation",
"dna",
"dna",
"recombination",
"cellular",
"structures",
"and",
"organelles",
"bacterial",
"pathogens",
"infectious",
"diseases",
"genomics",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"recombinant",
"proteins",
"genetic",
"loci",
"cell",
"membranes",
"biochemistry",
"cell",
"biology",
"nucleic",
"acids",
"genitourinary",
"infections",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"evolutionary",
"biology",
"evolutionary",
"processes",
"syphilis"
] |
2019
|
Identification of positively selected genes in human pathogenic treponemes: Syphilis-, yaws-, and bejel-causing strains differ in sets of genes showing adaptive evolution
|
Analysis of polymorphism and divergence in the non-coding portion of the human genome yields crucial information about factors driving the evolution of gene regulation . Candidate cis-regulatory regions spanning more than 15 , 000 genes in 15 African Americans and 20 European Americans were re-sequenced and aligned to the chimpanzee genome in order to identify potentially functional polymorphism and to characterize and quantify departures from neutral evolution . Distortions of the site frequency spectra suggest a general pattern of selective constraint on conserved non-coding sites in the flanking regions of genes ( CNCs ) . Moreover , there is an excess of fixed differences that cannot be explained by a Gamma model of deleterious fitness effects , suggesting the presence of positive selection on CNCs . Extensions of the McDonald-Kreitman test identified candidate cis-regulatory regions with high probabilities of positive and negative selection near many known human genes , the biological characteristics of which exhibit genome-wide trends that differ from patterns observed in protein-coding regions . Notably , there is a higher probability of positive selection in candidate cis-regulatory regions near genes expressed in the fetal brain , suggesting that a larger portion of adaptive regulatory changes has occurred in genes expressed during brain development . Overall we find that natural selection has played an important role in the evolution of candidate cis-regulatory regions throughout hominid evolution .
Over 30 years ago it was suggested that evolutionary changes at the level of gene regulation might have had a greater influence on human and chimpanzee phenotypic divergence than changes in proteins themselves [1] . This is exemplified by humans and chimpanzees being highly similar at the protein level yet manifesting considerable phenotypic differences . Until recently , the focus of evolutionary studies has been on changes occurring within the protein-coding regions of genes , while evidence from Drosophila [2] , [3] , rodents [4] , [5] , and primates [6]–[19] have all suggested that Darwinian selection can be an important driving force of evolutionary change within non-coding DNA . Microarray studies have identified numerous transcriptional differences between human and chimpanzee [20]–[25] , and polymorphisms in cis-regulatory regions have been directly associated with differences in gene expression levels in humans [26] , [27] . Therefore , single nucleotide polymorphisms ( SNPs ) in non-coding DNA can be associated with phenotypic differences , making them potential targets for natural selection . A simple model of gene regulation involves the binding of transcription factors to several short non-coding sequences , which are generally found upstream of the transcribed regions of genes ( forming a cis-regulatory region ) . Once a specific combination of transcription factors has bound to the DNA , the recruitment and assembly of the general transcriptional machinery will initiate gene transcription . However , gene regulation can also occur via additional processes , such as nucleosome positioning , distal enhancer elements , DNA methylation , and microRNA regulation . While a variety of computational methods have been developed to identify specific regulatory elements at the genomic scale , these methods can be prone to false positives due to the degeneracy and short size of DNA binding sites [28] . An alternative predictor of regulatory function at the genomic scale may be the evolutionary conservation between species [29] , [30] , although conservation scores can be prone to false negatives as not all functional sites are expected to be conserved [14] , [31] . However , the study of conserved non-coding sites within regions potentially enriched for cis-acting elements does not a priori specify any particular mode of regulation , and may provide a broader glimpse of the evolutionary processes acting on candidate cis-regulatory regions . We examined patterns of polymorphism and divergence in conserved non-coding sites ( CNCs ) in the flanking regions of 15 , 061 human genes , and the coding regions of 13 , 009 genes that were sequenced by Celera Genomics in 15 African Americans and 20 European Americans . We find evidence for selective constraint and adaptive evolution within candidate cis-regulatory regions , and find non-random patterns with respect to functional and transcriptional profiles of genes with higher probabilities of selection . Moreover , we find that patterns observed in candidate cis-regulatory regions are often distinct from those observed in protein-coding regions .
We defined a candidate cis-regulatory region as all CNCs within the introns , UTRs , or within 5 kb up- or downstream of the transcription start or stop site of a gene . CNCs were pooled in this way to increase the power to detect selection , and to capture signals of selection without regards to a specific mode of cis-regulation . However , we note there is a significant correlation between the probability of selection estimated from the 5′ upstream regions and the combined set of CNCs ( Figure S2 ) . To identify genes whose candidate cis-regulatory regions may be under selection , contingency tables were constructed containing the counts of polymorphic sites and fixed differences to the chimpanzee in the pooled flanking CNCs of a gene . The total number of synonymous polymorphisms and fixed differences in coding regions ( without respect to human and mouse sequence conservation ) were pooled to use as a neutral standard as in [35]; the use of pooled vs . local synonymous sites has little effect on our estimates of selection ( Figure S3 ) . In order to identify loci showing signatures of natural selection , we implemented the program mkprf [36] by conservatively setting no fixed variance on the prior distribution of the population scaled selection coefficient ( γ = 2Nes ) , which shrinks the estimates of γ ( Figure S4 ) . For each gene we quantified the probability that the estimate of γ falls within five bins: γ<−1 ( strong negative selection ) , −1<γ<−0 . 5 ( weak negative selection ) , −0 . 5<γ<0 . 5 ( nearly neutral ) , 0 . 5<γ<1 ( weak positive selection ) , and γ>1 ( strong positive selection ) ( Figure 3 ) . In order to compare different modes of selection , the probability of negative selection was defined as Pr ( γ<−0 . 5 ) , and the probability of positive selection was defined as Pr ( γ>0 . 5 ) . In order to control for population size changes that may affect our estimates of positive and negative selection , we incorporated demographic parameters when calculating the likelihood of our observed data in mkprf , the effect of which can be seen in Figure S5 . A population expansion is expected to increase levels of polymorphism at neutral sites to varying extents due to differences in local mutation rates . In a set of neutral loci simulated under a model of population expansion , failing to correct for demography results in a neutral locus showing strong signatures of negative selection with credibility intervals on the mean estimate of γ below 0 ( Table 4 ) , and a slight inflation in the probability that γ<0 at higher values ( Figure S6 ) . Demographic models including a population expansion in African Americans and a single population bottleneck in European Americans were fitted to the complete set of autosomal synonymous SNPs using the program prfreq [32] and incorporated into mkprf . However , a single bottleneck model was found to be a poor fit to the European American sample , and even a complex , multi-bottleneck model could not account for the excess of high frequency derived alleles at synonymous sites [32] . Therefore , we focus the majority of our conclusions on the African American sample . A complete list of genes including McDonald-Kreitman tables and mkprf results are available in Table S1 , Table S2 , Table S3 , Table S4 , Table S5 , Table S6 , Table S7 , and Table S8 . In order to evaluate the performance of mkprf , we performed Wright-Fisher forward simulations under the inferred demographic model in African Americans using the program SFS_CODE [37] . We find a significant correlation between simulated and estimated mean values of γ in datasets including various combinations of loci simulated under positive and negative selection ( Figure S7 ) . However , in a dataset including an equal number of loci simulated under positive and negative selection , the correlation is stronger for positively selected loci as compared to negatively selected loci , most likely reflecting a limited ability to distinguish between strong vs . weak negative selection . There is also a 10-fold increase in the number of genes returned with credibility intervals ( CIs ) on the mean estimate of γ>0 compared to CIs<0 ( neut+del+pos , Table 4 ) , as genes subject to strong enough negative selection often have reduced levels of polymorphism and divergence . Moreover , of the 11 , 000 simulated loci in the neut+del+pos dataset , 96% of loci in the positively selected class ( γ>0 ) had at least 1 informative site , whereas only 94% of loci in the negatively selected class ( γ<0 ) had at least 1 informative site making an estimate of γ even possible in mkprf . We note that across all datasets , only a limited number of the loci simulated under positive and negative selection have CIs on the mean estimate of γ above or below 0 ( Table 4 ) , suggesting there are likely additional loci subject to selection than identified by CIs alone . An important consideration in gene specific tests for selection is unequal sequence coverage , as genes with a smaller number of resequenced sites tend to have fewer informative sites ( fixed or polymorphic sites , Figure S8 ) . It is unlikely that we have complete coverage of all cis-regulatory regions of a gene due to a focus on the 5′ upstream regions of genes , which is potentially problematic for our comparisons if genes with fewer informative sites were to provide less reliable estimates of γ . As discussed in the previous paragraph this may be particularly relevant for negatively selected loci , as genes subject to strong negative selection are expected to have fewer informative sites . Our simulations confirm that a higher percentage of negatively selected loci have fewer than 4 informative sites ( neut+del+pos , 52% ) as compared to either positively selected ( 16% ) or neutral loci ( 28% ) . However , we find no significant difference in the distribution of γ when we compare loci with fewer or at least 4 informative sites when simulated under neutral , positive , and negative selection ( Figure S9 ) , suggesting that loci with fewer informative sites are unlikely to be problematic in our comparisons . The distributions of the number of resequenced sites at a locus in candidate cis-regulatory and protein-coding regions are shown in Figure S10 . We downloaded the Novartis Gene Expression Atlas 2 data from 72 normal human tissues in order to examine patterns of selection in candidate cis-regulatory regions with respect to gene expression signals [38] . Microarray expression profiles were available for 87% of genes in our tests for natural selection . Conflicting studies have shown that expression patterns in tissue-specific genes evolve more rapidly [39] , or more slowly [40] as compared to broadly expressed genes in comparisons of humans and mice . More recently , experimentally defined cis-regulatory regions were found to exhibit stronger degrees of selective constraint in genes expressed in a smaller number of tissues [18] . However , we find little correlation between the probability of negative selection and the absolute number of tissues in which a gene is expressed ( AA: Kendall's tau = −0 . 011 , p = 0 . 12; EAs: tau = −0 . 0060 , p = 0 . 40 ) , nor with the index of tissue specificity [41] , which includes additional information on the level of expression in each tissue ( AA: Kendall's tau = 0 . 0086 , p = 0 . 21; EAs: tau = 0 . 0024 , p = 0 . 73 ) . We also find no significant correlation between probabilities of selection and the mean and the maximum expression level of a gene across all tissues , suggesting that expression levels may not have a large impact on patterns of recent natural selection within candidate cis-regulatory regions . However , our simulations suggest that we likely have reduced power to identify significant correlations that are driven by negative rather than positive selection . We then assigned genes to each tissue of expression in order to examine differences across tissues in evidence for selection on candidate cis-regulatory regions ( Table S9 , Table S10 ) . Mann-Whitney U tests did not identify any tissues with a significantly higher probability of negative selection in either candidate cis-regulatory regions or nonsynonymous sites , suggesting that weak selective constraint may be a persistent factor affecting most human tissues , but likely reflects the limited power we have to detect negative vs . positive selection . On the other hand , we find that genes expressed in at least 3 tissues have a significantly higher probability of positive selection in candidate cis-regulatory regions ( FDR<10% , Table 5 ) , suggesting they may have an excess of positively selected loci . Notably , genes expressed in the fetal brain have a higher mean probability of positive selection as compared to genes expressed in other tissues , suggesting the importance of adaptive regulatory changes during brain development . Similarly , genes expressed in certain tissues of the adult brain , including the cerebellum peduncles and the medulla oblongata , also have a higher probability of positive selection in candidate cis-regulatory regions . Curiously , the medulla controls a variety of autonomic functions and is considered to be the most plesiomorphic structure of the brain [42] , and might be expected to have a high level of conservation across species . However , the medulla also contains several motor nuclei important for facial expression , mastication , tongue movements , and controlling sound amplitudes that reach the inner ear , which are hypothesized to have played an adaptive role in the evolution of facial expression , feeding , and speech in humans [43] . Microarray studies have found that differences in brain expression patterns are more pronounced between humans and other primates as compared to other tissues [20] , and that the majority of these differences are likely due to upregulation of brain-expressed genes in humans as compared to chimpanzees [21] . However , these findings have been controversial [22] , and a more recent study has found that differences in expression patterns between humans and chimpanzees are less pronounced in the brain as compared to heart , kidney , liver and testis [24] . Regional expression in parts of the brain is generally conserved between human and the more distantly related mouse [44] , raising the possibility that cognitive differences between species may be more likely to result from differential gene expression during development . While comparative microarray studies have generally focused on adult brain expression , our findings suggest that adaptive evolution might have had a larger impact on expression patterns in the fetal brain . We applied the same methodology to examine differences in selection acting on coding regions with regards to gene expression . In contrast to what we observed for candidate cis-regulatory regions , we find no evidence for a higher probability of positive selection on nonsynonymous sites in genes expressed in the fetal brain in either AAs ( Mann-Whitney U-test: p = 0 . 26 ) or EAs ( p = 0 . 78 ) . Our results are similar to a previous finding that positive selection in protein-coding regions is not elevated in genes expressed in the fetal brain [45] . In light of our results , it would seem that adaptive evolution of fetal brain development is influenced more strongly by changes at the regulatory level rather than at the protein-coding level . However , we note that genes expressed in the “whole brain” and cerebellum show a trend towards higher probabilities of positive selection within nonsynonymous sites in AAs ( p = 0 . 03 and 0 . 048 respectively , FDRs = 34% ) , suggesting that adaptive changes in the human brain may be the cumulative result of positive selection on regulatory regions during early development , and possibly on protein-coding regions in the adult brain . Previous studies have identified immune response and T cell-mediated immunity as processes that are enriched for genes showing signatures of positive selection in protein-coding regions [46] , [47] , highlight the importance of adaptive evolution in response to pathogens . We find that genes expressed in natural killer cells and T-cells both rank high among tissues with higher probabilities of positive selection in both coding and candidate cis-regulatory regions as compared to genes expressed in other tissues . Therefore , while genes expressed in the fetal brain show a different trend in coding and candidate cis-regulatory regions , positive selection in genes expressed in various immune cells may have occurred at both the coding and regulatory level . Evolutionary patterns within candidate cis-regulatory regions can also provide insight into the relative importance of functional categories in the evolution of modern humans . We generated a custom GOslim set containing 129 terms from the Gene Ontology database [48] , and performed Mann-Whitney U-tests to identify functional categories that have higher than expected probabilities of selection within candidate cis-regulatory regions ( Table S11 , Table S12 ) . In AAs , functional categories with a higher probability of positive selection in candidate cis-regulatory regions include regulation of cellular process ( GO:0050794 , p = 5 . 6×10−4 , FDR = 4% ) , protein modification ( GO:0006464 , p = 4 . 8×10−3 , FDR = 9% ) , and cell cycle ( GO:0007049 , p = 3 . 7×10−3 , FDR = 9% ) ( Table 6 ) . In EAs , the most significant terms include calcium ion binding ( GO:0005509 , p = 4 . 9×10−3 , FDR = 18% ) , organelle organization and biogenesis ( GO:0006996 , p = 5 . 4×10−3 , FDR = 22% ) , cell cycle ( GO:0007049 , p = 7 . 9×10−3 , FDR = 22% ) , and behavior ( GO:0007610 , p = 9 . 3×10−3 , FDR = 22% ) . Functional categories with a higher probability of negative selection in candidate cis-regulatory regions in AAs are generally less significant than positive selection , but include cytosol ( GO:0005829 , p = 8 . 6×10−3 , FDR = 16% ) , ribosome ( GO:0005840 , p = 0 . 02 , FDR = 16% ) , extracellular region ( GO:0005576 , p = 0 . 03 , FDR = 16% ) , and carrier activity ( GO:0005386 , p = 5 . 9×10−3 , FDR = 22% ) , and in EAs include proteinaceous extracellular matrix ( GO:0005578 , p = 0 . 01 , FDR = 19% ) , and extracellular space ( GO:0005615 , p = 0 . 03 , FDR = 19% ) ( Table 6 ) . We tested the same group of GO terms in our parallel tests for selection on nonsynonymous sites within coding regions , and find that different categories are identified as having higher probabilities of selection in nonsynonymous sites as compared to candidate cis-regulatory regions ( Table 6 ) . Terms showing significantly higher probabilities of positive and negative selection in nonsynonymous sites correspond to previously published results involving the same dataset with AAs and EAs pooled [35] . For example , we find that “transcription” has significantly higher probabilities of positive selection in nonsynonymous sites ( GO:0006350; AA: p = 1 . 1×10−3 , FDR = 8%; EA: p = 2 . 5×10−3 , FDR = 6% ) , and that “actin binding” has higher probabilities of negative selection in nonsynonymous sites ( GO:0003779: p = 8 . 0×10−3 , FDR = 25% ) . Transcription factors have frequently been reported as having a high degree of positive selection in protein-coding regions in general [35] , [45] , [49] , [50] , however this is not a trend that we observe in candidate cis-regulatory regions . Interestingly , in EAs we find that “calcium ion binding” has a significantly higher probability of negative selection at nonsynonymous sites ( GO:0005509 , p = 1 . 5×10−3 , FDR = 6% ) , but within candidate cis-regulatory regions it shows a higher probability of positive selection ( p = 4 . 9×10−3 , FDR = 18% ) . Therefore , evolutionary patterns in candidate cis-regulatory regions tend to exhibit different functional patterns than those found in protein coding regions of genes . We identified several genes that are likely to have undergone adaptive regulatory changes during human evolution ( 95% credibility interval on mean γ above 0 ) . NUDT16 shows the strongest evidence for positive selection in candidate cis-regulatory regions in both AAs ( Pr[γ>0 . 5] = 98 . 5% ) and EAs ( Pr[γ>0 . 5] = 97 . 4% ) . NUDT16 is a negative regulator of ribosome biogenesis , and may be involved in RNA decay in the nucleus . ITPR1 shows the second strongest evidence for positive selection in EAs ( Pr[γ>0 . 5] = 97 . 3% ) , however we find much weaker evidence for positive selection in AAs due to the presence of SNPs that are exclusive to AAs ( Pr[γ>0 . 5] = 70 . 7% ) . ITPR1 is essential for brain function , and is translated in response to synaptic activity in order to modulate calcium release from the endoplasmic reticulum . However ITPR1 also modulates calcium entry in the plasma membrane of B-cells , suggesting it may also have an important role in immunity . Functional studies have shown that the 3′ UTR of ITPR1 is required for dendritic localization in the mouse , however the majority of human-chimp fixed differences fall within conserved intronic sites rather than in the 3′ UTR region of this gene , suggesting that the target of selection is more likely to be an intronic cis-acting element ( or elements ) . Within nonsynonymous sites , ITPR1 demonstrates little evidence for either negative or positive selection in both AAs and EAs , suggesting that positive selection on ITPR1 likely involved adaptive substitutions within candidate cis-regulatory regions rather than within protein-coding regions . Additional genes with strong evidence for positive selection in candidate cis-regulatory regions in both populations include CACNA2D3 , a calcium channel subunit protein ubiquitously expressed in fetal tissues; OR2L13 , an olfactory receptor; and RNF167 , a gene involved in protein degradation . We also identified several genes that are likely to have experienced negative selection on candidate cis-regulatory regions ( 95% credibility interval on mean γ below 0 ) . The gene with the highest probability of negative selection in candidate cis-regulatory regions in AAs is FREM1 ( Pr[γ<−0 . 5] = 98 . 4% ) , a gene involved in the development of a number of epidermal structures in the mouse [51] , which also shows signatures of negative selection in EAs ( Pr[γ<−0 . 5] = 92 . 9% , but CI includes 0 ) . FREM1 is highly expressed in the dermis during mouse embryonic development , and truncation of the FREM1 protein results in blebbing ( blistering ) diseases that are similar to phenotypes observed in Fraser syndrome , and dystrophic epidermolysis bullosa in humans . The gene with the highest probability of negative selection in candidate cis-regulatory regions in EAs in KRT40 ( Pr[γ<−0 . 5] = 99 . 9% ) , a hair keratin protein that shows a similarly high probability of negative selection in AAs ( Pr[γ<−0 . 5] = 97 . 6% ) . However , positive values of Tajima's D in both EAs ( D = 1 . 80 ) and AAs ( D = 1 . 32 ) indicate that polymorphisms at KRT40 tend toward intermediate frequencies , suggesting that KRT40 may potentially be subject to balancing rather than negative selection . Other genes with strong evidence for negative selection in candidate cis-regulatory regions include SGCZ , a gene that may be important in the pathogenesis of muscular dystrophy and cardiomypathy; KIF19 , a gene involved in intracellular transport and a member of a superfamily of proteins important for brain functioning; DOCK1 , a gene thought to play a role in regulating phagocytosis during apoptosis; and TNNI3K , a cardiac-specific protein kinase . We also identify C14orf119 and C20orf117 as having strong evidence of negative selection in candidate cis-regulatory regions , however these genes have not been fully characterized . In order to examine patterns of natural selection on candidate cis-regulatory regions with respect to human disease , we identified 666 Mendelian disease genes using a hand-curated list of genes from the Online Mendelian Inheritance in Man database ( OMIM ) [52] , and 1 , 072 complex disease genes using the Genetic Association Database ( GAD ) [53] that were included in our scans for selection . We find that disease genes have a higher mean probability of negative selection within candidate cis-regulatory regions as compared to non-disease genes , however this trend is only suggestive in EAs , the population where the majority of diseases have likely been characterized ( Mann-Whitney U-test; OMIM: p = 0 . 23 in AAs , p = 0 . 011 in EAs; GAD: p = 0 . 29 in AAs , p = 0 . 06 in EAs ) . A link between negative selection and human disease has also been observed in protein-coding regions of the genome [35] , however genetic diseases can also be regulatory in nature [54] . There are several examples of disease-associated genes showing evidence for negative selection in candidate cis-regulatory regions in EAs . For example , LDB3 , is a gene expressed in skeletal and cardiac muscle for which several nonsynonymous mutations have been associated with myofibrillar myopathy ( OMIM:609452 ) and cardiomyopathy ( OMIM:601493 ) . Another gene , PLCE1 shows evidence for negative selection in both coding and candidate cis-regulatory regions , and homozygous mutations within the coding regions have been associated with type 3 nephrotic syndrome ( OMIM:610725 ) . Both LDB3 and PLCE1 show only moderate signatures of negative selection at candidate cis-regulatory regions in AAs . On the other hand , there are several examples of disease genes that show strong evidence of positive selection in candidate cis-regulatory regions . For example , CACNA2D3 displays a pattern where loss of heterozygosity in intronic sequences is associated with renal cell carcinoma; CENTG3 variants confer protection against the pathogenesis of polyglutamine disease in the brain; and ALG3 variants are associated with congenital disorder of glycosylation type Id , a metabolic disease . Direct comparisons between different classes of sites in mkprf may be confounded by differences in power to infer selection when there are varying degrees of selection on background loci in the dataset , particularly when there is no fixed variance on the prior distribution of γ . Our simulations show that as the proportion of negatively selected loci is varied , so does the number of genes showing strong signatures of positive selection due to a broader exploration of the parameter space ( CI>0 , Table 4 ) , despite the actual number of positively selected loci remaining constant . For example , by increasing the number of negatively selected loci from 0 ( neut+pos ) to 699 ( neut+wkdel+pos ) , to 1 , 485 ( neut+del+pos ) , while keeping the number of positively selected loci constant ( 1 , 824 loci ) , the number of genes with CIs>0 changes from 206 to 199 , to 203 ( Table 4 ) . If we increase the strength and the number of negatively selected loci to 4 , 632 loci ( neut+stdel+pos ) , we then identify 267 loci with CIs>0 ( Table 4 ) . More importantly , the distributions of γ for positively selected and neutral loci show visible differences when the number of negatively selected loci is varied , which may confound direct comparisons when different classes of sites are analyzed independently ( Figure S11 ) . In order to control for varying levels of selection on background loci , we ran a combined analysis with nonsynonymous , synonymous , and candidate cis-regulatory regions in a single run of mkprf . The effect of analyzing different classes of sites in a concurrent vs . independent analysis is a shift in the distribution of γ towards smaller values for both candidate cis-regulatory regions and synonymous sites , and a shift in the distribution of γ towards larger values for nonsynonymous sites , causing the distributions of γ to be more similar , and more closely centered around 0 for different classes of sites ( Figure S12 ) . First we looked for a correlation between selection acting on coding and candidate cis-regulatory regions , and observe what appears to be a complex relationship ( Figure 4 ) . In AAs we find a weak yet significant rank correlation between nonsynonymous and non-coding regions for both the probability of positive ( Kendall's tau = 0 . 055 , p = 1 . 6×10−11 ) and negative selection ( tau = 0 . 056 , p = 9 . 6×10−12 ) , and between synonymous and non-coding regions ( positive selection: tau = 0 . 060 , p = 2 . 1×10−13; negative selection: tau = 0 . 060 , p = 4 . 1×10−13 ) . The correlation between synonymous and nonsynonymous sites appears to be slightly stronger ( positive selection: tau = 0 . 098 , p<10−16; negative selection: tau = 0 . 096 , p<10−16 ) , as expected due to increased linkage disequilibrium from a closer proximity between sites . Results from the EA sample are similar . However , given the small values of tau , identifying genes with evidence for selection in coding regions may be a poor predictor of whether a gene will show evidence for selection in the same direction as candidate cis-regulatory regions . A recent study has found that genes with patterns of expression consistent with any direction of selection ( either positive or negative ) exhibit reduced rates of protein evolution on nonsynonymous sites [25] , which is consistent with only a weak correlation in the probability of selection between sites . The lack of a strong correlation in the mode of selection also suggests that the effect of linkage between candidate cis-regulatory and protein coding regions may be small , increasing the probability that we are detecting signatures of selection that are specific to non-coding regions . We next compared the overall distributions of the probability of selection at different classes of sites ( Figure 5 ) . Nonsynonymous sites have a higher mean probability of negative selection as compared to both candidate cis-regulatory regions and synonymous sites ( Mann-Whitney U-tests , p-values<10−16 ) , and candidate cis-regulatory regions have a significantly higher mean probability of negative selection as compared to synonymous sites ( p = 8 . 7×10−4 ) . We find that candidate cis-regulatory regions exhibit a significantly higher mean probability of positive selection as compared to nonsynonymous sites ( p<10−16 ) , however this is also true for synonymous sites ( p<10−16 ) , and synonymous sites have a marginally higher mean probability of positive selection as compared to candidate cis-regulatory regions ( p = 0 . 014 ) . Although synonymous sites may not evolve under strict neutrality , they may better represent the distribution expected under neutrality while taking into consideration local effects of linkage , mutation rates , and variability in effective population sizes across the genome . Therefore , differences in the distributions of the probability of positive selection between nonsynonymous and candidate cis-regulatory regions may be driven by more neutral , rather than more adaptive evolution in candidate cis-regulatory regions . However , it is possible we have limited power to identify differences in the degree of positive selection in protein-coding and candidate cis-regulatory regions , as the extent of positive selection within candidate cis-regulatory regions may be an underestimate due to their ascertainment . By restricting our analyses to human-mouse conserved sequences , it may have biased our dataset toward lower ratios of divergence/polymorphism based on neutral coalescent simulations ( Text S1 , Table S13 , Table S14 , Table S15 , Figure S13 , and Figure S14 ) . For example , if we restrict our analysis of nonsynonymous sites to those within human-mouse conserved regions ( i . e . use the same ascertainment as non-coding sites ) , all genes show strong evidence for strong selective constraint and have 95% credibility intervals below 0 on the mean estimate of γ ( Table 4 ) . Moreover , it is likely that only a small proportion of sites within candidate cis-regulatory regions are truly functional , as regulatory elements are often small . Therefore , the relative contribution of adaptive evolution at the level of gene regulation vs . changes in the actual protein remains an open question . An important consideration is whether selection on synonymous sites has affected our estimates of selection . While we observe no discernable relationship between the probability of positive or negative selection with GC content ( Figure S15 ) , negative selection may indeed affect a certain proportion of synonymous mutations [55] . Wright-Fisher simulations show that if purifying selection is acting on synonymous sites , we should expect a shift in the distribution of γ from a mean of 0 . 044 to 0 . 26 for a set of neutral loci ( Table 4 ) . On the other hand , selection on closely linked nonsynonymous sites may cause a reduction in polymorphism at synonymous sites , and could explain the signatures of positive selection we observe on synonymous sites ( Table 4 ) . Nevertheless , pooled synonymous sites provide a good fit to a neutral demographic model of expansion in the African American sample [32] , and were used as a neutral standard in the same way for all classes of sites . Therefore , any bias is expected to have a similar affect on candidate cis-regulatory , synonymous , and nonsynonymous sites . Our analysis of human polymorphism and divergence in conserved non-coding sites suggests that the evolution of candidate cis-regulatory regions is often driven by both positive and negative selection . Our findings reinforce the idea that the non-coding portion of our genome has an important functional and evolutionary role , and suggest that patterns of natural selection in non-coding DNA are often distinct from that of protein-coding regions . Many of the adaptive changes in candidate cis-regulatory regions might have occurred near genes expressed in the fetal brain , supporting the hypothesis that the evolution of the developing brain may be largely attributable to changes in gene regulation . Our results add to the increasing evidence that non-coding DNA is not all selectively neutral , and that selection on candidate cis-regulatory regions has played an important role throughout hominid evolution .
Sequencing and SNP detection was performed at Celera Genomics in 19 African Americans , 20 European Americans , and one chimpanzee as previously described [35] . Primers were designed to amplify the protein-coding exons of 23 , 363 genes according to Celera's human genome version R26k , which concurrently amplified non-coding flanking sequences around each exon . Primers were also designed to target human and mouse conserved sequences ( HMCS ) within 5 kb upstream of the first start codon ( for a total of 9 , 459 genes ) . HMCS were defined as sequences >100 bp and >70% identity between human and mouse , however the majority were between 250–500 bp in length . The bioinformatic pipeline is depicted in Figure S16 . We downloaded over 525 , 000 human and mouse conserved sequences ( HMCS ) from the UCSC genome browser [56] , of which 87 , 100 were aligned to the trace sequences using BLAST [57] . Sequences were oriented to hg17 using BLAT [58] , of which 85 , 641 mapped to a unique position on the human genome . The positions of all but five HMCS were updated to hg18 using liftOver [59] with a ratio of remapped bases set to 1 . A total of 83 , 379 HMCS were oriented to the public chimpanzee genome ( PanTro2 ) using syntenic alignments from the UCSC genome browser . HMCS were annotated to genes according to Refseq 19 with respect to sites 5 kb upstream of the transcription start site , 5 kb downstream of the transcription stop site , 5′ UTR , 3′ UTR , intron , and coding . Sites beyond 5 kb upstream or downstream from any known transcript were annotated as intergenic . All sites within HMCS that overlapped with any known coding exons were masked , leaving strictly conserved non-coding sites ( CNCs ) . Our resulting dataset included CNCs flanking a total of 15 , 061 autosomal genes . For comparisons to synonymous and nonsynonymous sites , we used the Celera Genomics exon resequence data from the same set of 35 people without regard to human and mouse conservation [35] . In the case of alternatively transcribed genes , we counted the total number of fixed and polymorphic sites in nonsynonymous sites as the union of all known transcripts . An analysis of admixture revealed that 4 out of the 19 African American individuals had high levels of European ancestry [33] , so these individuals were excluded from our study . In order to estimate the distribution of fitness effects on conserved non-coding sites ( CNCs ) in the flanking regions of genes for African Americans , we calculated the likelihood that the observed site frequency spectrum fits a neutral model , a model with a single estimate of γ , and model with a Gamma distribution of γ using the program prfreq [32] . Estimates of γ for intergenic CNCs were calculated relative to synonymous sites , and yielded no evidence for natural selection on intergenic CNCs . Estimates of γ for CNCs in the flanking regions of genes were then calculated relative to intergenic CNCs rather than synonymous sites in order to control for the effect of ascertainment based on conservation . Demographic parameters were simultaneously inferred from intergenic CNCs , which produced a similar model to that inferred from synonymous sites . In order to identify candidate cis-regulatory regions subject to positive and negative selection we implemented the program mkprf [36] , [60] , which estimates the posterior distribution of the population scaled selection coefficient ( γ = 2Nes ) for individual loci . For a neutral comparison we pooled the number of fixed and segregating synonymous sites from the exon resequencing data , similar to the approach taken in Bustamante et al . [35] . We updated the mkprf framework to incorporate the effect of non-stationary demography using the Poisson Random Field approach [61] , and used the maximum likelihood demographic parameters from Boyko et al . [32] that were estimated from the frequency information at synonymous coding sites . In this model , the African population exhibited a 3-fold expansion approximately 6 , 800 generations ago whereas the European population exhibited a bottleneck followed by a more recent expansion . We implemented mkprf by conservatively setting no fixed variance on the prior distribution of γ , as the number of genes with credibility intervals ( CIs ) above or below 0 has been shown to be positively correlated to the variance set on the prior distribution of γ [62] . For example , by not fixing the variance we identify 8 genes with CIs>0 and 7 genes with CIs<0 in our analysis of CNCs , whereas by fixing the variance to 8 we identify 11 genes with CIs>0 and 60 genes with CIs<0 . We performed Wright-Fisher forward simulations under the inferred demographic model for AAs using the program SFS_CODE [37] ( Table 4 ) . For each scenario a total of 11 , 000 loci were simulated using the length distribution of resequenced sites in candidate cis-regulatory regions ( or synonymous sites ) , the observed mutation rate from pooled synonymous sites ( θ = 5 . 91×10−4 ) , partial linkage between sites ( rho = θ ) , and a splitting from the chimpanzee ancestor 20*2N generations ago to match the observed fixed/segregating ratio in pooled synonymous sites ( N = 7 , 778 , the estimated ancestral population size estimated from synonymous sites [32] ) . Two sets of loci were simulated under a neutral demographic model ( one for candidate cis-regulatory regions , one for synonymous sites ) , and three sets of loci were simulated under different selective regimes ( one for candidate cis-regulatory regions , one for nonsynonymous , and one for synonymous sites ) . For candidate cis-regulatory regions a distribution of selective effects were drawn from a mixture of normal distributions assuming that most loci were under weak selection or nearly neutral ( N = 10 , 500 , mean = 0 , s . d . = 0 . 5 ) , but with some loci having more extreme selection coefficients ( N = 500 , mean = 0 , s . d . = 5 ) . This distribution assumed an equal number of genes exposed to positive and negative selection , and allowed us to evaluate our method under a general condition ( neut+del+pos , Table 4 ) . Loci with no informative sites were discarded , leaving a total of 9 , 863 loci in the neutral demographic set , and 9 , 707 loci in the neutral demographic+selection set that had at least 1 fixed or polymorphic site . In order to evaluate our method under different conditions , we randomly substituted half of the negatively selected genes for neutral loci ( neut+wkdel+pos ) , half of the negatively selected and all of the positively selected genes for neutral loci ( neut+wkdel ) , all of the negatively selected loci for neutral loci ( neut+pos ) , and all of the positively selected loci for neutral loci ( neut+del ) . For nonsynonymous sites a distribution of selective effects was drawn from an exponential distribution with rate parameter 0 . 2 , and truncated to the nearest integer . We then substituted all of the negatively selected loci that were simulated for nonsynonymous sites for an equivalent number of neutral loci in the neut+pos dataset ( neut+stdel+pos ) . For synonymous sites a distribution of selective effects was drawn from an exponential distribution with rate parameter 0 . 8 , reflected across the y-axis to be negative , and truncated to the nearest integer . All correlations with gene attributes were run using the continuous distribution of the probability of positive and negative selection ( Pr[γ>0 . 5] and Pr[γ<−0 . 5] ) . A Kendall's tau rank correlation coefficient was calculated in order to test for correlations between the probability of positive or negative selection and continuous biological variables . Mann-Whitney U-tests were used to test for differences in the mean probability of selection with discrete data , and multiple testing was considered using the false discovery rate [63] in the R statistical package using q . value . We created a custom GOslim including 129 terms , and annotated genes to the most terminal child to look for functional categories with higher probabilities of natural selection . In order to examine patterns of selection with regards to the transcriptional profiles of genes , we downloaded the Novartis Gene Expression Atlas 2 data [38] from 72 normal human tissues from the UCSC genome browser [56] . A gene was considered to be expressed in a tissue if the signal was >350 , and in the case of genes with multiple transcripts the average expression level for each tissue was calculated . The index of tissue specificity ( τ ) was estimated for each gene according to Yanai et al . [41] .
|
It has been suggested that changes in gene expression may have played a more important role in the evolution of modern humans than changes in protein-coding sequences . In order to identify signatures of natural selection on candidate cis-regulatory regions , we examined single nucleotide polymorphisms obtained from the complete re-sequencing of conserved non-coding sites ( CNCs ) in the flanking regions of over 15 , 000 genes in 35 humans . Patterns of allele frequencies in CNCs indicate the presence of both positive and negative selection acting on standing variation within these candidate cis-regulatory regions , particularly for the 5′ and 3′ UTRs of genes . Gene-specific tests comparing levels of polymorphism and divergence identify several genes with strong signatures of selection on candidate cis-regulatory regions and suggest that the biological characteristics of genes subject to selection are different between coding and candidate cis-regulatory regions with respect to gene expression and function . For example , we find stronger signatures of positive selection in candidate cis-regulatory regions near genes expressed in the fetal brain , which we do not observe in a concurrent analysis on protein-coding regions . Our results suggest that both positive and negative selection have acted on candidate cis-regulatory regions and that the evolution of non-coding DNA has played an important role throughout hominid evolution .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] |
[
"genetics",
"and",
"genomics/genomics",
"computational",
"biology/population",
"genetics",
"genetics",
"and",
"genomics/comparative",
"genomics",
"evolutionary",
"biology/evolutionary",
"and",
"comparative",
"genetics",
"evolutionary",
"biology/human",
"evolution",
"genetics",
"and",
"genomics/gene",
"expression",
"computational",
"biology/comparative",
"sequence",
"analysis",
"evolutionary",
"biology/genomics",
"evolutionary",
"biology/bioinformatics",
"computational",
"biology/genomics",
"genetics",
"and",
"genomics/bioinformatics",
"genetics",
"and",
"genomics/population",
"genetics"
] |
2009
|
Evolutionary Processes Acting on Candidate cis-Regulatory Regions in Humans Inferred from Patterns of Polymorphism and Divergence
|
Tick-borne diseases are the most prevalent vector-borne diseases in Europe . Knowledge on the incidence and clinical presentation of other tick-borne diseases than Lyme borreliosis and tick-borne encephalitis is minimal , despite the high human exposure to these pathogens through tick bites . Using molecular detection techniques , the frequency of tick-borne infections after exposure through tick bites was estimated . Ticks , blood samples and questionnaires on health status were collected from patients that visited their general practitioner with a tick bite or erythema migrans in 2007 and 2008 . The presence of several tick-borne pathogens in 314 ticks and 626 blood samples of this cohort were analyzed using PCR-based methods . Using multivariate logistic regression , associations were explored between pathogens detected in blood and self-reported symptoms at enrolment and during a three-month follow-up period . Half of the ticks removed from humans tested positive for Borrelia burgdorferi sensu lato , Anaplasma phagocytophilum , Candidatus Neoehrlichia mikurensis , Rickettsia helvetica , Rickettsia monacensis , Borrelia miyamotoi and several Babesia species . Among 92 Borrelia burgdorferi s . l . positive ticks , 33% carried another pathogen from a different genus . In blood of sixteen out of 626 persons with tick bites or erythema migrans , DNA was detected from Candidatus Neoehrlichia mikurensis ( n = 7 ) , Anaplasma phagocytophilum ( n = 5 ) , Babesia divergens ( n = 3 ) , Borrelia miyamotoi ( n = 1 ) and Borrelia burgdorferi s . l . ( n = 1 ) . None of these sixteen individuals reported any overt symptoms that would indicate a corresponding illness during the three-month follow-up period . No associations were found between the presence of pathogen DNA in blood and; self-reported symptoms , with pathogen DNA in the corresponding ticks ( n = 8 ) , reported tick attachment duration , tick engorgement , or antibiotic treatment at enrolment . Based on molecular detection techniques , the probability of infection with a tick-borne pathogen other than Lyme spirochetes after a tick bite is roughly 2 . 4% , in the Netherlands . Similarly , among patients with erythema migrans , the probability of a co-infection with another tick-borne pathogen is approximately 2 . 7% . How often these infections cause disease symptoms or to what extend co-infections affect the course of Lyme borreliosis needs further investigations .
Lyme borreliosis is the most prevalent tick-borne disease in humans , and is caused by spirochetes of the Borrelia burgdorferi sensu lato complex [1–3] . The most common clinical manifestation of early localized Lyme borreliosis is erythema migrans ( EM ) , an expanding skin lesion occurring after several days or weeks at the site of the tick bite . Other sporadically reported symptoms in this early stage of disease are malaise and viral-like symptoms . Disseminated Lyme borreliosis displays more severe manifestations that can involve a patient’s nervous system , joints , skin , and in rare cases the heart [1–3] . Tick-borne encephalitis ( TBE ) is the most common tick-borne central nervous system infection caused by the tick-borne encephalitis virus ( TBEV ) . Its clinical spectrum ranges from fever to mild meningitis and severe meningoencephalitis with or without paralysis [4] . In several European countries , there have been marked increases in the incidence of Lyme borreliosis and TBE over the past ten to twenty years [5–7] . In the Netherlands , a retrospective study among general practitioners has shown a continuing increase in consultations for tick bites and EM between 1994 and 2009 [8 , 9] . The increasing number of tick bites , adding up to 1 . 1 million tick bites in 2009 [8] , poses a growing risk of disseminated Lyme borreliosis and perhaps also of other tick-borne diseases . In the Netherlands , Ixodes ricinus ticks transmit several Borrelia burgdorferi s . l . genospecies , but are also infected with a variety of established or potentially pathogenic microorganisms , such as Borrelia miyamotoi , Anaplasma phagocytophilum , Candidatus Neoehrlichia mikurensis , several Babesia species , Rickettsia helvetica , R . monacensis and TBEV [10–15] . These ticks often carry multiple pathogens; at least one-third of the I . ricinus ticks carrying B . burgdorferi s . l . are co-infected with one or more pathogens from a different genus [12] , implying frequent exposure and possibly subsequent infection with several pathogens when humans are bitten by ticks . Remarkably , little is known about the incidences and clinical presentations of other tick-borne diseases than Lyme borreliosis and TBE . In general , disease caused by these other tick-borne pathogens , are associated with febrile illnesses with fever , headache , myalgia and malaise [16–21] . However , in immunocompromised patients chronic infections with severe clinical manifestations and even mortality have been described [17 , 18 , 21 , 22] . In the Netherlands , one single case of anaplasmosis has been reported in 1999 [23] , and one case of B . miyamotoi disease in an immunocompromised patient in 2012 [18] . It has been suggested that the severity of disease in Lyme borreliosis is affected by co-infections with other tick-borne pathogens [24–28] . Therefore , co-infections of B . burgdorferi s . l . with different tick-borne pathogens may possibly contribute to the high variety of clinical manifestations that are associated with Lyme borreliosis . Several reasons can be appointed for the absence in reporting of tick-borne diseases other than Lyme borreliosis and TBE , and the diagnosis of co-infections with other pathogens in Lyme borreliosis patients . Firstly , most of these infections might be self-limiting without overt or characteristic symptoms , often a clear-cut case definition of patients infected with one of these pathogens has not been established yet . Secondly , a poor performance or non-existence of supportive laboratory tests in routine medical microbiological settings . Thirdly , the lack of awareness among health professionals . Here , we aim to investigate i ) whether infection with tick-borne pathogens other than B . burgdorferi s . l . can be shown in patients with early localized Lyme borreliosis and in people exposed to tick bites in the Netherlands , and to determine ii ) the clinical picture of patients with DNA of tick-borne pathogens in their blood . Our approach is to test for the presence of nucleic acid ( DNA/RNA ) of the specific pathogens in human blood through amplification with PCR , especially since currently; there is no other specific laboratory diagnostic to detect infection with most of these tick-borne pathogens . Compared to DNA amplification with PCR , available serological tests generally have a low specificity and or sensitivity , particularly during the early phase of infection . In addition , although culturing is considered the most reliable method in proving the presence of microorganisms , it is time consuming , costly and often not possible for all pathogens .
Ticks , EDTA-blood and questionnaire data were available from a nationwide prospective observational study among patients who consulted one of 307 enrolling general practitioners for a tick bite or EM between May 2007 and December 2008 in the Netherlands , as described in detail [29] . All participants gave written informed consent , all minors who participated in the study had consent given from a parent/guardian , and the study protocol ( number 07-032/K ) was approved by the medical ethics committee of the University Medical Centre in Utrecht , the Netherlands . Patients were not eligible for participation when they were younger than six years of age , or when the tick bite had occurred outside the Netherlands . At enrolment , participants received the first set of study materials , containing a brochure about the study , an enrolment questionnaire , and materials for collection and mailing of first blood samples and removed ticks . Ticks removed from the skin were submitted in a small tube with 70% ethanol . In total , 314 ticks were obtained from 293 participants , of which 278 patients consulted their physician for a tick bite , and fifteen patients consulted their physician with an EM . Four ticks ( 1% ) were larvae , 167 ( 53% ) nymphs , 135 ( 43% ) adult ticks , and for eight ticks , the developmental stage could not be determined , as they had been damaged too much during removal from the patient’s skin . No other tick species than I . ricinus were identified . At enrolment , two tubes of blood were collected , 7 ml in a serum tube and 5 ml in an EDTA tube . Three months after enrolment , follow-up questionnaires and a consecutive 7 ml serum sample was collected from the tick bitten patients and from the EM patients after standard antibiotic treatment [29 , 30] . Seven patients who consulted their physician for a tick bite and in whom EM developed within the three month follow-up duration of the prospective study , were categorized in EM patient-group of the current study . EDTA-blood samples were available for molecular testing from 335 tick bitten patients and 291 EM patients . After arrival at the laboratory , ticks were stored at –20°C in ethanol . DNA was extracted using the DNeasy Blood and Tissue Kit ( Qiagen , Hilden , Germany ) according to the manufacturer’s instructions extraction for ticks . After total DNA extraction from ticks and amplification by PCR , reverse line blotting ( RLB ) was performed for Borrelia- , Ehrlichia- , Anaplasma- , Rickettsia- and Babesia-species . Further identification by DNA sequencing was performed as described [11 , 31] . PCR products that specifically reacted to the generic ( “catch all” ) probes , but that could not be further specified to the ( geno ) species level were designated as “untypeable” . Furthermore , our RLB analysis could not distinguish between B . garinii and B . bavariensis [12] . The presence of B . miyamotoi in ticks was tested by a real-time PCR amplification in 302 ticks , and Candidatus Neoehrlichia mikurensis in 312 ticks . The presence of TBEV RNA could not be screened in the tick samples , since only DNA had been extracted from these samples . Individual test results of the tick analyses were not reported to the participants or their physicians , in accordance with the informed consent form . Extraction of whole nucleic acid of the EDTA-blood samples were performed using robot-extraction ( MagNA Pure Compact Extraction Robot; Roche , Basel , Switzerland ) from 400 μL of EDTA-plasma ( Nucleic Acid Isolation Kit I; Roche ) according to the manufacturer’s instructions in a diagnostic laboratory setting . All 626 samples were analyzed with different real-time PCRs based on various genes specific for the microorganism of interest namely; B . burgdorferi s . l . , B . miyamotoi , A . phagocytophilum , Candidatus Neoehrlichia mikurensis , spotted fever Rickettsia's carried out on a LightCycler 480 ( Roche Diagnostics Nederland B . V , Almere , the Netherlands ) . For primers and probes , see S1 Table ( supplementary data ) . Reactions were done in a final volume of 20 μl with iQ multiplex Powermix , 3 μl of sample and 0 . 2 μM for all primers and different concentrations for probes . Positive plasmid controls and negative water controls were used on every plate tested . For detection of TBEV , multiplex a reverse transcription real-time PCR was performed as described before [32] . In brief , reactions were done in a final volume of 20 μl with TaqMan Fast Virus 1-Step Master Mix ( Thermo Fisher scientific , USA ) , 5 μl of sample and 0 . 2 μM for all primers and 0 . 2 μM probes ( S1 Table ) were added to the master mix and internal control was added to all the samples . With 20 min reverse transcription step at 50°C , denaturation at 95°C for 30 s and 50 cycles of 95°C for 10 s and 60°C for 30 s . The amplification was performed on a Roche LightCycler 480 instrument . For Babesia genospecies , we performed a conventional PCR targeting the 18S rRNA gene on all the blood samples [11] , followed by sequencing . To minimize cross contamination and false-positive results , negative controls were included in each batch tested by PCR . In addition , DNA/RNA extraction , PCR mix preparation , sample addition , and PCR analyses were performed in separated air locked dedicated labs . On all samples that were found positive in the real-time PCR , conventional PCRs were performed for confirmation on one or more targets followed by Tris-Borate-EDTA-agarose gel-electrophoresis . PCR products were sequenced , and these were compared with reference sequences from Genbank using Unweighted Pair Group Method with Arithmetic Mean-based ( UPGMA ) hierarchical clustering . Individual test results of these molecular analyses on EDTA-blood were not reported to the participants or their physicians , in accordance with the informed consent form . The prevalence of microorganism DNA detection in ticks and in EDTA-blood was calculated with 95% confidence intervals ( 95%CI ) based on mid-P exact . Characteristics of persons with or without DNA detected in blood by PCR were compared in Chi-square or Fisher’s exact test . We looked for associations between DNA detected in EDTA-blood by PCR and DNA detected in available ticks from the participants , tick engorgement , patient-reported tick attachment duration , antibiotic treatment at enrolment , and patient-reported symptoms at enrolment and after three months . Using multivariate logistic regression , we explored for associations between DNA detected in blood by PCR and self-reported symptoms at enrolment and follow-up . All reported clinical symptoms ( at enrolment and follow-up ) were included as predictive variables in the multivariate logistic regression models , after which the models were optimized using backwards elimination , until all predictive variables that were maintained in the model were statistically significant contributors ( p<0 . 05 ) . Statistical analyses were performed with SAS 9 . 4 ( SAS Inc . ) .
Table 1 shows the number of DNA sequences of the pathogens detected in 314 ticks obtained from 293 participants . Borrelia burgdorferi s . l . DNA was detected in 92 ( 29% ) ticks , as published earlier [29] . The ticks contained DNA of Candidatus Neoehrlichia mikurensis ( 5 . 4% ) , A . phagocytophilum ( 1 . 0% ) , Rickettsia species ( 22% ) , Babesia species ( 3 . 5% ) . and B . miyamotoi ( 2 . 3% ) . DNA of microorganisms of two or more genera were detected in 34 ticks ( 11% ) . Among the 92 B . burgdorferi s . l . positive ticks , 30 ( 33% ) also carried a pathogen of a different genus . About half of the ticks ( 149/314 , 47% ) tested negative for all genera . Table 2 shows the prevalence of DNA detection of tick-borne pathogens in EDTA-blood samples of 335 tick bitten patients and 291 EM patients , using various ( real-time ) PCRs . Only one ( 0 . 2% ) of 626 blood samples tested positive for B . burgdorferi s . l . and one ( 0 . 2% ) for B . miyamotoi in the real-time PCRs multiplex , both with high Ct values . In another multiplex , five blood samples ( 0 . 8% ) were positive for A . phagocytophilum and seven ( 1 . 1% ) for Candidatus Neoehrlichia mikurensis . Three ( 0 . 5% ) blood samples for Babesia genospecies yielded a sequence in conventional PCR , in which genetic analyses showed to be B . divergens . None of the samples were found positive for spotted fever Rickettsia's or TBEV . All seven of the Candidatus Neoehrlichia mikurensis sequence yielded a partial groEL sequence and five out of seven could also be confirmed on a separate gene , namely gltA . The seven groEL are 100% identical to each other as were the five gltA sequences ( Fig 1 ) . Four out of five A . phagocytophilum positives yielded a partial groEL sequence after nested PCR . The four groEL are almost identical to each other , with just one or two mismatches . Nevertheless , all four sequences are part of zoonotic variant of the A . phagocytophilum , ecotype I [33] . Three of the tested blood samples for Babesia genospecies yielded a sequence in conventional PCR for the ribosomal 18S rRNA gene , and showed to be identical to B . divergens sequences . Extensive efforts to generate a B . miyamotoi sequence failed . Accession numbers of the obtained sequences are: LC167302 , LC167303 , LC167304 , LC167305 . The prevalence of DNA of a tick-borne pathogen other than B . burgdorferi s . l . detected in blood from persons after a tick bite was 2 . 4% ( Table 2 ) , this number was similar to EM patients ( 2 . 7% ) . In the blood of one person DNA of both A . phagocytophilum and B . divergens were detected . Altogether , DNA of tick-borne pathogens was detected in the blood of sixteen persons . The characteristics of these sixteen participants are summarized in Table 3 , with regard to age , gender , self-reported clinical symptoms , antibiotic treatment and tick exposure at enrolment and during the three month follow-up period . Eight of the sixteen cases had submitted ticks at enrolment . Among these eight ticks , six ticks tested negative in PCR , and in two ticks DNA was detected of a different genus than the tick-borne pathogens that had been detected in the EDTA-blood of the corresponding participants ( Table 3 ) . We did not observe associations between detection of tick-borne pathogen DNA in EDTA-blood and; patient-reported tick attachment duration , with tick engorgement , with antibiotic treatment at enrolment . Seven out of sixteen cases reported clinical symptoms at enrolment or during the three month follow-up period , such as myalgia ( 3 cases ) , skin rash ( 2 cases ) , tingling sensations in limbs ( 2 cases ) , fatigue , arthralgia , headache , pain in limbs , and gastrointestinal symptoms/vomiting . Using multivariate logistic regression , we compared the prevalence of self-reported symptoms , tick attachment duration and tick engorgement among cases with and without DNA of tick-borne pathogens detected in blood . Compared to the cases that tested negative by PCR , the cases with DNA of tick-borne pathogens detected in their blood sample were not more likely to report any of the named clinical symptoms at enrolment or at follow-up .
In this study , DNA of tick-borne microorganisms was detected and identified in ticks and human blood samples ( Tables 1 and 2 ) . The limitations of this methodology are well known; hence , the interpretation of these results should be done with caution [34] . In order to unequivocally prove the presence of the corresponding infectious agents in ticks or blood , their viability should be tested by in vitro culture or infection experiments of laboratory animals . Also , the absence of DNA of a pathogen cannot be interpreted as the absence of the infectious agent . Besides the technical detection limits of PCR-based methods , the timing of sample collection after a tick bite and start of an antibiotic treatment , as well as the tissue tropism of the pathogen strongly affect the ability of pathogen detection [35 , 36] . The latter is corroborated in this study: Only in one out of the 291 patients with an erythema migrans ( EM ) -a skin infection caused by B . burgdorferi s . l . —the DNA of this pathogen was detected in blood ( Table 2 ) . This finding confirms that the chance of detecting B . burgdorferi s . l . DNA in blood samples of confirmed Lyme borreliosis patients is very low [35] . Rickettsia helvetica and R . monacensis were both not detected in the 626 blood samples whereas , recently molecular evidence for their presence in cerebrospinal fluid of neuroborreliosis patients and in a skin sample of an EM patient was found [37 , 38] . The absence of TBEV in blood samples can further be explained by its extremely low infection rates in ticks and focal geographic distribution in the Netherlands [15] . Candidatus Neoehrlichia mikurensis , A . phagocytophilum , B . miyamotoi and Babesia species are all pathogens that can be expected in blood because of their biology and tissue tropism [16 , 20 , 39 , 40] . The tick samples were screened by a different method ( RLB ) than the blood samples ( real-time PCR ) . In 314 ticks removed from humans a wide variety of tick-borne pathogens were detected namely , Borrelia afzelii , Borrelia garinii , Borrelia burgdorferi sensu stricto , Borrelia valaisiana , Babesia microti , Babesia venatorum , Babesia divergens , Anaplasma phagocytophilum , Candidatus Neoehrlichia mikurensis , Rickettsia helvetica , Rickettsia monacensis and Borrelia miyamotoi . All these pathogens have been found in questing ticks from field studies in the Netherlands before [41–43] . The infection rate of tick-borne pathogens other than B . burgdorferi s . l . varied from 0 . 3% ( B . divergens and R . monacensis ) up to 18 . 8% ( R . helvetica ) . With an estimated incidence of 1 . 1 million tick bites per year , human exposure to a tick-borne pathogen other than B . burgdorferi s . l . and TBEV varies from roughly 3500 for B . divergens , and 3500 for R . monacensis to 207 , 000 persons for R . helvetica . Among the 322 , 000 persons exposed to B . burgdorferi s . l . through a tick bite , roughly 105 , 000 are simultaneously exposed to another pathogen . In addition , exposure to more than one tick-borne pathogen can occur when people have more than one tick bite at once or several consecutive tick bites . Clearly , not all exposure to tick-borne pathogens results in human infection . Based on the development of an EM or seroconversion , the risk of infection with B . burgdorferi s . l . after tick bites was estimated to be 5 . 1% [29] . In this study , evidence for infection comes from the detection of Candidatus Neoehrlichia mikurensis , A . phagocytophilum , B . divergens , B . miyamotoi and B . burgdorferi s . l . DNA in the blood of sixteen individuals after exposure to a tick bite . None of these cases reported to be immunocompromised , and all the EM patients were treated with antibiotics according to the guidelines for treatment of Lyme borreliosis [30] . Mild clinical symptoms were reported by seven out of sixteen PCR-positive cases . However , using multivariate logistic regression , we did not detect associations between DNA detected in blood and self-reported symptoms at enrolment and follow-up . Furthermore , we did not find associations between detection of DNA of tick-borne pathogens in blood and; PCR positive ticks , patient-reported tick attachment duration , tick engorgement , and antibiotic treatment at enrolment . The lack of statistically significant associations may be due to the mildness of symptoms amongst immune-competent patients , and to a lesser degree due to insufficient numbers of PCR-positive cases per pathogen genus in our analyses . In this study , Candidatus Neoehrlichia mikurensis infection was observed in 1 . 1% ( 95%CI 0 . 5%–2 . 2% ) . Candidatus Neoehrlichia mikurensis infections have been described in immunocompromised patients [44] , and more recently in immune-competent patients with relatively mild symptoms in China , Poland , and Sweden [45–48] . Anaplasma phagocytophilum infection was found in 0 . 9% ( 95%CI 0 . 3%–2 . 0% ) of the persons exposed to tick bites in the Netherlands ( Table 2 ) . Genetic analyses of the DNA sequences showed the highest similarity to the zoonotic A . phagocytophilum ecotype I [33] . Evidence for A . phagocytophilum infection is primarily based on its molecular , microscopic or serological detection most disease cases [20 , 49] . There is serological evidence that A . phagocytophilum infection occurs in the absence of disease symptoms [50] . Babesia divergens infection was observed in 0 . 5% ( 95%CI 0 . 1%–1 . 3% ) of the persons exposed to tick bites . In Europe , only two cases of human babesiosis have been reported in immune-competent patients , one due to B . divergens [51] . Only one patient with EM was possibly infected with B . miyamotoi 0 . 2% ( 95%CI 0 . 0%–0 . 8% ) . The presence of B . miyamotoi DNA could only be determined by real-time PCR , and several attempts to confirm this finding by conventional PCR was unsuccessful . This patient had received antibiotic treatment at enrolment for his EM , so a low bacterial load in blood due to the treatment could be an explanation for the high Ct value . Evidence for infection with B . miyamotoi in Europe comes from one immunocompromised case [18] , and a seroprevalence study in people exposed to tick bites [52] . Altogether , the probability of infection with a tick-borne pathogen other than Lyme spirochetes after tick bites in the Netherlands is roughly 2 . 4% ( 95%CI 1 . 1%–4 . 5% ) . This number is similar to the probability of a co-infection with another tick-borne pathogen in patients with EM ( 2 . 7% , 95%CI 1 . 3%–5 . 2% ) . Interestingly , one patient in this study had a co-infection with A . phagocytophilum and B . divergens . The severity of self-reported symptoms of the seven EM patients with a co-infection was indistinguishable from patients only having EM . No indications were found that infection with a tick-borne pathogen other than B . burgdorferi s . l . caused overt symptoms that would indicate a corresponding illness . The low number of persons with a tick bite or EM that were identified with an tick-borne pathogen infection other than B . burgdorferi s . l . , in combination with the limited medical assessments , and the used method of pathogen detection are not sufficient to infer how often tick-borne pathogens other than B . burgdorferi s . l . ( and TBEV ) cause disease . Also , to what extend they affect the diagnoses and the etiology of Lyme borreliosis . Furthermore , the ability for a pathogen to cause disease depends also on extrinsic factors for example the immune status of its host . The high exposure to tick-borne pathogens other than B . burgdorferi s . l . and TBEV , and their ability to cause infection in the general population , warrants increased awareness , knowledge , improvement of diagnostic tests and a clear-cut clinical case definitions in an European setting . Only when better laboratory tests are available for these tick-borne diseases , their impact as a co-infection with Lyme borreliosis can be assessed .
|
Two most common tick-borne diseases in Europe are Lyme borreliosis and tick-borne encephalitis . Ticks transmit many more pathogens , causing neglected diseases such as anaplasmosis , babesiosis , rickettsiosis and neoehrlichiosis . These diseases are seldom diagnosed , due to their mild and non-characteristic symptoms , but also due to lack of awareness and availability of diagnostic tests . Using molecular detection techniques ( polymerase chain reaction or PCR ) , we estimated the frequency of tick-borne infections in humans after a tick bite and in patients with the first symptoms of Lyme borreliosis , an erythema migrans . About half of the ticks that fed on humans carried one or more tick-borne pathogens , and approximately 2 . 5% of people that were bitten by ticks were infected with a tick-borne pathogen other than Lyme borreliosis or tick-borne encephalitis . Co-infections of a tick-borne pathogen in patients with an erythema migrans was also approximately 2 . 5% . Based on these findings , we estimated the incidence of tick-borne infections other than Lyme borreliosis in the Netherlands . How often these infections cause disease or to what extend co-infections affect the course of Lyme borreliosis needs further investigations .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"borrelia",
"infection",
"rheumatology",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"parasite",
"groups",
"body",
"fluids",
"ixodes",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"animals",
"parasitic",
"protozoans",
"parasitology",
"organisms",
"bacterial",
"diseases",
"apicomplexa",
"lyme",
"disease",
"protozoans",
"molecular",
"biology",
"techniques",
"ticks",
"bacteria",
"bacterial",
"pathogens",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"borrelia",
"burgdorferi",
"artificial",
"gene",
"amplification",
"and",
"extension",
"medical",
"microbiology",
"epidemiology",
"microbial",
"pathogens",
"borrelia",
"molecular",
"biology",
"hematology",
"disease",
"vectors",
"arthropoda",
"arachnida",
"blood",
"anatomy",
"babesia",
"physiology",
"biology",
"and",
"life",
"sciences",
"polymerase",
"chain",
"reaction"
] |
2016
|
Molecular Detection of Tick-Borne Pathogens in Humans with Tick Bites and Erythema Migrans, in the Netherlands
|
C . canimorsus 5 has the capacity to grow at the expenses of glycan moieties from host cells N-glycoproteins . Here , we show that C . canimorsus 5 also has the capacity to deglycosylate human IgG and we analyze the deglycosylation mechanism . We show that deglycosylation is achieved by a large complex spanning the outer membrane and consisting of the Gpd proteins and sialidase SiaC . GpdD , -G , -E and -F are surface-exposed outer membrane lipoproteins . GpdDEF could contribute to the binding of glycoproteins at the bacterial surface while GpdG is a endo-β-N-acetylglucosaminidase cleaving the N-linked oligosaccharide after the first N-linked GlcNAc residue . GpdC , resembling a TonB-dependent OM transporter is presumed to import the oligosaccharide into the periplasm after its cleavage from the glycoprotein . The terminal sialic acid residue of the oligosaccharide is then removed by SiaC , a periplasm-exposed lipoprotein in direct contact with GpdC . Finally , most likely degradation of the oligosaccharide proceeds sequentially from the desialylated non reducing end by the action of periplasmic exoglycosidases , including β-galactosidases , β-N-Acetylhexosaminidases and α-mannosidases .
Capnocytophaga are capnophilic Gram negative bacteria that belong to the family of Flavobacteriaceae in the phylum Bacteroidetes and colonize the oral cavity of diverse mammals including humans [1] , [2] . Capnocytophaga canimorsus , a usual member of dog's mouths flora [3] , [4] , was discovered in 1976 [5] in patients that underwent dramatic infections after having been bitten , scratched or simply licked by a dog . These infections occur , worldwide , with an approximate frequency of one per million inhabitants per year . They generally begin with flu symptoms and evolve in a few days into fulminant septicaemia and peripheral gangrene with a mortality as high as 40% [5] , [6] , [7] , [8] , [9] . A few recent observations help understanding the high aggressiveness of C . canimorsus for humans . First , C . canimorsus are able to escape complement killing and phagocytosis by human polymorphonuclear leukocytes ( PMN's ) [10] , [11] . They also escape detection and phagocytosis by macrophages , which results in a lack of release of pro-inflammatory cytokines [12] . In addition to this passive evasion from innate immunity , 60% of the strains are able to block the killing of Escherichia coli phagocytosed by macrophages [3] , [10] and some strains even block the onset of pro-inflammatory signalling induced by an E . coli lipopolysaccharide ( LPS ) stimulus [12] . The molecular bases of these immunosuppressive mechanisms are not understood yet . However , their study led to the serendipitous discovery that the fastidious C . canimorsus grow readily upon direct contact with mammalian cells including phagocytes . This property was found to be dependent on a sialidase ( SiaC ) allowing C . canimorsus to harvest amino sugars of glycan chains from host cell glycoproteins [13] . Recently , we reported the complete 2 , 571 , 405-bp genome sequence and the surface proteome of strain Cc5 ( Manfredi et al . , submitted ) . Among others , this study unravelled the existence of 13 complex feeding systems encoded by polysaccharide utilization loci ( PULs ) , a hallmark of the Cytophaga-Flavobacteria-Bacteroides ( CFB ) group [14] , [15] . The archetype of these systems is the Sus system , pioneered by the laboratory of A . Salyers and allowing Bacteroides thetaiotaomicron to forage starch . It is composed of the surface-exposed SusCDEF protein complex [15] , [16] and the SusAB periplasmic proteins [17] . SusC resembles a TonB-dependent transporter essential for energy-dependent import of starch oligosaccharides into the periplasm [18] while SusD is a α-helical starch-binding lipoprotein [19] , [20] . SusE and SusF are other surface-exposed lipoproteins that reinforce starch binding [17] . Finally , the outer membrane α-amylase SusG hydrolyses surface-bound starch [19] . B . thetaiotaomicron has 88 of these PULs , identified essentially by the presence of a pair of adjacent susC-like and susD-like alleles . Interestingly , expression of some PULs is upregulated in the presence of mucin O-glycans or glycosaminoglycans ( GAGs ) , indicating that B . thetaiotaomicron also forages on host glycans , primarily the O-glycosylated mucin [14] but these glycoprotein foraging systems have not been characterized so far . Although Streptococcus oralis , a firmicute from the human oral flora and S . pneumoniae have been shown to remove and metabolize N-linked complex glycans of human glycoproteins [21] , [22] , [23] , no Sus-like-N-linked glycan foraging system has been described in detail . One of the 13 Sus-like systems from C . canimorsus 5 is involved in the capacity to grow on mammalian cells and to deglycosylate glycoproteins . It is encoded by chromosome locus PUL5 and accounts for 12% of the Cc5 surface proteins . Since it contributes to survival in mice . it can be seen as a new type of bacterial virulence factor ( Manfredi et al . , submitted ) . Here , we present a detailed molecular characterization of this PUL5-encoded foraging complex and we show that it deglycosylates human immunoglobulins G ( IgG ) . Among others , we show that it cleaves N-linked glycan moieties between two GlcNAc residues and we show its functional relation with sialidase .
PUL5 consists of the five genes Ccan_08700–Ccan_08740 . Ccan_08700 encodes a SusC-like integral outer membrane ( OM ) protein presumably forming a pore in the OM while Ccan_08710 is a SusD-like protein presumably involved in substrate binding [20] . Since the locus was shown to confer the capacity to deglycosylate proteins ( Manfredi et al . , submitted ) , we named the five genes gpd ( for glycoprotein deglycosylation ) and we called gpdC and gpdD the genes encoding homologs to SusC and SusD , respectively . The five gpd genes seem to be organized as an operon in the order gpdC , gpdD , gpdG , gpdE and gpdF ( Fig . 1A ) . GpdG is predicted to be an endo-β-N-acetylglucosaminidase and GpdE has similarities with the Concanavalin A-like lectins/glucanases superfamily on its 108 C-terminal amino acids and could have a substrate-binding role analogous to that of GpdD . Finally , GpdF shows homology to the galactose-binding domain-like superfamily on its 136 C-terminal amino acids suggesting again a role in glycan binding . In order to investigate what is the function of the individual Gpd proteins we constructed single gpd genes knockout strains . None of the knockout mutants was significantly affected in its growth on blood agar plates . In contrast , deletion of any of the gpdC , -D , -G or -E genes led to a clear reduction of growth on HEK293 cells while deletion of gpdF had only a slight effect ( Fig . 1B ) . Complementation of the deleted genes with plasmid-borne genes expressed from the natural gpdC promoter completely restored growth to the wildtype ( wt ) level indicating that none of the mutation was polar . In order to determine whether the reduced growth of the mutants was due to a defect in protein deglycosylation , we incubated wt Cc5 bacteria and the gpd mutant bacteria with fetal calf serum protein fetuin , taken as a standard glycoprotein . Fetuin contains 3 O-linked glycans ( accounting for 20% of fetuin-bond carbohydrates ) and 3 N-linked glycans ( 80% of fetuin-bond carbohydrates ) [24] . We monitored glycosylation by staining with Sambucus nigra agglutinin ( SNA ) , a lectin that recognizes terminal sialic acids on glycans . As shown in Fig . 1C , fetuin that had been incubated with wt Cc5 reacted much less with SNA and appeared as two , still sialylated smaller degradation products . As shown by Manfredi et al . ( submitted ) , this indicated that partial deglycosylation had occurred and progressed further than a simple desialylation . In contrast , fetuin that was incubated with the gpdC , -D , -G and -E mutant bacteria was unaffected , indicating that no desialylation occurred in the absence of these gpd genes , although sialidase SiaC [13] was not directly affected . Fetuin incubated with the gpdF mutant showed a slight desialylation indicating that fetuin deglycosylation was not completely abolished as with the other mutants . Fetuin glycosylation was also monitored by immuno-blotting with anti-fetuin antibodies . As shown in Fig . 1D , the size of fetuin was shifted down after incubation with wt Cc5 bacteria while the protein migration rate was unchanged after incubation with the gpdC , -D , -G and -E mutant bacteria . After incubation with gpdF mutant bacteria , fetuin did undergo a size shift but not as marked as when incubated with wt bacteria . Taken together these results indicate that partial fetuin deglycosylation was strictly dependent on the activity of proteins GpdC , -D , -G , -E and , to a lesser extend -F . Finally , our data strongly suggest that the defect in growth of the gpd mutants onto HEK293 cells was due to a defect in the ability to deglycosylate host glycoproteins . GpdG is annotated as an endo-β-N-acetylglucosaminidase ( Manfredi et al . , submitted ) , i . e an endo-glycosidase that cleaves N-linked glycan structures at the base of the glycan in between two GlcNAc molecules . Hence , it should leave one GlcNac molecule attached to the protein . Fetuin is reported to be glycosylated on the three asparagine residues N99 , N156 and N176 [24] . Analysis by liquid chromatography-mass spectrometry ( LC-MS ) of trypsin-digested fetuin showed that the main glycosylation site resides on N156 . The m/z is in accordance with the peptide LCPDCPLLAPLNDSR carrying the GlcNAc5Man3Gal3Sial3 sugar ( Fig . 2A ) . N176 was found to carry a sugar with a Hex6HexNAc5NeuAc4 composition , but its site occupancy was much lower than N156 . Only trace amounts of glycans were found attached to N99 . After incubation of fetuin with wt Cc5 bacteria , LC-MS analysis revealed the presence of a peptide whose mass indicated that only one GlcNAc moiety remained linked to N156 ( Fig . 2B ) . The fragmentation spectrum of this peptide fully confirmed the presence of the GlcNAc moiety on N156 ( Fig . 2C ) . Due to the low site occupancy of N176 , deglycosylation of N176 to the GlcNAc moiety was too weak to be detected . The conversion of GlcNAc5Man3Gal3Sial3 to GlcNAc on N156 suggests an endo-β-N-acetylglucosaminidase dependent deglycosylation . To confirm that fetuin deglycosylation was due to the Gpd complex activity and in particular to the GpdG glycosyl hydrolase activity , we then analysed fetuin after incubation with the gpdG knockout bacteria . Fetuin incubated in the presence of these mutant bacteria turned out to remain fully glycosylated ( Fig . 2D ) indicating that no cleavage occurred in the absence of the enzyme . The sequence of GpdG was then compared to those of two endo-β-N-acetylglucosaminidases , namely EndoS from Streptococcus pyogenes capable of deglycosylating N-linked glycans from the γ chain of human immunoglobulins [25] , and EndoF from Flavobacterium meningosepticum capable of cleaving off high-mannose and complex glycan N-linked from several glycoproteins including immunoglobulins [26] . It appeared that a chitinase motif present in these two enzymes was conserved in GpdG ( FDGFDIDWE ) . In order to further confirm the endo-β-N-acetylglucosaminidase activity of GpdG we substituted the essential E205 residue [26] with a glycine and tested the growth on HEK293 cells of the gpdG mutant strain expressing in trans the GpdG catalytic mutant . As shown in Fig . 3A , the GpdG catalytic mutant was impaired in growth . We then tested the fetuin deglycosylation ability of the GpdG catalytic mutant . As shown by the lectin staining in Fig . 3B and by the immuno-blotting in Fig . 3C , bacteria endowed with the GpdG catalytic mutant were completely impaired in fetuin deglycosylation . We conclude from all these experiments that GpdG is an endo-β-N-acetylglucosaminidase . Since GpdG has the same chitinase motif as EndoF and EndoS , known to deglycosylate N-linked glycans from the γ chain of human IgGs [25] , [26] , we tested whether the Gpd complex would also be able to deglycosylate the heavy chain of IgGs . Cleavage of the N297-linked glycan moiety by EndoS was shown to determine a size shift of ∼3 KDa [25] . After incubation of purified human IgG with wt Cc5 bacteria , the molecular mass of the γ chain underwent a slight size shift ( Fig . 4A and B ) while the mass of the light chains was unchanged ( Fig . 4A ) . In contrast incubation with ΔgpdG knockout bacteria did not alter the γ chain size indicating that the cleavage was GpdG dependent . To confirm that the size reduction of the γ chain was due to the removal of the glycan moiety , IgG was stained with SNA . As shown in Fig . 4C , the SNA signal of the γ chain was significantly reduced after incubation with wt Cc5 . In contrast the γ chains remained fully glycosylated after incubation with ΔgpdG bacteria . Analysis by LC-MS of trypsin-digested IgG showed that N297 from this sample of IgG mainly bears a GlcNAc4Man3Gal2Sial1Fuc1 chain or a GlcNAc4Man3Gal2Fuc1 chain ( Fig . 5A–B ) . After incubation of this sample of IgG with wt Cc5 bacteria , LC-MS analysis revealed the presence of a peptide whose mass indicated that only a GlcNAc1Fuc1 moiety remained linked to N297 ( Fig . 5C ) . The fragmentation spectrum of this peptide fully confirmed the presence of the GlcNAc1Fuc1 moiety on N297 ( Fig . 5D ) . LC-MS analysis of IgG after incubation with ΔgpdG bacteria revealed exactly the same profile as in the untreated sample indicating that no deglycosylation has occurred ( Fig . 5E–F ) . The conversion of GlcNAc4Man3Gal2Sial1Fuc1 and GlcNAc4Man3Gal2Fuc1 to GlcNAc1Fuc1 on N297 indicates an endo-β-N-acetylglucosaminidase ( GpdG ) dependent deglycosylation . These data indicated that , like F . meningosepticum and S . pyogenes , C . canimorsus has the capacity to deglycosylate IgGs . The GpdD , -G , -E and –F proteins belong to the OM and surface proteomes of Cc5 ( Manfredi et al . , submitted ) . In addition , these proteins are endowed with a signal peptidase II consensus signal peptide . Altogether , this suggests that they could be lipoproteins anchored to the outer leaflet of the outer membrane and exposed at the surface of the bacterium ( Manfredi et al . , submitted ) . In order to determine whether the lipidation of the Gpd proteins is required for their function , we generated soluble periplasmic versions of GpdD and GpdG by substituting the cystein residue of the lipobox with a glycine . We then tested the ability of the periplasmic variants of GpdD and GpdG to complement the growth deficiency of the gpdD and gpdG knockout strains on HEK293 cells . As shown in Fig . 6 , both the GpdD and GpdG periplasmic variant were unable to complement the growth deficiency indicating that lipid modification is necessary for the proper localization and function of the proteins . This conclusion was reinforced by the fact that bacteria endowed with periplasmic GpdD or GpdG were unable to deglycosylate fetuin ( Fig . 6 ) . Hence , we infer that GpdD and GpdG are lipoproteins that are anchored in the outer leaflet of the outer membrane and exposed to the bacterial surface . The same presumably applies to GpdE and GpdF since they have also a lipobox and they are also part of the surface proteome ( Manfredi et al . , submitted ) . In order to assay whether the five Gpd proteins interact with each other to form a complex at the bacterial surface , we performed a two-step affinity purification with a His-Strep tagged version of GpdC . Analysis by immuno-blot and mass spectrometry ( Fig . 7 ) of the purified fraction revealed the presence , together with GpdC , of GpdD , -G , -E and –F , indicating a stable interaction between all these proteins . Furthermore , six other proteins , among which SiaC ( Fig . 7 ) , co-purified with the complex . SiaC has been previously shown [13] to be essential to sustain growth of Cc5 in the presence of eukaryotic cells due to its role in the glycoprotein deglycosylation process . We thus focused our attention on the sialidase-Gpd complex interaction . The co-purification of SiaC with GpdC strongly suggested that SiaC is associated to the Gpd complex , although it is encoded far away from PUL5 . However , unlike the Gpd proteins , sialidase was not identified in the surface proteome of Cc5 ( Manfredi et al . , submitted ) . On the other hand , earlier immunofluorescence assays suggested that sialidase is localized on the bacterial surface and removal of the signal sequence of sialidase prevented growth on cells [13] . In order to better understand the interplay between SiaC and Gpd proteins in the glycoprotein deglycosylation process , we decided to clarify its localization . Since the sialidase sequence analysis revealed the presence of a signal peptide with a lipobox in the N-terminal sequence , we first sought to determine whether SiaC is a lipoprotein . We incubated Cc5 and sialidase mutant ( siaC ) bacteria encoding SiaCC17Y in the presence of tritiated palmitate and analyzed the total proteins by SDS-PAGE and fluorography ( Fig . 8A ) . Sialidase appeared indeed to be lipidated and the C17Y mutation completely prevented this lipid modification . The analysis of outer membrane proteins isolated by sarcosyl extraction confirmed that sialidase but not its C17Y variant was associated with the OM ( Fig . 8B ) . We conclude from these experiments that SiaC is a lipoprotein anchored into the outer membrane . In order to define whether it is exposed towards the outside like GpdDGEF or towards the periplasm , we tested whether the periplasmic SiaC17Y could restore the growth deficiency of the siaC mutant strain . In contrast to what was observed for GpdD and GpdG , expression of SiaC17Y in trans did fully restore the growth defect ( Fig . 8C ) indicating that the localization of sialidase in the periplasm and the absence of association with the outer membrane did not prevent its function . This data pointed to the direction of a periplasmic localization of SiaC rather than a surface-exposed localization as was previously suggested [13] . The association between sialidase and the Gpd complex obviously suggests that the two work cooperatively . This was already suggested by the fact that the gpd mutant bacteria did not remove the terminal sialic acid residues from fetuin , although SiaC was functional in these mutants ( Fig . 1C ) . We then tested the ability of the siaC knockout bacteria to deglycosylate fetuin . SNA lectin staining ( Fig . 8D ) and immuno-blotting ( Fig . 8E ) clearly showed the same fetuin deglycosylation pattern for the wt and siaC mutant bacteria . These results indicate that the endo-cleavage of fetuin N-glycans , operated by the Gpd complex is completely independent from the activity of SiaC . However , the evidence that SiaC activity is essential for growth on Hek293 cells ( Fig . 8C ) , suggests that removal of the glycan terminal sialic acid is nevertheless a crucial step for the subsequent glycan catabolism process . This indicates that the Gpd complex acts upstream of SiaC . Since the Gpd complex includes the GpdC porin-like protein , this sequential order is perfectly compatible with a periplasmic localization of sialidase . Sialic acid removal would thus occur in the periplasm after the glycan has been cleaved off and transported through the GpdC OM channel . If this model was correct , the interaction between the periplasmic SiaC and the GpdC complex could only occur through a direct interaction with GpdC , since the other Gpd proteins are surface exposed . To test this prediction , we expressed a C-terminally Strep-His double tagged GpdC in a gpdCDGE multi knockout strain and we performed a two-step affinity purification of GpdC . The analysis by immuno-blotting ( Fig . 8F ) of the fractions eluted after the second purification step showed that SiaC did indeed co-purify with GpdC indicating that SiaC and GpdC do indeed interact directly with each other . The complete deglycosylation complex would thus consist of the surface-exposed lipoproteins GpdDGEF and the periplasm-exposed lipoprotein SiaC , all of them associated to the porin-like GpdC ( Fig . 8 ) .
Our previous work has shown that C . canimorsus deglycosylates surface glycoproteins from the host and sustains its growth on the glycan moieties [13] . Here , we showed that this deglycosylating activity is achieved by the joined action of the PUL5-encoded Gpd complex and sialidase [13] . PUL5 consists of the five gpdCDGEF genes . GpdC , an homolog of the archetypal SusC [17] , likely represents the specific OM porin of the system . GpdD is an homolog of SusD , a starch-binding protein [16] , [20] and hence most likely a glycoprotein-binding protein . On the basis of their annotation , we propose that GpdE and GpdF are also glycan-binding proteins . GpdG was annotated as an endo-β-N-acetylglucosaminidase ( Manfredi et al . , submitted ) and this annotation was shown to be correct . Indeed mass spectrometry analyses demonstrated that GpdG deglycosylates the tribranched GlcNAc5Man3Gal3Sial3 glycan structure linked to N156 from the model glycoprotein fetuin , leaving one GlcNac residue on the protein . GpdDGEF were predicted to be lipoproteins ( Manfredi et al . , submitted ) . Replacement of the critical cysteine of the lipoprotein signal peptide from GpdD and GpdG completely abolished the deglycosylating activity , indicating that a periplasmic location did not sustain the activity . These data , together with the fact that the two proteins belong to the surface proteome indicate that these two lipoproteins are exposed to the surface and not to the periplasm . We assume the same is true for GpdE and F since , like GpdD , they are thought to bind glycans , they contain a lipobox and they belong to the surface proteome . Interestingly , all the five Gpd proteins could be co-purified with the porin-like GpdC , indicating that they all form one single complex at the bacterial surface . Unexpectedly , not only GpdD , -G , -E and -F co-purified with GpdC but also SiaC . Although SiaC was known to be part of the catabolic process , SiaC is not encoded together with GpdCDGEF ( Manfredi et al . , submitted ) and it was not anticipated that the interaction would be so close . SiaC turned out to be also a lipoprotein but , unlike GpdD and GpdG , it was still functional when it was directed to the periplasm , unlipidated . We inferred from this observation that , contrary to our initial report , SiaC is a periplasm-oriented lipoprotein . Thus , the observations presented here suggest the model illustrated in Fig . 9: The surface-exposed GpdCDEF complex captures the N-linked complex glycan moieties of glycoproteins , which are then detached from the protein by GpdG and internalized by GpdC . As soon as they reach the periplasm , SiaC removes the terminal sialic acid . This sequence of events is strongly supported by the observation that gpd mutant bacteria do not desialylate fetuin , although SiaC is functional in these mutants ( Fig . 1 ) . After desialylation , the oligosaccharide would be sequentially degraded by periplasmic exoglycosidases and the monosaccharides would be transferred to the cytosol . This last step of the model is supported by the fact that the genome encodes three putative β-galactosidases ( Ccan_01530 , Ccan_15520 , Ccan_17480 ) , five putative β-N-acetylhexosaminidases ( Ccan_03860 , Ccan_04040 , Ccan_16820 , Ccan_17870 , Ccan_20090 ) and four putative α-mannosidases ( Ccan_00510 , Ccan_01900 , Ccan_04050 and Ccan_16220 ) , all of them endowed with a signal peptide I or II , and none of them surface exposed ( Manfredi et al . , submitted ) . The β-galactosidase and α-mannosidase activities were confirmed in the crude extract ( data not shown ) . The three β-galactosidases seemed actually redundant since they could all be individually knocked out without affecting the growth on cells ( data not shown ) . This global model strikingly reminds the archetypal Sus system shown to consist of one single complex made of SusCDEF [16] . It is thought that SusG , an endo-acting enzyme , generates internal cuts in a bound starch molecule and releases oligosaccharides larger than maltotriose , which are then transported by SusC into the periplasmic compartment . In the periplasm , glycoside hydrolases SusA and SusB then degrade the oligosaccharides into their component sugars prior to final transport to the cytosol [27] , [28] . The two systems are thus remarkably conserved , although they adapted to different complex saccharides . To our knowledge , the Gpd system is the first Sus-like system devoted to foraging N-linked glycoproteins . It contributes to sustain growth of C . canimorsus at the expenses of cultured cells ( Manfredi et al . , submitted ) . Since C . canimorsus has 13 PULs ( Manfredi et al . , submitted ) , it is very likely that some of them could be devoted to the harvest of O-linked glycans , but this activity has not been identified thus far . The best approach would probably be to look for upregulation in the presence of O-linked glycoproteins , as was done in B . thetaiotaomicron [29] . Deglycosylation of N-linked glycans is not unprecedented among pathogens and commensals . As mentioned earlier , two streptococci , S . pyogenes and S . oralis have this remarkable property . In the case of S . pyogenes , this activity is exerted towards IgGs by secreted endoglycosidase EndoS and it does not seem to play a major role in nutrient acquisition [25] . In contrast , in S . oralis , the activity was shown to sustain growth [30] . It is interesting to notice that S . oralis , like C . canimorsus , is emerging as an important opportunistic pathogen originating from the oral flora . This commonality between two very different bacteria from the same ecosystem suggests first that the capacity to deglycosylate host proteins is a favourable trait in the mouth ecosystem and , second , could favour opportunistic infections . Deglycosylation of IgGs is very likely to contribute to a generalized infection as discussed by Collin and Olsen [25] but , for C . canimorsus , one cannot exclude that deglycosylation of other host proteins would also significantly contribute to pathogenesis . Our data demonstrate that PUL-encoded lipoproteins are surface-exposed . Prolipoproteins are exported through the Sec pathway and then acylated at the periplasmic leaflet of the inner membrane ( IM ) by the sequential action of glyceryl transferase , O-acyl transferase ( s ) and prolipoprotein signal peptidase ( signal peptidase II ) . A mature lipoprotein harbours as a first aminoacid a cysteine residue that is lipid modified with a N-Acyl diacyl Glyceryl group which serves to anchor the protein to the IM . In Gram-negative bacteria some lipoproteins are destined for the OM . These proteins are extracted from the IM , transported across the periplasm and inserted in the inner leaflet of the OM by the Lol pathway ( for review see refs [31] , [32] . Insertion of lipoproteins into the outer leaflet of the OM is however established in some pathogens like Borrelia but the pathway is neither well documented nor well understood [32] . Since bacteria from the Cytophaga-Flavobacteria-Bacteroides group massively insert lipoproteins in the outer leaflet of the OM , we postulate that they have an original system dedicated to the transport of lipoproteins across the OM but this system still needs to be identified and investigated .
Mutagenesis of Cc5 wt has been performed has described in ref [33] with slight modifications . Briefly , replacement cassettes with flanking regions spanning approximately 500 bp homologous to direct gpd framing regions were constructed with a three-fragment overlapping-PCR strategy . First , two PCRs were performed on 100 ng of of Cc5 genomic DNA with primers A and B ( Table 2 ) for the upstream flanking regions and with primers C and D for the downstream regions . Primers B and C contained an additional 5′ 20-nucleotide extension homologous to the resistance ermF insertion cassette . The ermF resistance cassette was amplified from plasmid pMM106 DNA with primers 5502 and 5503 . All three PCR products were cleaned and then mixed in equal amounts for PCR using Phusion polymerase ( Finnzymes ) . The initial denaturation was at 98°C for 2 min , followed by 12 cycles without primers to allow annealing and elongation of the overlapping fragments ( 98°C for 30 s , 50°C for 40 s , and 72°C for 2 min ) . After the addition of external primers ( A and D ) , the program was continued with 20 cycles ( 98°C for 30 s , 50°C for 40 s , and 72°C for 2 min 30 s ) and finally 10 min at 72°C . Final PCR products consisted in gpd::ermF insertion cassettes and were then digested with PstI and SpeI for cloning into the appropriate sites of the C . canimorsus suicide vector pMM25 . Resulting plasmids were transferred by RP4-mediated conjugative DNA transfer from E . coli S17-1 to C . canimorsus 5 to allow integration of the insertion cassette . Transconjugants were then selected for presence of the ermF cassette , checked for sensitivity to cefoxitin and the deleted regions were sequenced . Plasmid pPM1 , used for complementation and expression of the Gpd proteins , is a derivative of the E . coli- C . canimorsus shuttle vector pMM47A . 1 [33] . pMM47A . 1 ermF promoter region was cleaved with SalI and NcoI and the 117 nucteotides upstream the gpdC starting codon sequence , containing the putative gpdC promoter , was cloned using the same restriction sites . Full length gpdC , -D , -G , -E and -F were amplified with the specific primers listed in Table 3 and cloned into plasmid pPM1 into NcoI and XbaI restriction sites leading to the insertion of a glycine at position 2 . The E205G substitution inactivating the catalytic site of GpdG was introduced by site directed mutagenesis by overlapping PCR using primers 5008/6061 and 6060/6055 and cloned in pPM1 using NcoI and XbaI restriction sites leading to plasmid pFR10 ( gpdG* ) . The C17G substitution of GpdD was introduced by site directed mutagenesis amplifying by PCR using primers 6056 and 6057 and cloning NcoI/XbaI in pPM1 leading to plasmid pFR8 . The C21G substitution of GpdG was introduced by site directed mutagenesis amplifying by PCR using primers 6054 and 6055 and cloning NcoI/XbaI in pPM1 leading to plasmid pFR9 . The C17Y substitution of SiaC was introduced by site directed mutagenesis amplifying by inverse PCR using primers 5045 and 5046 using as pMM52 as template leading to plasmid pMM121 . 1 . C-terminal His-Strep double tagged gpdC was amplified by two-step overlapping PCR using primers 5081 , 5467 and 5530 and cloned in pMM47 . A using SalI and SpeI restriction sites leading to plasmid pPM3 . Bacteria were collected from blood agar plates and resuspended in PBS at OD600 = 1 . 100 µl of bacterial suspensions were then incubated with 100 µl of a fetal calf serum fetuin ( Sigma F2379 ) solution ( 0 . 1 g . l−1 ) for 120 minutes at 37°C . As negative control , 200 µl of 1∶2 diluted fetuin solution alone was incubated for 120 minutes at 37°C . Samples were then centrifuged for 5 min at 13000× g , supernatant collected and loaded in a 12% SDS gel . Samples were analyzed by immunoblotting ( Fetuin , Rabbit anti-Bovine RIA , UCBA699/R1H , ACCURATE CHEMICAL & SCIENTIFIC CORPORATION ) and lectin stainings were performed with Sambucus nigra lectin ( SNA ) according to manufacturer recommendations ( DIG Glycan Differentiation Kit , 11210238001 , Roche ) . Bacteria were collected from blood agar plates and resuspended in PBS at OD600 = 1 . 100 µl of bacterial suspensions were then incubated with 100 µl of a purified human IgG ( Invitrogen , 02-7102 ) solution ( 0 . 5 g . l−1 ) for 180 minutes at 37°C . As negative control , 200 µl of 1∶2 diluted IgG solution alone was incubated for 180 minutes at 37°C . Samples were then centrufiged for 5 min at 13000× g , supernatant collected and and loaded in a 12% SDS gel . Samples were analyzed by Coomassie blue staining , immunoblotting [Goat Anti-Human IgG ( Fc specific ) -FITC antibody , F9512 Sigma] and lectin stainings were performed with SNA according to manufacturer recommendations ( DIG Glycan Differentiation Kit , 11210238001 , Roche ) . Fetuin ( Sigma F2379 ) and human IgG ( Invitrogen , 02-7102 ) were reduced with 10 mM TCEP at 37°C for 1 hour and alkylated with 50 mM iodoacetamide for 15 min at room temperature . Fetuin and IgG were digested with trypsin at an enzyme to protein ratio of 1∶50 ( w/w ) at 37°C overnight . The peptides were desalted on C18 StageTips ( Thermo Fisher Scientific , Reinach , Switzerland ) according to the manufacurer's recommendations . The fetuin and IgG peptides were analysed on an LTQ Orbitrap instrument ( Thermo Fisher , San José , CA , USA ) coupled to an Agilent 1200 nano pump according to ( Manfredi et al . submitted ) . Bacteria were collected from blood agar plates and resuspended in 3 ml ice cold HEPES 10 mM ( pH 7 . 4 ) at OD600 = 1 . Bacterial suspensions were then sonicated on ice until they turned clear and spun at 15600× g for 2 minutes at 4°C . Supernatants were transferred and centrifuged again for 30 minutes at 15600× g at 4°C . Pellets were resuspended in 2 ml HEPES 10 mM with 1% sarcosyl ( N-Lauroylsarcosine sodium salt , Sigma ) and incubated at room temperature for 30 minutes . Finally , samples were centrifuged at 15600 g for 30 min at 4°C and pellet resuspended in 0 . 1 ml HEPES . Samples were checked for quality and quantity on silver stained SDS-PAGE and analysed by MS/MS . Cc5 ΔgpdC bacteria harbouring plasmid pPM3 , expressing a C-terminal His-Strep double tagged GpdC , or harbouring plasmid pPM2 , expressing GpdC without any tag ( Mock ) , were grown for 2 days at 37°C in the presence of 5% CO2 on sheep blood agar plates . Bacteria from 6 plates were scraped and lysed in 35 ml of 25 mM Tris-HCl , 150 mM NaCl , 0 . 2% triton , 1% NP-40% , 1% sodium deoxycholate , pH 7 . 6 . For His affinity purification , the lysates were clarified by centrifugation ( 10 min at 18500 g at RT ) and the supernatant was diluted 1∶2 in PBS , 10 mM Imidazole , in the presence of proteinase inhibitor ( cOmplete , Mini , EDTA-free Protease Inhibitor Cocktail Tablets , Roche ) . 3 . 5 ml of 50% slurry Chelating sepharose Fast Flow beads ( GE Healthcare ) was first coupled to Ni2+ according to the manufacturer instructions and then 1 . 75 ml of resin was added to the solution and incubated overnight at 4°C on a rotating wheel . The solution was then loaded into a column and the resin washed first with 25 column volumes ( CV ) of high salt buffer ( 50 mM Tris , 500 mM NaCl , pH 8 ) and then with 5 CV of low salt buffer ( 50 mM Tris , 100 mM NaCl , pH 8 ) . Proteins were then eluted from the resin with 2 CV of elution buffer ( 50 mM Tris , 100 mM NaCl , 350 mM Imidazole , pH 8 ) . The material eluted from the Ni2+ column was then diluted 1∶2 in PBS and 1 ml of 50% slurry ( 0 . 5 ml CV ) Strep-Tactin Superflow resin ( IBA , cat No: 2-1206-002 ) was added . The solution was then incubated overnight at 4°C on a rotating wheel . The solution was then loaded into a column and the flow through reloaded into the resin 2 more times . The resin was then washed 4 times with 10 CV of Buffer W ( 100 mM Tris , 150 mM NaCl , 1 mM EDTA , pH 8 ) and proteins eluted in 3 steps with 0 . 5 ml elution buffer ( 100 mM Tris , 150 mM NaCl , 1 mM EDTA , 2 . 5 mM desthiobiotin , pH 8 ) . The proteins present in the elution fractions were identified by MS and immunoblotting , using anti-His for GpdC detection , anti-GpdG and anti-SiaC . GpdC-sialidase co-purification was performed exactly as described above using Cc5 ΔPUL5 bacteria harbouring pPM3 plasmid or harbouring plasmid pPM2 ( Mock ) . Proteins present in the elution fractions were identified by immunoblotting with anti-Strep antibodies to detect GpdC and anti-SiaC . Bacteria were inoculated to HeLa epithelial cells ( ATCC CCL-2 ) in complete DMEM at 37°C with 5% CO2 at a moi of 20 . 15–16 h post infection , [9 , 10-3H] palmitic acid ( 48 Ci/mmol; Perkin-Elmer Life Sciences ) was added to a final concentration of 50 µCi/ml and incubation was continued for 8–9 h , by which time the bacterial culture had reached approximately 108 bacteria/ml as described elsewhere [13] . Supernatants of 2×1 ml were collected without detaching epithelial cells from the wells . Bacteria corresponding to approximately 2×108 cfu were then collected by centrifugation and pellets were combined from 2 ml and stored at −20°C until they were processed . Pellets were resuspended in 0 . 1 ml PBS TritonX 1% to lyze bacteria and sialidase was immuno-precipitated by addition of 10 µl rabbit polyclonal anti-SiaC for 1 h at RT on a rotating wheel . Protein A agarose slurry ( Sigma ) was then added in equal amounts for 30 min under constant rotation at RT . Samples were then centrifuged at 14000× g for 2 min at RT , supernatant was discarded and pellets were washed with 0 . 5 ml PBS 0 . 1% Triton which was repeated 4 times . Captured proteins were eluted by addition of 50 µl Lämmli buffer ( 1% SDS , 10% glycerol , 50 mM dithiothreitol , 0 . 02% bromophenol blue , 45 mM Tris , pH 6 . 8 ) for 5 min at 85°C . Samples were centrifuged again and supernatant was carefully separated from the agarose beads and loaded on SDS PAGE gels using 10% polyacrylamide . After gel electrophoresis , gels were fixed in 25∶65∶10 isopropanol∶water∶acetic acid overnight and subsequently soaked for 30 min in Amplify ( Amersham ) . Gels were vacuum dried and exposed to SuperRX autoradiography film ( Fuji ) for 13days until desired signal strength was reached .
|
Capnocytophaga canimorsus are Gram-negative bacteria from the normal oral flora of dogs and cats . They cause rare but severe infections in humans that have been bitten or simply licked by a dog or cat . Fulminant septicemia and peripheral gangrene with a high mortality are the most common symptoms . A surprising feature of these bacteria is their capacity to feed by foraging the glycan moieties of glycoproteins from animal cells , including phagocytes . Here we show that C . canimorsus can also deglycosylate human IgGs reinforcing the idea that this property of harvesting host glycoproteins may contribute to pathogenesis . We also unravel the complete deglycosylation system which belongs to a large family of systems devoted to foraging complex glycans , found exclusively in the Capnocytophaga-Flavobacteria-Bacteroides group , and whose archetype is the starch harvesting system Sus . It is the first system devoted to deglycosylation of glycoproteins to be characterized .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"gram",
"negative",
"microbial",
"pathogens",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"bacterial",
"pathogens",
"pathogenesis"
] |
2011
|
The N-glycan Glycoprotein Deglycosylation Complex (Gpd) from Capnocytophaga canimorsus Deglycosylates Human IgG
|
Intestinal microfold ( M ) cells possess a high transcytosis capacity and are able to transport a broad range of materials including particulate antigens , soluble macromolecules , and pathogens from the intestinal lumen to inductive sites of the mucosal immune system . M cells are also the primary pathway for delivery of secretory IgA ( SIgA ) to the gut-associated lymphoid tissue . However , although the consequences of SIgA uptake by M cells are now well known and described , the mechanisms whereby SIgA is selectively bound and taken up remain poorly understood . Here we first demonstrate that both the Cα1 region and glycosylation , more particularly sialic acid residues , are involved in M cell–mediated reverse transcytosis . Second , we found that SIgA is taken up by M cells via the Dectin-1 receptor , with the possible involvement of Siglec-5 acting as a co-receptor . Third , we establish that transcytosed SIgA is taken up by mucosal CX3CR1+ dendritic cells ( DCs ) via the DC-SIGN receptor . Fourth , we show that mucosal and systemic antibody responses against the HIV p24-SIgA complexes administered orally is strictly dependent on the expression of Dectin-1 . Having deciphered the mechanisms leading to specific targeting of SIgA-based Ag complexes paves the way to the use of such a vehicle for mucosal vaccination against various infectious diseases .
The mucosal immune system comprises the largest part of the entire immune system , and the mucosal surface represents the primary site of entry for pathogenic agents . SIgA has long been recognized as a first line of defense in protecting the intestinal epithelium from enteric pathogens and toxins . It is generally assumed that SIgA acts primarily through receptor blockade , steric hindrance , and/or immune exclusion . In recent years evidence has emerged indicating that SIgA promotes the uptake and delivery of Ags from the intestinal lumen to DC subsets located in gut-associated lymphoid tissues ( GALTs ) , and influences inflammatory responses normally associated with the uptake of highly pathogenic bacteria and potentially allergenic antigens . This particular feature of SIgA , called reverse transcytosis , is mediated by epithelial M cells [1] . However , although the potentially useful properties of M cells on SIgA uptake are now well known , the receptor ( s ) whereby SIgA is taken up and transported by M cells remain ( s ) elusive . SIgA reverse transcytosis was first invoked to account for the binding of rabbit SIgA to M cells in Peyer's patches ( PPs ) of suckling rabbits [2] . Colloidal gold particles coated with IgA were subsequently detected within M cell cytoplasmic vesicles and in the extracellular space of M cell pockets [3] . Endogenous SIgA was also shown to bind to human PP M cells in paraffin sections of human ileum [4] . In frozen sections , labeled SIgA could be visualized bound at the apical surface , in transit through intracellular vesicles , in the intraepithelial pocket , and on basolateral processes extending toward the basal lamina . In a mouse ligated ileal loop assay , mouse SIgA , human SIgA2 , but not human SIgA1 , bound to PP M cells [4] . Structural changes could explain the differences in reverse transcytosis between these subtypes . The IgA1 hinge features a 16 amino-acid insertion , lacking in IgA2 , comprising a repeat of eight amino acids decorated with 3–5 O-linked oligosaccharides [5] , [6] . Recombinant IgA1 with a deleted hinge region gained M cell binding function , which was interpreted as the M cell's binding site comprising both domains Cα1 and Cα2 , juxtaposed in mouse IgA and human IgA2 [4] . Overall , IgA2 contains 4 N-glycosylation sites ( Asn166 , Asn263 , Asn337 , Asn459 ) . In dimeric IgA , the Fc regions of the two monomers are linked end to end through disulfide bridges to the J chain [7] . IgA , with or without bound secretory component ( SC ) , selectively adheres to the apical surfaces of mouse PP M cells [4] . To date , only a limited number of M cell receptors and their ligands have been identified , but most of these receptors are expressed in M cells and neighboring enterocytes as well . Some important pathogen recognition receptors , such as toll-like receptor-4 , platelet-activating factor receptor , and α5β1 integrin have been identified on the surface of human and mouse M cells [8] , [9] . The sialyl Lewis A ( CA19 . 9 ) antigen lectin reacts with 80% of human M cells and , in contrast to the other ligands , binds only weakly to the enterocytes of the follicle-associated epithelium ( FAE ) . Moreover , there is a wide variation in marker expression between M cells of different species and even between M cells at different portions of the intestine within the same species [10] . Indeed , M cells in murine , but not human , PP are preferentially bound with Ulex europaeus agglutinin–1 ( UEA-1 ) , a lectin specific to α-l-fucose residues [11] . A first mouse M cell–specific monoclonal antibody ( mAb NKM 16-2-4 ) [12] displaying specificity for α ( 1 , 2 ) -fucose–containing carbohydrate moieties was produced . Glycoprotein 2 ( GP2 ) was also shown to be specifically expressed on M cells of mouse and human PPs [13]–[15] and serves as an endocytic receptor for luminal antigens [16] . Another M cell marker , clusterin , is expressed in M cells and follicular DCs at inductive sites of human GALTs [14] . To date , the molecular partner ( s ) involved in SIgA reverse transcytosis has ( have ) not been identified in mice or in humans . In this work , we sought to map the structural feature ( s ) responsible for the selective interaction between murine SIgA and M cells . Since it is impossible to keep M cells in culture , one valuable approach consists in using cell culture models that mimic essential features of the FAE tissue . An in vitro model was used , based on the co-culture of polarized Caco-2 cells grown on inverted inserts and exposed to human Raji B lymphocytes [17] , [18] . Following optimization in terms of functionality and reproducibility , we evaluated the transport of wild-type ( wt ) and mutant human IgA2 across newly differentiated M-like cells in comparison with other Ab isotypes . We found that glycosylation sites and in particular sialylation of the Cα1 region of IgA2 are required for M-like cell-mediated reverse transcytosis . We demonstrate for the first time that Dectin-1 expressed on the surface of M cells acts as a receptor involved in SIgA reverse transcytosis both in vitro and in vivo . Siglec-5 receptor seems also to participate in reverse transcytosis . Such a selective interaction has functional consequences in vivo , since targeting of HIV p24-SIgA complexes after oral delivery promotes the production of systemic and mucosal Ag-specific Abs in wt mice only , and not in Dectin-1 KO animals .
The model was adapted as described in the Methods section to optimize its reproducibility ( Figure 1a ) . Prior to adding the lymphocytes , the tightness of the Caco-2 cell monolayer was checked by measuring transepithelial electrical resistance ( TEER ) . The decrease in TEER observed after 5 d of co-culture is indicative of Caco-2 cell conversion into M cells ( Figure 1b ) [19] but not a result of the deterioration of tight junction organization , as reflected by preserved ZO-1 immunolabeling ( Figure 1c ) , of either mono- or co-cultures . M cells display a reduced brush border at their apical surface and an invaginated basolateral membrane , forming a pocket filled with immunoreactive cells [3] . Transmission electron microscopy shows that mono-cultures of Caco-2 cells exhibit a well-developed brush border with tightly packed microvilli , whereas in co-cultures with Raji cells , M-like cells characterized by the effacement of microvilli and enfolded lymphocytes are present ( Figure 1d1 ) . Moreover , the presence of desmosomes between M-like cells and the neighboring cells reveals their enterocytic origin ( Figure 1d1 , inset ) . Using scanning electron microscopy analysis , we observed in mono-cultures that all Caco-2 cells possessed a regular brush border and well-developed tight junctions , whereas in co-cultures , approximately 20–30% of Caco-2 cells expressed short and irregular microvilli ( Figure 1d2 ) [20] . Immunolabeling of M-like cells with CA19 . 9 and enterocytes with UEA-1 ( Ulex europaeus isoagglutinin I ) [21] indicated a similar percentage of conversion , assuming a surface equivalence for M and Caco-2 cells ( Figure 1e1 ) . This was further verified by co-localization of human IgA2 with M cells labeled with CA19 . 9 mAb ( Figure 1e2 ) , in agreement with Mantis et al . [4] . To verify functional Caco-2 cell conversion into M cells , the transport of yellow/green-conjugated , 0 . 2 µm nanoparticles ( NPs ) across mono- and co-cultures was examined . NPs have previously been used to study transcytosis in various M-like cell models in vitro and in vivo [22] . The number of transported NPs recovered in the basal medium was 5 . 5-fold higher in the co-cultures , compared to Caco-2 cell mono-cultures ( p<0 . 001 ) ( Figure 1f ) . The sum of these data confirmed that the in vitro model of human FAE allowed efficient Caco-2 cells to M-like cell conversion to occur ( 20–30% ) , and importantly , with a high level of reproducibility . Wt and truncated/mutated Ab constructs depicted in Figure 2 were cloned in the pGTRIO expression vector , stably transfected in CHO cells , produced in the culture supernatant , and purified by affinity chromatography as described in the Methods section . SDS-PAGE performed under reducing and nonreducing conditions confirmed the expected molecular weight for the light and heavy chains of the various constructs produced , and indicated assembly despite reduced formation of disulfide bridges between heavy and light chains ( Figure 2 ) , a feature commonly encountered while expressing IgA Abs in CHO cells [23] . One feature of M cells is their ability to transport a broad range of materials including Abs from the lumen to the underlying follicles . Specific retro-transport of Abs was compared between mono- and co-cultures using a luciferase ( Luc ) -IgA fusion protein . The Luc tag did not affect the Ab functionality ( unpublished data ) and allowed for sensitive quantification . As shown in Figure 3a , a significant exclusive transport of the IgA2 monomer ( m-IgA2 ) across the cell monolayer harboring M-like cells was observed ( p = 0 . 03 ) . No significant transport of m-IgA1 , IgG , or IgE was detected . Specificity of IgA reverse transcytosis was further confirmed in vivo by using a ligated murine intestinal loop . IgA positive cells were 30 times more abundant than IgG positive cells in PPs ( Figure 3b ) . Dimerization , by incorporation of the J chain , or association with human SC did not modify IgA2 uptake by M-like cells ( Figure 3a ) . Next , mapping of regions and domains involved in IgA2 reverse transcytosis was performed with recombinant IgA2 lacking various portions of the heavy chain C-terminus ( Figure 3a ) . IgA2 monomer depleted of Cα2 , Cα3 , and the tailpiece ( m-IgA2 dCα2/3 ) crossed M-like cells as well as m-IgA2 wt , whereas IgA2 monomer depleted of the tailpiece ( m-IgA2 dPB ) and IgA2 monomer depleted of both the Cα3 and tailpiece ( m-IgA2 dCα3 ) were not transported . The hinge region did not influence uptake as m-IgA2 dCα2/3 and m-IgA2 Cα1 ( IgA2 with only the Cα1 constant region ) gave similar results . Strikingly , M-like cell-mediated transport of IgA2 with only the Cα1 constant region was equivalent to wt m-IgA2 . These results demonstrate that in the in vitro model , the Cα1 region of IgA2 is sufficient to allow reverse transcytosis through M-like cells . Subclasses of human IgA are also different with respect to the number of N-glycosylation sites . In order to determine whether N-glycans present on IgA2 could influence their uptake and transport by M-like cells , transcytosis of a battery of constructs with engineered glycosylation sites was compared . As shown in Figure 4a , the efficiency of reverse transcytosis was highly dependent on the number of glycosylation sites . Indeed , there was a significant decrease in the transport of m-IgA2 G2 , G1 , G0 , and m-IgA2 Cα1 G0 compared with m-IgA2 . These findings were confirmed by enzymatic digestion of m-IgA2 by PNGase , an amidase that cleaves between the innermost GlcNAc and asparagine residues of high mannose , hybrid , and complex oligosaccharides from N-linked glycoproteins . Sialic acid ( Sia ) can occur in different glycosidic linkages , most typically at the exposed , nonreduced ends of oligosaccharide chains attached to a wide variety of proteins like IgA [24] . To assess the function of Sia in IgA binding to M-like cells , IgA was exposed to neuraminidase , which has the capacity to selectively cleave the glycosidic linkages of neuraminic acids . An important and significant decrease in transport of IgA2 lacking Sia , resembling that measured for m-IgA2 G0 or m-IgA2+PNGase , was observed . Identical results were obtained with another recombinant IgA2 Ab molecule specific for CD20 , with SIgA purified from colostrum and with plasma IgA treated by neuraminidase or PNGase ( Figure 4b ) . The absence of remaining carbohydrates or Sia on the different IgA was verified by Western blot using labeling with lectins ( Figure 4c ) , while the integrity of the IgA2 polypeptide following enzymatic treatment was verified by analysis on SDS-PAA gels ( unpublished data ) . These results demonstrate the essential role of IgA glycosylation sites , and in particular , Sia in the reverse transcytosis of IgA2 by M-like cells . As the above results provide solid evidence of the contribution of glycosylation to reverse transcytosis , we postulated that the IgA2 receptor of M-like cells is a glucan receptor . Blocking experiments were performed using a series of β-glucans , mono- , and disaccharides . A statistically significant decrease in IgA2 transport was observed in the presence of β-glucans including curdlan , laminarin , and zymosan ( Figure 5a ) . No inhibition was observed with other members of the family or with mono- or disaccharides . To further explore the possible involvement of glycans in IgA2 binding to M-like cells , Abs directed against the most common sugar receptors were used in blocking experiments . The use of an anti-Dectin-1 mAb targeting this β-glucan receptor led to an almost complete inhibition of IgA2 reverse transcytosis ( Figure 5b ) . In contrast , blocking of the mannose receptor with an anti-CD206 mAb or of the lipopolysaccharide receptor with anti-TLR4 and anti-CD14 mAbs did not influence IgA2 transport . Consistently , the presence of Dectin-1 was observed on M-like cells present in co-culture conditions only ( Figure 5c–e ) . Other receptors that have been described as being involved in IgA transport were evaluated . Transferrin receptor expressed by enterocytes ( CD71 ) [25] , which binds IgA1 Abs , did not block IgA2 passage , thus confirming the exclusive transport of IgA2 by M-like cells . Similarly , targeting of the human myeloid IgA Fc receptor ( CD89 ) [26] with a specific mAb did not block the transport of IgA2 ( Figure 5b ) . To confirm the inability of desialylated IgA2 to target M-like cells in vitro , blocking experiments were also carried out using mAbs directed against various Siglecs , a family of receptors that specifically recognize Sia [27] . The unique involvement of Siglec-5 in IgA2 reverse transcytosis was demonstrated ( Figure 5b ) , in contrast to all the other members of the family . The surface of M-like cells was Siglec-5+ in co-culture conditions only , with no labeling observed in mono-cultures ( Figure 5c–e ) . Moreover , N-Acetylneuraminic acid severely affected SIgA2 reverse transcytosis ( Figure 5a ) . Mabs to either Dectin-1 or Siglec-5 strongly inhibited transport of IgA2 in vitro , reaching up to 90% when added together ( Figure 5b ) . Binding of monomeric IgA2 and SIgA to Dectin-1 and Siglec-5 was verified by ELISA using recombinant Dectin-1 and Siglec-5 as coating molecules ( Figure 6a ) . In support of previous data , deglycosylated IgA2 and IgA1 were unable to recognize Dectin-1 and Siglec-5 . Specificity of IgA2 recognition was further confirmed by immunofluorescence ( Figure 6b ) and flow cytometry ( Figure 6c ) using HEK cell transfectants expressing both Dectin-1 and Siglec-5 . Importantly , co-localization between partners of the triad was observed in both types of analyses . Taken together , these results highlight the prominent role of Dectin-1 and Siglec-5 as receptors that mediate intestinal IgA2 reverse transcytosis . To verify the validity of data obtained using the in vitro model of human FAE , SIgA transport was also analyzed in vivo in a mouse ligated intestinal loop containing a PP [28] . As shown in Figure 7a1 and 7a2 , mouse SIgA-Cy3 was present on the surface of , and inside , UEA-1+ or GP2+ M cells , thus confirming the in vitro binding data in the in vivo context . Co-localization of mouse SIgA on Dectin-1+ cells in the FAE confirmed the role of Dectin-1 in SIgA binding in vivo as well ( Figure 7b ) . In support of these data , in a Dectin-1 KO mouse model , we observed no co-localization between SIgA-Cy3 and UEA-1+ M cells and no reverse transcytosis of SIgA-Cy3 in PPs ( Figure 7c ) . The interaction between Dectin-1 or Siglec-5 with IgA2 was similarly observed in human PPs . Immunolabeling with green-labeled IgA2 and red-labeled Dectin-1 or Siglec-5 of patient biopsies displayed specific co-localization between the Ab and Dectin-1 ( Figure 7d ) or Siglec-5 ( Figure 7e ) . No specific immunofluorescence of secondary IgG Abs was obtained on human M cells ( Figure 7f ) . These results prompted us to compare the outcome of oral immunization in wt C57BL/6 mice and Dectin-1 KO mice using SIgA as an intestinal delivery system targeting M cells . As Dectin-1 is also expressed by DCs or macrophages , one can argue that such cells intercalating within the FAE may “pollute” the Dectin-1 signal on M cells in vivo . To solve this issue , Dectin-1 KO mice reconstituted with wt bone marrow cells ( chimeric-KO:wt ) and wt mice reconstituted with Dectin-1 KO bone marrow cells ( chimeric-wt:KO ) were immunized . Confirmation of the correct reconstitution in the chimeric mice was obtained by flow cytometry ( Figure 8a/b ) and immunofluorescence ( Figure 8c ) analysis on peripheral blood leukocytes . Positive control of immunization was obtained by subcutaneous administration of nanoparticulated vaccine polylactic acid ( PLA ) -p24 , which induced strong immune response in mice ( Figure 8d/e ) [29] . HIVp24 was chosen as a vaccine candidate antigen for its relatively low molecular weight , thus reducing the risk of disturbing the overall structure of SIgA after covalent coupling . Administration of p24-SIgA in an intestinal ligated loop resulted in the presence of the complex in the SED region of PPs ( unpublished data ) . Moreover , p24-SIgA complexes administered orally co-localized with Dectin-1+ cells in the FAE region ( Figure 9a ) . Oral immunizations with p24-SIgA were performed in wt , Dectin-1 KO , chimeric-wt:KO , and chimeric-KO:wt mice as described in the Methods section . As intestinal immunization is well known to induce both mucosal and systemic responses [30] , serum and feces samples were collected 1 wk after the last immunization . p24-specific IgG and IgA titers were measured following immunization of wt and chimeric-wt:KO mice only ( Figure 9b and 9c ) . Moreover , the levels of p24-specific IgG and IgA responses in these mice were 25-fold higher than those obtained after oral immunization with the p24 polypeptide only . No antigen-specific response was measured in Dectin-1 KO and chimeric-KO:wt mice , thus confirming the essential role of Dectin-1 in SIgA reverse transcytosis . Taken together , these results indicate that reverse transcytosis of the p24-SIgA complex is strictly Dectin-1-dependent and results in the potentiated passage of the hooked Ag , which is subsequently processed to trigger the onset of mucosal and systemic Ab responses . In order to examine SIgA2 transport from the intestinal lumen to DCs located in the SED region of PPs , we took advantage of the recent demonstration that SIgA is recognized by DCs via the DC-SIGN receptor [31] . SIgA uptake by DCs was analyzed in vivo in a PP-containing ligated intestinal loop from wt mice with DC-SIGN-specific immunostaining and from CX3CR1-GFP transgenic mice . Figure 10a and 10b show specific localizations of p24-SIgA on DC-SIGN+ and SIgA-Cy3 on CX3CR1-GFP+ DCs present in the SED region . However , the strictly equivalent of DC-SIGN has not been described in mice , yet several homologues have been documented [32] . We assume that DC-SIGN-positive staining results from cross-reactivity with one of these murine homologues . We next assess the relevance of these findings in the human in vitro system . HeLa transfectants stably expressing DC-SIGN added to the compartment bathing the basolateral pole of Caco-2 cells were used as surrogates of DCs populating the SED region of PPs [33] . Control of DC-SIGN expression was demonstrated by the inhibition of gp120 binding on HeLa-DC-SIGN+ cells by specific blocking mAbs ( unpublished data ) . The binding of IgA2 that had previously crossed the monolayer containing M-like cells was observed by immunostaining of HeLa-DC-SIGN ( Figure 10c ) , but not with wt HeLa cells used as a negative control . In another control , CHO expressing Langerin placed in the basolateral compartment did not bind transcytosed IgA2 ( unpublished data ) . The transport of integral IgA2 through M cells , and also their preserved capacity to interact with DC-SIGN+ DC–expressing cells is another indication of the steps involved to ultimately lead to immune responses as detected above . These findings have also been confirmed by flow cytometry using human monocyte-derived DCs known to express DC-SIGN ( Figure 10d ) . The sum of these data shed light on the biochemical partners involved in reverse transcytosis of SIgA by PPs . SIgA is first taken up by M cells via the Dectin-1 receptor and/or Siglec-5 , and is subsequently targeted to mucosal CX3CR1+ DCs bearing the DC-SIGN receptor . In the context of immune complexes , this process explains the functional production of mucosal and systemic Ab responses to the associated antigen .
M cells possess a high transcytotic capacity , allowing a wide range of materials to be transported including particulate Ags , soluble macromolecules , and pathogens . They are delivered from the intestinal lumen to inductive sites of the mucosal immune system . M cells are also the primary route through which SIgA are delivered to the GALT . Corthésy et al . have previously shown that after selective interaction with M cells , SIgA are targeted to DCs located in the SED region of PP , resulting in limited mucosal and systemic immune responses against a non-self-associated protein Ag [34] . Selective adherence to the apical surface of M cells is a prerequisite for efficient transepithelial transport , but the identity of receptors involved in SIgA endocytosis has remained elusive . In the current study , we investigated the transport of human SIgA2 across a model mimicking human FAE . At the level of IgA2 , we provide evidence that both the Cα1 domain and associated glycosylation , more particularly Sia residues , are involved in M-like cell-mediated reverse transcytosis , while at the receptor level , both Dectin-1 and Siglec-5 have been identified as essential partner in the process . Finally , we validate our in vitro results upon analysis of murine and human tissues , ultimately demonstrating that Dectin-1/Siglec-5-mediated uptake of SIgA-based complexes results in productive mucosal and systemic antigen-specific Ab responses . Initially , we studied reverse transcytosis of IgA2 across human M-like cells using a cell culture model that reproduces features of the FAE tissue . We confirmed that human IgA2 , with or without J chain and/or bound SC , but not IgA1 , IgG , or IgE , selectively bound to the apical surface of in vitro differentiated human M-like cells . Using a battery of deletion mutants , we demonstrated that domains Cα2 and Cα3 of IgA2 are dispensable to keep reverse transcytosis through M-like cells highly active . Low or absent transport of m-IgA2 dCα3 and m-IgA2 dPB comprising the Cα1 region suggests that subtle structural changes may affect optimal folding of these two particular recombinant proteins . Our in vitro results obtained with human cells do not totally correlate with the in vivo results of Mantis et al . , who showed that both domains Cα1 and Cα2 were required for IgA binding to mouse PP M cells [4] . Differences in the expression systems for IgA constructs ( deletion versus domain swab ) and the glycosylation pattern may explain this discrepancy . It is conceivable that a critical density of glycans must be present to ensure uptake , as was recently described for Dectin-1 efficiently binding β-glucan polymers [35] . In conclusion , our study unequivocally demonstrates that IgA transport requires the presence of the properly glycosylated Cα1 domain within the Ab structure . The model opens the path toward in vitro assays of transport across reconstituted FAE , examination of the mechanisms of uptake , and investigation into vaccine or intestinal microbe delivery . Sia residues on pathogens interact with Siglecs , which are expressed in the hemopoietic , immune , and nervous systems . Glycosylation patterns on pathogens are frequently used for adherence to , and passage across , the mucosal epithelium and in particular M cells in the FAE [36]–[38] . Similarly , it is conceivable that abundant carbohydrates located on the surface of SIgA may intervene in the process of selective recognition of M cells . The sum of our data confirms this working hypothesis , and demonstrates the prominent influence of glycosylation on the uptake of IgA2 by M cells . Additional experiments dealing with deletion of particular glycosylation sites and enzymatic desialylation allowed us to confirm the role of Sia residues in reverse transcytosis ( Figure 4 ) . Several members of the β-glucan superfamily were also identified as competitors of IgA2 transcytosis ( Figure 5 ) . This adds to the multiple functions of carbohydrates in SIgA including , for example , neutralization of bacterial toxins [39] and interaction with commensal bacteria [40] . Having unraveled the structural features responsible for the selective transport of SIgA in the reconstituted FAE model , we sought to identity the receptor ( s ) by which SIgA is taken up and transported by M cells . The use of blocking Abs against known IgA receptors including CD89 and CD71 did not prevent SIgA2 reverse transcytosis . These data , combined with the sufficient role of the Cα1 region of IgA2 in M-cell-mediated reverse transcytosis , led to the conclusion that no other known IgA receptor ( pIgR , Fcα/μ receptor , and the asialoglycoprotein receptor ) was involved in the process . Given the established involvement of Sia and β-glucan moieties , we speculated that the IgA2 receptor of M cells is a glucan- and/or Sia-receptor . Our work provides evidence of the presence of Dectin-1 on M-like cells , together with its involvement in reverse transcytosis of SIgA2 . Dectin-1 is a type II transmembrane protein of the C-type lectin family , expressed by myeloid phagocytes ( macrophages , DCs and neutrophils ) , which recognizes β-glucans in fungal cell walls and transduces signals triggering phagocytosis and the production of reactive oxygen species [41] , [42] . In contrast , as recognition of soluble ligands by Dectin-1 does not lead to inappropriate activation signaling [35] , its presence on M cells is consistent with simple SIgA capture and internalization . Co-operation between Fc galactosylation and Dectin-1–inducing anti-inflammatory activities suggests that Dectin-1 is capable of working in combination with other partners in the cell plasma membrane . In view of the involvement of Sia in IgA2 reverse transcytosis via M cells , we investigated whether a Siglec receptor could serve this function . The majority of Siglecs , including CD33-related Siglecs like Siglec-5 , appears to be naturally masked owing to cis-interactions with adjacent Sia . Unmasking of Siglecs can also occur in some cases by cellular activation or by exposure to sialidases . The unmasked Siglec would then be capable of de novo interactions with surrounding ligands in the environment . This could result in increased interactions with exogenous materials including glycosylated SIgA . Such a scenario of Siglec serving as a co-receptor has been reported in the case of HIV-1 entry mediated by CD4 in macrophages . Preparation of murine duodenal ligated loops validated the results generated in the in vitro model of human FAE . This method has proven valuable in documenting the interaction of mouse IgA with PP M cells [4] . Tissue immunolabeling both confirmed the transport of SIgA2 by UEA-1+ and GP2+ M cells , and that of murine SIgA by Dectin-1 ( Figure 7a and b ) . However , the absence of cross-reactivity of the anti-human CD170 mAb prevented us from confirming the role of Siglec-5 in the reverse transcytosis of SIgA in mice . Consistent with the in vitro data gathered in the model based on human cells , human biopsy analyses resulted in specific co-localization between IgA2-GFP and Dectin-1 ( Figure 7d ) or Siglec-5 ( Figure 7e ) . Finally , oral immunization of wt , Dectin-1 KO , or chimeric mice with p24-SIgA complexes unambiguously demonstrated that reverse transcytosis of SIgA is strictly dependent on Dectin-1 expressed on M cells . The further confirmation of the essential role of Dectin-1 in the in vivo context provides an explanation to the uptake of antigen-bearing SIgA by M cells , a feature resulting in systemic and mucosal immune responses [43] . The lack of a murine functional ortholog of human Siglec-5 prevented us from confirming the associated role of Siglec-5 in SIgA reverse transcytosis in vivo [44]–[46] . In vivo , the uptake of murine SIgA by murine CX3CR1+ DCs present in the GALT could also be documented ( Figure 10b ) . In the SED region , CX3CR1+ DCs play a central role in antigen sampling [47] . In contrast to CD103+ DCs , CX3CR1+ cells represent a nonmigratory gut-resident population , which displays poor T-cell stimulatory capacity [48] , [49] . In contrast to CD103+ DCs that serve classical DC functions and initiate adaptive immune responses in local lymph nodes , CX3CR1+ populations might modulate immune responses directly in the mucosa and serve as a first line barrier against invading enteropathogens . This supports the low activation properties of SIgA targeting antigen to DCs in the SED region [50] , [51] . A recent study has shown that small intestine goblet cells function as passages delivering the low molecular weight soluble dextran ( 10 kDa ) to CD103+ DCs [52] , which promote IgA production , imprint gut homing on lymphocytes , and induce the development of regulatory T cells . As HIVp24 is administered in the form of a complex with SIgA ( 400 kDa ) , we believe that this pathway need additional characterization before it can be considered as operative for large molecules . Transcytosis across M cells is known to enable the selective transport of particulate antigens in the absence of any assessable damage [53] , [54] . This holds true for soluble SIgA , as the transcytosed Ab released by M cells in the human in vitro and murine in vivo models was still able to specifically target cells expressing DC-SIGN in the basolateral environment ( Figure 10a/c/d ) . In mucosal tissues such as the rectum , uterus , and cervix , DC-SIGN is abundantly expressed by DCs present in the lamina propria and PPs , further substantiating the importance of the localization of DC-SIGN+ DCs as a first line of defense against viruses and pathogens . Delivery in the form of SIgA-based immune complexes may thus combine the onset of limited immune responses , which translates into the absence of spurious inflammatory reactions . Moreover , this receptor , by binding to ICAM-3 , favors the generation of antigen-specific suppressive CD4+ T cells , which produce IL-10 [55] , a cytokine that intervenes in both intestinal homeostasis and the production of local IgA . This work defines Dectin-1 expressed on the surface of M cells as a receptor involved in SIgA reverse transcytosis both in vitro and in vivo . Besides bringing new information on the mechanism involved in SIgA retro-transport , deciphering the identity of such receptors may lead to the further development of mucosal vaccines targeting M cells . In future work , it will be critical to test the expression of Dectin-1 on other mucous membranes such as nasal/bronchial , endocervical , or buccal mucosa order to evaluate the broad applicability of this finding to active and passive immunization . As a perspective to future works , one can argue that intestinal villous M cells serving as an antigen gateway for the sampling of gut bacteria and inducing Ag-specific immune responses in a PP-independent manner [56] may contribute to SIgA reverse transcytosis as well .
Pullulan from Aureobasidium pullulans , mannan from Saccharomyces cerevisiae , α-Lactose , L-fructose , glycogen from bovine liver , sucrose , curdlan from Alcaligenes faecalis , laminarin from Laminaria digitata , and zymosan from Saccharomyces cerevisiae were all purchased from Sigma-Aldrich . Anti-human Dectin-1/CLEC 7A polyclonal Ab ( pAb ) ( goat IgG ) , anti-human CD14 mAb ( mouse IgG1 ) , anti-human TLR4 pAb ( goat IgG ) , anti-human CD170 mAb ( Siglec-5 ) ( mouse IgG1 ) , and anti-human CD329 mAb ( Siglec-9 ) ( mouse IgG2a ) were all purchased from R&D Systems . Anti-human CD206 mAb ( mouse IgG1 ) ( mannose receptor ) was purchased from Ozyme . Anti-human CD22 mAb ( Siglec-2 ) ( mouse IgG1 ) , anti-human CD33 mAb ( Siglec-3 ) ( mouse IgG1 ) , anti-human CDw328 mAb ( Siglec-7 ) ( mouse IgG1 ) , and anti-human CD169 mAb ( Siglec-1 ) ( mouse IgG1 ) were purchased from AbD Serotec . Anti-human CD71 mAb ( mouse IgG1 ) was purchased from Cliniscience . Anti-human CD89 mAb ( mouse IgG1 ) was purchased from Abcam . All Abs were blocking and used according to the procedure provided by the manufacturer . Yellow-green carboxylated or aminated latex particles ( FluoSpheres ) with a mean diameter of 0 . 2 µm were purchased from Molecular Probes . Both the human intestinal cell line Caco-2 cell ( clone 1 ) ( obtained from Dr . Maria Rescigno , University of Milan-Bicocca , Milan , Italy ) [57] and CHO cells were cultured in Dulbecco's modified Eagle's medium ( DMEM ) ( PAA ) supplemented with 10% ( v/v ) fetal bovine serum ( FBS , Thermo-Fisher ) , 1% ( v/v ) nonessential amino-acids ( PAA ) , and 1% ( v/v ) penicillin-streptomycin ( PAA ) . The human Burkitt's lymphoma cell line Raji B ( American Type Culture Collection ) was cultured in RPMI 1640 supplemented with 10% ( v/v ) FBS , 1% ( v/v ) nonessential amino-acids , 1% ( v/v ) L-glutamine , and 1% ( v/v ) penicillin-streptomycin . The inverted FAE model ( Figure 1a ) has been previously reported [20] . Several major changes were made and are listed below . Inverted Transwell polycarbonate inserts ( 12 wells , pore diameter of 3 . 0 µm , Corning ) were coated with Matrigel , a basement membrane matrix ( BD Biosciences ) prepared in pure DMEM to a final protein concentration of 100 µg/ml for 1 h at room temperature . The coating solution was removed and inverted inserts washed with 300 µl of DMEM . Caco-2 cells ( 3×105 ) , resuspended in 300 µl of supplemented DMEM , were seeded on the lower insert side and cultured overnight . The inserts were then inverted and placed in a 12-well culture dish and kept for 9 d . Raji B cells ( 5×105 ) , resuspended in supplemented DMEM , were then added to the basolateral compartment of the Caco-2 cells , and co-cultures were maintained for 5 d . Mono-cultures of Caco-2 cells , cultivated as above but without the Raji B cells , were used as controls . Finally , the inserts were inverted in six-well plates , and a piece of silicon tubing ( 14×20 mm , Labomoderne ) was placed on the basolateral side of each insert . Cell monolayer integrity , both in mono- and co-cultures , was controlled by measurement of TEER using an Endohm tissue resistance chamber ( Endohm-12 , World Precision Instruments ) connected to a Millicell-ERS Ohmmeter ( Millipore ) . The resistance of medium alone ( 9 Ω×cm2 ) was considered as background resistance and subtracted from each TEER value . Barrier function of the tight junctions was also analyzed by zonula occludens-1 ( ZO-1 ) immunolabeling ( see next section ) . Cells morphologically similar to M cells were discriminated from Caco-2 cells using transmission electron microscopy ( TEM ) and scanning electron microscopy ( SEM ) . TEM and SEM were used to evaluate morphological cell changes after co-culture with Raji cells . Mono- and co-cultures were washed twice in HBSS and fixed in 4% ( v/v ) formaldehyde . Ultra-thin sections of cell-covered filters were prepared for TEM analysis by standard methods , as previously described [58] . Observations were made using a Hitachi H-800 and a Digital camera Hamamatsu AMT XR40 . Samples processed for SEM analyses were dehydrated , dried at critical point , and gold coated . Pictures of cell monolayers were obtained with a Thermo Noran Quest 2 L Hitachi S 3000N . Since no human-specific M cell markers have yet been identified , the microvilli-free morphology of M-like cells was used to identify and quantify them by SEM . Mono-cultures were used as controls . Characterization and quantification of M-like cells in co-cultures was further verified by immunolabeling . Inserts were washed in HBSS to eliminate residual medium , incubated in 4% paraformaldehyde for 30 min , permeabilized with 0 . 1% Triton X-100 ( Sigma-Aldrich ) , and blocked with PBS containing 5% FBS for 15 min at room temperature . Immunolabeling was performed using a combination of GFP-IgA2 , anti-human ZO-1 mAb ( Invitrogen ) , and mouse anti-human CA19 . 9 ( Dako ) [10] . Each reagent was diluted to 1/100 , and incubated for 2 h at room temperature . 1/200 dilutions of secondary antibodies labeled with a fluorochrome were incubated for 1 h at room temperature . After two washes , inserts were air-dried , mounted with Fluoprep ( BioMerieux ) , and observed by Immunofluorescence microscopy ( Eclipse Ti , Nikon ) . Nanoparticle ( NP ) ( yellow-green fluorescent , 0 . 2 µm carboxylate-modified FluoSpheres beads ) transport by polarized Caco-2 cells was evaluated in HBSS medium . NP concentration was adjusted to 4 . 5×109 NPs/ml and vortexed for 1 min to dissociate possible aggregates . NP suspension was added to the apical side of cell monolayers ( 400 µl ) and the inserts were incubated at 37°C for 90 min . Basolateral solutions were then sampled and the number of transported particles was measured by flow cytometry ( Facs Calibur , Becton Dickinson ) . The measurements were based on both fluorescence and particle size . Light and heavy chain encoding genes from a human TNF-alpha–specific IgA Ab were cloned in a single vector ( pGTRIO ) designed for efficient Ab expression in HEK293 and CHO cell lines . pGTRIO is a derivative of pVITRO2 ( Cayla-InvivoGen , Toulouse , France ) , a multigenic plasmid that contains two distinct transcription units . In pGTRIO , the antibiotic resistance gene is under the control of the EF1 alpha/HTLV promoter combined with the CMV enhancer that together constitutes a third transcription unit with the EF1 polyadenylation signal . The kappa constant region was cloned downstream of the FerL promoter together with the CMV enhancer , and the heavy chain constant regions were cloned downstream of the FerH promoter together with the human aldolase A enhancer . Unique restriction sites were introduced upstream of each constant region in order to allow the cloning of the variable region as SgrAI-BsiWI and AgeI-NheI fragments for VL and VH , respectively . All variable heavy chain regions were fused at the C-terminal end of secreted luciferase . CHO cells were transfected with pGTRIO constructs using the LyoVec system ( Cayla-InvivoGen ) in accordance with the manufacturer's instructions . Stable transfectants were selected in antibiotic-containing medium and screened for the production of Abs with a neutralizing activity on the HEK-Blue TNF-alpha/IL1-beta reporter cells ( Cayla-InvivoGen ) stimulated with TNF-alpha . IgA preparations were purified using Kappa affinity chromatography , IgG preparations were purified using protein G affinity chromatography , and IgE preparations were purified using protein L affinity chromatography . Ig-Luc constructs specific for TNF-alpha maintained their ability to block the cytokine , indicating proper assembly and folding . The following Abs were obtained by this method ( Figure 2 ) : human m-IgA2 ( monomer ) ; human GFP-IgA2 ( monomer ) ; human d-IgA2 ( + J chain - dimer ) ; murine m-IgA ( monomer ) ; human IgE; human IgG1; human m-IgA1 ( monomer ) ; human m-IgA2 G0 ( no glycosylation – monomer ) ; human m-IgA2 G1 ( 1 glycosylation ( Asn263 ) - monomer ) ; human m-IgA2 G2 ( 2 glycosylations ( Asn263 and Asn469 ) – monomer ) ; human m-IgA2 dPB ( without basal part – monomer ) ; human m-IgA2 dCα3 ( without basal part and Cα3 – monomer ) ; human m-IgA2 dCα2/3 ( without basal part , Cα3 and Cα2 – monomer ) ; human m-IgA2 Cα1 G0 ( only Cα1 without glycosylation ) ; and human m-IgA2 Cα1 ( only Cα1 with glycosylation ) . M-IgA2 was desialylated and deglycosylated with neuraminidase and PNGase , respectively ( Enzymatic CarboRelease Kit , QA-Bio ) . Purity and assembly of the Abs were controlled by SDS-PAGE ( Figure 2b/c ) . IgA deglycosylation was detected under normal conditions by using standard Western blot protocol with a combination of UEA-1-HRP and WGA-HRP lectins ( Sigma-Aldrich ) . For visualization of the mouse SIgA retrotranscytosis , a polymeric IgA Ab from the hybridoma clone IgAC5 specific to S . flexneri serotype 5a LPS [59] was obtained as previously described [60] . Purified free human SC was produced in Chinese hamster ovary cells [61] . SIgA molecules were obtained by combining in PBS pIgA molecules with a 2-fold excess of human SC for 2 h at room temperature according to the conditions described in the study by Rindisbacher et al . [62] . Cy3-SIgA molecules were obtained by conjugation with indocarbocyanine ( Cy3 ) using the FluoroLink mAb Cy3 labeling kit ( Amersham Biosciences ) according to the procedure provided by the manufacturer . Transport experiments were performed in HBSS at 37°C for 90 min with 10 µg of Ab conjugated with luciferase ( Luc ) . Basolateral solutions were then recovered and the number of retro-transcytosed Ab-Luc measured by luminometry ( Tristar LB941 , Berthold Technologies ) using the Gaussia Luc Assay Kit ( Biolux ) according to the procedure provided by the manufacturer . Ab-Luc transport was expressed as a mean value ± S . E . M . For inhibition experiments , cell monolayers were first preincubated apically with 5 mg of inhibitor in HBSS for 90 min at 37°C , and washed with HBSS , before adding the Ab-Luc suspension . All transport experiments were carried out in triplicate and were standardized with m-IgA2 ( ratio RLU/µg of Ab ) . Maxisorp 96-well plates were either coated with 50 µl of recombinant human Dectin-1/CLEC7A ( 5 µg/ml ) ( R&D Systems ) , 50 µl of recombinant human Siglec-5 ( 5 µg/ml ) ( R&D Systems ) , or 50 µl of an equal mixture of both Dectin-1 and Siglec-5 proteins and incubated O/N at 4°C . The wells were then washed three times with PBS and saturated with 200 µl of blocking solution ( PBS+3% BSA ) at room temperature for 1 h . The blocking solution was then discarded and 100 µl of m-IgA2 , colostrum IgA , m-IgA2+PNGase , or m-IgA1 were added at a concentration of 5 µg/ml . After 1 h of incubation at room temperature , wells were washed three times with PBS , and bound IgA was detected using biotinylated goat anti-human IgA ( Southern Biotech ) followed by streptavidin-HRP ( Amersham ) . Results are expressed as the means of OD ± SEM . Six-week-old C57BL/6 mice were purchased from Charles River Laboratories ( Lyon , France ) . CX3CR1-GFP transgenic mice were obtained from Maryline Cossin ( Joseph Fourier University , France ) . Dectin-1 knockout mice [63] , chimeric-KO:wt mice , chimeric-wt:KO mice , C57BL/6 mice , and CX3CR1-GFP transgenic mice were hosted at the University Hospital Unit for animal testing ( Saint-Etienne , France ) . For ileal loop preparation , mice were starved overnight , anesthetized by intra-peritoneal injection of a mix of ketamine and xylazine ( 100 and 10 mg/kg animal weight , respectively ) , and kept warm at 37°C throughout the surgical procedure . We administered 100 µl of a 1 mg/ml solution of SIgA-Cy3 or p24-SIgA diluted in PBS into a 1 . 5-cm ileal loop containing a PP . Upon completion of the experiment , the mice were sacrificed by cervical dislocation and the piece of intestine was removed , extensively washed with PBS , fixed for 2 h in 3% paraformaldehyde , and included in optimal cutting tissue ( OCT ) embedding solution . We captured 7-µm sections ( Leica cryostat model CM1950 , Leica Microsystems ) on Ultra+ superfrost microscope slides ( VWR International ) and stained for M cells . Slides were washed in PBS to eliminate residual OCT embedding solution , and blocked with PBS containing 5% FBS for 30 min at room temperature . Abs diluted to 20 µg/ml were incubated for 2 h at room temperature . The slides were then washed in PBS , air-dried , and mounted with Fluoprep ( Biomérieux ) . Slides were observed by immunofluorescence microscopy ( Eclipse , Nikon ) . Immunolabeling was performed using a combination of UEA-1-FITC ( Sigma-Aldrich ) , anti-human Dectin-1/CLEC 7A pAb , anti-human CD170 mAb ( R&D Systems ) , anti-human GP2 mAb ( MBL ) , and p24-specific Ab directly labeled with PE ( Santa Cruz Biotechnology ) . The protocol followed the guidance of the regional Ethics Committee for Animal Testing ( CREEA ) ( Permit Number No . 69387487 ) . Informed and consenting patients who had undergone upper duodenal endoscopy for routine diagnostic purposes ( e . g . , dyspepsia and chronic diarrhea ) with normal intestinal mucosa provided four to six biopsy samples from the distal duodenum . Biopsies were fixed for 2 h in 3% paraformaldehyde and included in OCT embedding solution , before being cryosectioned using a Leica cryostat model CM1950 . We captured 7 µm sections on Ultra+ Superfrost microscope slides , and they were stained for M cells as described for mouse intestine and observed by immunofluorescence microscopy . Immunolabeling was performed using a combination of GFP-IgA2 , anti-human Dectin-1/CLEC 7A pAb , and anti-human CD170 mAb ( R&D Systems ) . We housed 5–9-wk-old Dectin-1 KO and wt C57BL/6 males in individually ventilated cages at least 7 d prior to being irradiated . During this time and throughout the remainder of the experiment , animals were also maintained on sterile food and acidified in sterile water ( containing 0 . 004% HCl ) . Animals received two doses of full body irradiation at 5 Gy ( 2×500 rads ) . Each dose was separated by a 3 h interval to limit gastrointestinal problems . Irradiated mice were returned to individually ventilated cages for 24 h . Bone marrow was isolated from the femurs and tibia of donor Dectin-1 KO and wt C57BL/6 males under sterile conditions in the absence of red blood cell lysis . Nucleated cells were counted on a haemocytometer . Irradiated Dectin-1 KO mice each received 2×106 total nucleated bone marrow cells from wt C57BL/6 mice intravenously via the lateral tail vein . The phenotype of these mice , named chimeric-KO:wt , is thus wt at the systemic level and Dectin-1 KO at the mucosal level . Irradiated wt C57BL/6 mice were similarly injected with the same number of Dectin-1 KO donor cells . The phenotype of these mice , referred to as chimeric-wt:KO , is Dectin-1 KO at the systemic level and wt at the local level . Animals were maintained in individually ventilated cages as described above for a further 6 wk . Five weeks after bone marrow injections , 50 µl of tail vein blood was taken from each animal to characterize the cell phenotype , and red blood cells lysed for 2 min at room temperature in 1× Pharmlyse buffer . Cells were then washed twice in phosphate-buffered saline and counted . Cells were incubated for 15 min in FACS block ( HBSS+2 mM NaN3 , 0 . 5% BSA , and 5% heat inactivated rabbit serum ) containing 6 µg/ml Fc-receptor blocking mAb ( clone 24G2 ) , prior to addition of 10 µg/ml biotinylated anti-Dectin-1 mAb ( clone 2A11 ) or the rat biotinylated isotype control IgG2b for 30 min on ice . After three washes in FACS wash ( HBSS complemented with 2 mM NaN3 and 0 . 5% BSA ) , cells were incubated in FACS block containing 1/200 APC-conjugated streptavidin ( Invitrogen ) for 20 min on ice . Cells washed three times were analyzed on a FACSCalibur ( Becton-Dickinson ) and data analyzed using FlowJo software . Statistical analyses were performed using the InStat version 2 . 01 from the GraphPad Software , and the unpaired two-tail Mann–Whitney U test was applied . Significance limit was set at p≤0 . 05 .
|
Secretory IgA ( SIgA ) antibodies are secreted into the gut lumen and are considered to be a first line of defense in protecting the intestinal epithelium from gut pathogens . SIgA patrol the mucus and are usually known to help immune tolerance via entrapping dietary antigens and microorganisms and other mechanisms . SIgA , in complex with its antigens , can also be taken back up by the intestinal epithelium in a process known as reverse transcytosis . SIgA can thereby promote the uptake and delivery of antigens from the intestinal lumen to the Gut-Associated Lymphoid Tissues ( GALT ) , influencing inflammatory responses . This reverse transcytosis of SIgA is mediated by specialized epithelial M cells . Because M cells possess the ability to take up antigens and are therefore important to the local immune system , they are a key target for the specific delivery of novel mucosal vaccines against various diseases . M cell receptors that take up the SIgA-antigen complexes , which serve as mucosal vaccine vehicles , represent an important aspect of this vaccine strategy . The identification of SIgA receptor ( s ) on the surface of M cells has , however , remained elusive for more than a decade . In this study , we now identify Dectin-1 and Siglec-5 as the key receptors for M cell–mediated reverse transcytosis of SIgA complexes . We further find that the glycosylation modification , and particularly sialylation , of SIgA is required for its uptake by M cells . We show that , when administered orally in complex with SIgA , the HIV p24 antigen is taken up in a strictly Dectin-1-dependent manner to stimulate a mucosal and systemic antibody response . These findings are considered important for understanding gut immunity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2013
|
Dectin-1 Is Essential for Reverse Transcytosis of Glycosylated SIgA-Antigen Complexes by Intestinal M Cells
|
Beta-cell apoptosis and failure to induce beta-cell regeneration are hallmarks of type 2-like diabetes in mouse models . Here we show that islets from obese , diabetes-susceptible New Zealand Obese ( NZO ) mice , in contrast to diabetes-resistant C57BL/6J ( B6 ) -ob/ob mice , do not proliferate in response to an in-vivo glucose challenge but lose their beta-cells . Genome-wide RNAseq based transcriptomics indicated an induction of 22 cell cycle-associated genes in B6-ob/ob islets that did not respond in NZO islets . Of all genes differentially expressed in islets of the two strains , seven mapped to the diabesity QTL Nob3 , and were hypomorphic in either NZO ( Lefty1 , Apoa2 , Pcp4l1 , Mndal , Slamf7 , Pydc3 ) or B6 ( Ifi202b ) . Adenoviral overexpression of Lefty1 , Apoa2 , and Pcp4l1 in primary islet cells increased proliferation , whereas overexpression of Ifi202b suppressed it . We conclude that the identified genes in synergy with obesity and insulin resistance participate in adaptive islet hyperplasia and prevention from severe diabetes in B6-ob/ob mice .
The hallmark of type 2 diabetes ( T2D ) is a relative hypoinsulinemia which is unable to compensate insulin resistance and is a result of beta-cell failure [1] . In mouse models of obesity , differences in beta-cell adaptation to increased insulin demand are due to an underlying genetic background causing either diabetes susceptibility or resistance . B6-ob/ob mice lacking leptin on the C57BL/6 background become obese and insulin resistant but do not develop hyperglycemia because of massive beta-cell proliferation and high serum insulin levels . In contrast , the C57BLKS mice carrying the ob/ob mutation do not adapt to increased insulin requirements and die of severe hyperglycemia [2 , 3] . Here we compared the genetics and pathomechanisms of the diabetes resistant B6-ob/ob mouse with the diabetes-prone New Zealand Obese ( NZO ) mouse that represents an excellent model for polygenic obesity and type 2 diabetes and resembles the human disease . By positional cloning we have previously identified the adipogenic and diabetogenic genes Tbc1d1 [4] , Ifi202b [5] , and Zfp69 [6] of which the human orthologues also appear to be involved in the progression of human obesity and T2D [6–8] . Interestingly , restriction of carbohydrates protects diabetes-prone mice ( NZO , db/db on C57BLKS background ) from hyperglycemia , but re-exposure to carbohydrates causes rapid development of hyperglycemia and beta-cell apoptosis within a few days [9–11] . We have recently developed a dietary intervention which allowed a very fast and synchronized beta-cell failure in NZO-mice . By feeding a carbohydrate-free diet for about 15 weeks and subsequent treatment with carbohydrates we were able to follow early pathogenic alterations in the islets in a time course of 16 and 32 days . We discovered a rapid induction of hyperglycemia , a decreased AKT signaling and subsequent induction of apoptosis in the islets of Langerhans from NZO mice [10 , 12] . In the present study we used an integrated approach to identify novel disease genes . We compared the transcriptome profile of islets from NZO and B6-ob/ob mice that were treated with carbohydrates only for two days subsequent to the carbohydrate-free feeding and combined the expression pattern with a MetaCore enrichment pathway analysis and with QTL identified in the F2 ( NZOxB6 ) [13] . Several transcripts of cell-cycle regulation are likely to be associated with the pathogenesis of diabetes because their expression in islets of Langerhans is non-responsive to an in-vivo glucose challenge in the diabetes-sensitive strain . Mapping of differentially expressed genes to the major diabetes QTL Nob3 on chromosome 1 identified 15 genes of which 7 exhibited an exclusive expression in one strain . Direct evidence was obtained for four genes of which three appear to be involved in the compensatory capacity of B6-ob/ob islets: Lefty1 , Pcp4l1 , and Apoa2 exclusively expressed in B6-ob/ob islets increased proliferation . In contrast , Ifi202b which is not expressed in B6 suppressed proliferation when overexpressed in primary islet cells .
We have previously reported that a specific dietary regimen consisting of pretreatment with carbohydrate-free diet for 15 weeks and a subsequent intervention with a carbohydrate-containing diet results in completely different reactions in diabetes-susceptible NZO and diabetes-resistant B6-ob/ob mice . NZO mice develop hyperglycemia ( 20 . 2 ± 2 . 2 mM blood glucose at day 16 ) and lose most of their beta-cells by day 32 ( Fig 1A , upper panels ) , whereas B6-ob/ob mice were able to compensate after a transient increase in blood glucose concentrations showing 8 . 0 ± 0 . 1 mM blood glucose at day 16 [12] . This might be due to a massive islet hyperplasia that was detected at the end of the intervention state as demonstrated in a representative overview of total pancreas sections of mice before and after receiving carbohydrates ( Fig 1A , upper panels ) . Morphometric analysis demonstrated that size of islets in B6-ob/ob mice increased in response to the carbohydrate challenge rather than their number ( Fig 1A , lower panels ) . In order to confirm that the carbohydrate intervention induces proliferation of islet cells of B6-ob/ob mice we determined Ki-67 positive nuclei in pancreatic sections of NZO and B6-ob/ob mice that were treated with carbohydrates for 2 , 4 , 8 , and 32 days . At no time point an induction of proliferation occurred in NZO islets , whereas B6-ob/ob islets showed a transient induction of proliferation . Highest numbers of Ki-67 positive islet cells were detected at day 2 , and these numbers dropped to the initial level up to day 32 of treatment ( Fig 1B ) . For the identification of genes and pathways that participate in the prevention of diabetes we performed RNAseq based transcriptome analyses of isolated islets of NZO and B6-ob/ob mice two days after the diet switch . MetaCore enrichment pathway analysis with 2882 differentially expressed islet genes in NZO and B6-ob/ob ( log2 FC > │0 . 6│; P<0 . 05 ) revealed an upregulation of 22 genes involved in regulation of cell cycle and 8 genes of glutathione metabolism in B6-ob/ob islets , whereas NZO islets exhibited an elevated expression of 29 transcripts mediating cell adhesion and of 7 genes associated with inflammation ( Table 1 and Fig 2 ) . In order to identify the genes responsible for the diabetes resistance of B6-ob/ob and the diabetes sensitivity of NZO mice we projected the differentially expressed islet genes to the major diabesity QTL we have recently identified in a F2 generation of NZO x C57BL/6 . This QTL ( Nob3 ) with LODscore values of 16 . 1 , 16 . 0 , 4 . 0 for body weight , body fat , and blood glucose , respectively was located on the distal part of Chr . 1 [13] . By defining a minimal log2 FC of │1 . 5│ we found 15 genes to be differentially expressed within the peak region ( 162 . 6 to 192 . 6 Mbp ) of Nob3 ( Table 2 ) . Among these we found 6 genes with an exclusive expression ( <15% rest mRNA ) in B6-ob/ob islets ( Lefty1 , Pcp4l1 , Apoa2 , Mndal , Slamf7 , Pydc3 ) and one , Ifi202b solely in NZO islets ( Fig 3 ) . These fundamental differences were confirmed by qRT-PCR in a new batch of isolated islets ( S1 Fig ) . In contrast to the RNAseq results we detected higher Apoa2 mRNA levels in NZO islets , however , the Apoa2 expression in B6-ob/ob islets was strikingly higher . A major difference in the response to the carbohydrate challenge between NZO and B6-ob/ob islets was the lack and induction of proliferation , respectively ( Figs 1 and 2 ) . We therefore tested if genes within Nob3 which exhibited the major expression difference in islets of both mice , Ifi202b , Lefty1 , Pcp4l1 , Apoa2 , Mndal , Slamf7 , and Pydc3 directly modify islet cell proliferation . We overexpressed each gene in primary islet cells of B6 mice by adenoviral-mediated infection and subsequently analyzed the proliferation capacity by detecting the BrdU incorporation . In order to estimate the levels of beta-cells within primary islet cells and to test if beta-cells are able to proliferate we infected cells with an empty virus , incubated the cells with BrdU and performed a co-staining of BrdU and insulin 72 h later ( Fig 4A ) . As expected , most of the cells were beta-cells and could be distinguished from fibroblasts due to their different shape . We next infected primary islet cells with viruses encoding for ApoA2 , Lefty1 , Pcp4l1 , Mndal , Slamf7 , Ifi202b , and Pydc3 in combination with BrdU incorporation . The overexpression of each candidate was visualized by immunostaining of the respective myc-tag; the infection rate was between 40 and 90% ( Fig 4B–4I ) . Notably , every islet cell culture was accompanied by a growth of fibroblast-like cells , however , as these cells started to grow after virus treatment they were not infected and thereby not myc-positive ( Fig 4 ) . Presumably , according to individual differences of islet isolation , digestion , and growth the BrdU incorporation rate varied between the experiments in a range from 1 . 2 to 7 . 4% ( mean: 3 . 6 ± 0 . 6% ) . In most cases BrdU was detected in insulin-positive cells [95% ( AdØ ) , 96% ( AdApoa2 ) , 93% ( AdLefty1 ) , 91% ( AdPcp4l1 ) , 88% ( AdMndal ) , 93% ( AdSlamf7 ) , 91% ( AdIfi202b ) , and 94% ( AdPydc3 ) ] . Infection of cells with the Lefty1 , Pcp4l1 , and Apoa2 encoding viruses increased BrdU incorporation by 87% , 80% , and 92% , respectively; overexpression of Ifi202b decreased the number of BrdU positive cells by 43% . All other candidates were without effect ( Fig 5 ) . Previous human studies have linked type 2 diabetes traits to chr . 1q21-24 in multiple samples [14] . It was shown that seven variants increased the diabetes risk significantly through 5 pairs of interactions . One significant variant pair was a SNP in APOA2 ( rs6413453 ) interacting with calsequestrin-1 ( CASQ1; rs617698 ) [14] . As no association was described for LEFTY1 so far we tested a population of 1 , 865 individuals with a median age of ~39 years and a BMI of 28 . 5 kg/m² that was recruited from the Tübingen Family ( TUEF ) study for type 2 diabetes and characterized before [15] . The participants were genotyped for 10 SNPs tagging the LEFTY1 gene and 2 kb of its 5’-flanking region for association analysis . After adjusting for gender , age , and oGTT-derived insulin sensitivity the minor T-allele of rs3806259 was nominally associated with elevated insulin levels during oral glucose tolerance tests ( P = 0 . 0034 ) , whereas the minor allele of rs360060 was nominally associated with a lower insulin release ( P = 0 . 0052; S2 Fig ) . Both SNPs were also nominally associated with elevated ( P = 0 . 0593 ) and reduced ( P = 0 . 0111 ) C-peptide release , respectively .
The present data demonstrated that islets of diabetes-resistant and diabetes-prone mouse strains respond inversely to a glucose challenge: In B6-ob/ob islets , the induction of proliferation resulted in a strong increase in beta-cell mass , which appears to be responsible for an increase in plasma insulin concentration and a sufficient control of blood glucose even under glucotoxic conditions , whereas NZO islets lack the ability to initiate cell cycle progression and to maintain the survival pathway . We identified 7 genes within the diabesity QTL Nob3 which exhibit a nearly exclusive expression in islets of either diabetes-susceptible or diabetes-resistant mice . Three genes , Lefty1 , Pcp4l1 , and Apoa2 are the most likely candidates to suppress and one gene , Ifi202b to accelerate T2D and beta-cell failure in mice . To clarify the genetic basis of the inverse reaction of NZO and B6-ob/ob islets on the glucose challenge and to identify diabetes suppressor and enhancer genes we aligned the transcriptome data to the major susceptibility locus identified in former association studies from a NZOxC57BL/6J cross , the Nob3 with a LOD score of 16 for adiposity and of 4 for hyperglycemia [13] . In total , 15 differentially expressed islet genes are located in the Nob3 . Focusing on transcripts with the strongest effect revealed six genes ( ApoA2 , Lefty1 , Mndal , Pcp4l1 , Pycd3 , and Slamf7 ) to be exclusively expressed in B6-ob/ob and one , Ifi202 in NZO islets . The most prominent molecular alterations were identified for Ifi202b and Lefty1 . Ifi202b which was recently identified as an obesity gene in our group by positional cloning carries a microdeletion that includes the promoter and the first exon of Ifi202b and results in a loss of Ifi202b expression in most tissues from B6 mice such as white adipose tissue , liver , and skeletal muscle [5] , and as shown here , also in islets . Since the overexpression of Ifi202b in primary islet cells suppressed proliferation we conclude that it participates in the inability of NZO mice to compensate with an increased beta-cell mass when blood glucose levels rise . Consistent with this conclusion , Clarke et al . ( 2010 ) demonstrated that the overexpression of IFI16 , a human orthologue of Ifi202b , decreases cellular proliferation by growth arrest in the G1 phase of the cell cycle [16] . Lefty1 , one candidate exclusively expressed in islets of diabetes-resistant B6-ob/ob mice , is an atypical member of the TGF-β family that inhibits Nodal by direct binding to Nodal [17–19] . Recently , Zhang et al . ( 2008 ) reported that Nodal and Lefty1 are expressed in the pancreas during embryogenesis and islet regeneration and demonstrated that Lefty1 activates MAPK and AKT in vitro [20] . Our in vitro studies in primary islet cells clearly supported that Lefty1 expression increases proliferation as detected by an increased BrdU incorporation . As Lefty1 is a central regulator of the Smad3 pathway by antagonizing Nodal , we screened our RNAseq and pathway enrichment data for downstream transcripts and indeed found several targets exhibiting an altered expression as indicated in Fig 6A . Nodal activates Smad3 , a transcription factor which exhibits a higher expression in NZO than in B6-ob/ob islets ( Figs 6A and S3 ) . Smad3 is a tumor suppressor which inhibits cell proliferation and promotes apoptosis [21] . Accordingly , the lower expression of Smad3 in B6-ob/ob islets as a consequence of the presence of Lefty1 could trigger beta-cell proliferation in this diabetes-resistant model ( Fig 6A ) . Specifically , in pancreatic beta-cells Smad3 activation was shown to promote the Cdc25a ubiquitination and degradation via two ubiquitin ligases , the SCFbeta-TrCP-complex and the anaphase-promoting complex ( APCCdh1 ) [22] . Cyclin A2 ( Ccna2 ) , another target to be suppressed by the APCCdh1-complex [23 , 24] , is elevated specifically in B6-ob/ob islets in response to carbohydrate feeding and plays a pivotal role in cell-cycle regulation together with the targets Cdk1 and Cdk2 that are also upregulated as determined by qRT-PCR ( Fig 6B ) [25] . In contrast , without Lefty1 NZO islets display higher Smad3 levels that might participate in raising the expression of ECM molecules such as collagen and fibronectin [26 , 27] and an increased expression of Mmp7 and Mmp12 that cleave the ECM proteins vitronectin and laminin ( Fig 2 and Table 1 ) . We found that two SNPs in human LEFTY1 were nominally associated with altered insulin release in a glucose tolerance test , but not with glycaemia or diabetes . This association seems to suggest that LEFTY1 could be a minor contributor to the pathogenesis of the human disease . However , the finding needs validation in an independent study before a definitive conclusion can be drawn . No significant association between diabetes-related traits and the genotype of 15 SNPs in IFI16 , the human orthologue of Ifi202b , was found . Besides Lefty1 , Apoa2 showed an excessive expression in the B6-ob/ob islet and in primary islet cells also a positive effect on proliferation . In fact SNPs in human APOA2 were shown to associate with type 2 diabetes in a Utah case-control sample [14] . However , this finding was not confirmed in French Caucasian subjects [28] . Pcp4l1 ( Purkinje cell protein 4 like 1 ) was also sufficient to increase BrdU incorporation in primary islet cells . Little is known about the function of Pcp4l1 . It is described to be expressed during early brain development where it is localized to the midbrain-hindbrain boundary in a pattern that partially overlaps with Wnt1 , Pax2 and Fgf8 expression domains [29] . Its closest related protein is Pcp4/PEP-19 , a calmodulin-binding protein exhibiting a neuronal IQ motif . However , in contrast to Pcp4/PEP-19 , Pcp4l1 lacks the capacity of calmodulin binding presumably due to a nine-residue glutamic acid-rich sequence that is located in proximity to the IQ motif [30] . A link of Pcp4l1 and diabetes has not been described so far . In summary , our data indicate that Lefty1 , Apoa2 , Pcp4l1 and Ifi202b modify beta-cell proliferation , and suggest that the alteration of their expression contributes to the development of severe diabetes in the NZO mouse strain .
Male NZO/HIBomDife mice ( R . Kluge , German Institute of Human Nutrition , Nuthetal , Germany ) from our own colony and male B6 . V-Lepob/ob/JBomTac ( ob/ob ) mice ( Charles River Laboratories , Italy ) were housed in groups of at most five per cage ( type II/III macrolon ) at a temperature of 21 ± 2°C with a 12:12 h light-dark cycle ( lights on at 6:00 a . m . ) . Animals had free access to food and water at any time . Starting at the age of 5 weeks onwards , all mice received a carbohydrate-free diet ( -CH ) containing 68% fat ( w/w ) and 20% protein ( w/w ) with a total metabolisable energy of 29 . 3 kJ/g . At the age of 18 ± 1 weeks , subgroups of the animals received a carbohydrate-enriched diet ( +CH ) containing 28% fat ( w/w ) , 20% protein ( w/w ) and 40% ( w/w ) metabolisable carbohydrates for either 48 hours or 32 days . The metabolisable energy content of the +CH was 21 . 9 kJ/g . Due to the soft texture of the diets mice had access to wooden gnawing sticks in order to avoid excessive teeth growth . The ingredients of the diets are given in S1 Table . Pancreatic tissue excised immediately after Isofluran anesthesia , was fixed in 4% formaldehyde for 24 h and embedded in paraffin according to standard procedures . Longitudinal serial sections ( 2–3 μm thickness , sampling intervals 100 μm ) were prepared . After re-hydration , slices were incubated with primary antibodies against insulin ( 1:50 , 000; 1 h; room temperature; Sigma ) and Ki-67 ( 1:50; overnight; 4°C; Dako; Hamburg; Germany ) . Bound primary antibodies were detected with Histofine Immuno-POD Polymer ( Nichirei Biosciences , Tokyo , Japan ) against mouse or rat , followed by Diaminobenzidine ( DAB ) incubation for visualization of protein localization . Detection of apoptotic cells by a TUNEL-staining was performed by the ApopTag Plus Peroxidase In Situ Apoptosis Detection Kit ( Billerica , US ) according to manufacturer’s instructions . Microscopic investigation and photo documentation were performed with the combined light and fluorescence microscope ECLIPSE E-100 ( Nikon , Düsseldorf , Germany ) in combination with video camera equipment ( CCD-1300CB , Vosskühler , Osnabrück , Germany ) and the analysis system NIS elements ( Nikon ) . Total pancreas photographs were made by a MIRAX MIDI scanner ( Zeiss , Oberkochen , Germany ) and evaluated by a software package ( AxioVision , Zeiss ) . Morphometric analyses of Ki-67 staining were performed by counting the number of positive nuclei in 3 sectional planes of one pancreas with the help of the analysis software NIS elements ( Nikon ) . Mean values derived from each individual mouse were used to calculate mean ± s . e . m . ( 4–6 animals in each group ) . Insulin stainings were used to determine islet size , islet number and islet mass of B6-ob/ob and NZO mice . Similarly , quantifications were made from 3 sectional planes of one pancreas and means from each animal were used to calculate mean ± s . e . m . ( 3–5 animals per group ) . This analysis was realized by using the AxioVision software package ( Zeiss ) . All morphometric quantifications were conducted via an unbiased stereological approach . Isolation of islets was performed by a modified protocol of Gotoh et al . ( 1990 ) from 18 ± 1 week old NZO and B6-ob/ob mice [31] . Briefly , the pancreas was perfused by injection of 3 ml of Collagenase-P ( Roche , Mannheim , Germany ) ( 1 mg/mL ) in Hank’s buffered salt solution ( HBSS ) also containing 25 mM HEPES and 0 . 5% ( w/v ) BSA into the common bile duct . Subsequently , the perfused pancreas was digested in 2 ml of collagenase solution for 8 to 10 min at 37°C . With the help of a cannula ( 18G x 11’2 ) islets were mechanically detached from exocrine parts , and washed for several times with HBSS and RPMI 1640 containing 10% FCS and 100 U/mL penicillin/streptomycin ( PeSt ) . Islets were passaged for several times in RPMI 1640 by hand-picking and finally collected in Eppendorf tubes . Total islet RNA preparation was performed with the RNAqueous-Micro Kit ( Life Technologies , Darmstadt , Germany ) . In order to apply high-quality RNA for RNAseq analysis we assessed RNA integrity by using a Bioanalyzer and an appropriate kit ( RNA6000 nano kit , Agilent , Santa Clara , USA ) . All preparations were made according to manufacturer’s recommendations . Only RNA with a minimal RNA integrity number of 8 . 0 ( RIN ) was used . RNA concentration was measured in 1 μL with a spectrophotometer ( NanoDrop ND-100 UV/Vis , PEQLAB , Erlangen , Germany ) at 260 nm . RNA sequencing and evaluation of data was performed by LGC genomics ( Berlin , Germany ) . Total islet-RNA of NZO and B6-ob/ob mice fed with the carbohydrate-free diet up to the age of 18 weeks and with or without the carbohydrate-containing diet for 2 days was isolated with the RNAqueous-Micro Kit ( Life Technologies , Darmstadt , Germany ) . Subsequently , first-strand cDNA synthesis was performed with the whole amount of RNA , random hexamer primer and GoScript reverse transcriptase ( Invitrogen ) . qRT-PCR was performed with 12 . 5 ng cDNA in an Applied Biosystems 7500 Fast Real-time PCR system with TaqMan Gene Expression Assays ( Life Technologies , Darmstadt , Germany ) and the TaqMan Gene Expression Master Mix . Data were normalized to the expression of cyclophilin ( Ppia ) as endogenous control ( ΔCt-Method ) [32] . For overexpression of candidates in primary islet cells islets from C57BL/6 mice fed a high-fat diet until an age of 20 ± 2 weeks were isolated and recovered for 48 hours in RPMI 1640 ( 10% FCS , 1% PeSt ) in humidified 5% CO2 , 95% air at 37°C . Islets were digested with Accutase ( GE Healthcare , Buckinghamshire , GB ) in a volume of 75 μL per 100 islets for 6 min at 37°C and cells separated by aspirating . The digestion was stopped by addition of RPMI 1640 ( 10% FCS; 1% PeSt ) , cells were centrifuged ( 1000xg , 5 min , RT ) and the supernatant removed . Resuspended cells were seeded on poly-L-lysine coated coverslips in 24 well plates at a corresponding amount of 1 . 5 x 105 cells/well . Cells were maintained in RPMI 1640 ( 10% FCS , 1 & PeSt ) containing 11 mmol/L glucose for 3–4 days . Infection was carried out by a 24 h incubation with adenoviruses carrying cDNA from Pydc3 ( MOI100 ) , Lefty1 ( MOI100 ) , Slamf7 ( MOI10 ) , Mndal ( MOI100 ) , Apoa2 ( MOI100 ) , Pcp4l1 ( MOI100 ) ( Vector Biolabs ) and Ifi202b ( MOI500 ) ( ViraQuest , North Liberty , USA ) under the control of a CMV-promotor and in conjunction with the corresponding empty viruses . Due to the very low proliferation capacity of primary islet cells BrdU labeling ( 100 μmol/L ) was performed for 72 h after virus infection as described by Tsukiyama and colleagues [33] . Afterwards , cells were fixed with 4% paraformaldehyde . After permeabilization of cell membranes ( 0 . 2% saponin ) , DNA denaturation ( 2 M HCl ) and trypsinization ( 0 . 1% trypsin ) cells were incubated with primary antibodies against BrdU ( 1:500 , 1 h , room temperature; Sigma ) plus anti myc-antibody ( 1:500 , Santa Cruz , USA ) . Bound primary antibodies were detected with fluorophor-labeled secondary antibodies against rat ( Alexa Fluor546 , 1:400 , Invitrogen ) and rabbit ( Alexa Fluor488 , 1:400 , Invitrogen ) and documented with a confocal microscope ( TCS SP-2-Confocal Laser scanning microscope , Leica Microsystems , Heidelberg , Germany ) . Statistical evaluation was performed by blinded quantification of 10–12 photographs of at least 2 coverslips per infection and mean ± s . e . m . was calculated from 4–5 independent experiments . A study population of 1 , 865 individuals with complete insulin , C-peptide , and glucose measurements during a standard 5-point oral glucose tolerance test ( oGTT ) was recruited from the Tübingen Family study for type 2 diabetes [15] . From these data , insulin release was estimated as area under the curve ( AUC ) insulin0–30 divided by AUC glucose0–30 and as AUC C-peptide0–30 divided by AUC glucose0–30 . AUCs were calculated according to the trapezoid method . oGTT-derived insulin sensitivity was calculated as proposed by Matsuda and DeFronzo [34] . Based on publicly available information from 1000 Genomes ( http://browser . 1000genomes . org/Homo_sapiens/Info/Index ) , we analysed the complete human LEFTY1 and IFI16 genes and 2 kb each of their 5’-flanking regions and identified 10 and 15 , respectively , non-linked representative SNPs tagging all other common SNPs ( minor allele frequency ≥ 0 . 1 ) in these loci with r² ≥ 0 . 8 . The tagger function of Haploview 4 . 2 ( http://www . broadinstitute . org . haploview/haploview ) was used . The 10 LEFTY1 tagging SNPs , i . e . , rs7551003 G>T , rs72754996 G>A , rs193327 C>G , rs3806259 C>T , rs2013363 G>T , rs360074 G>A , rs10915896 C>T , rs360073 G>A , rs360060 G>A , and rs360058 G>T , were all non-coding and located in the 5’-flanking region of the gene . The 15 non-coding IFI16 tagging SNPs were rs72704962 T>C , rs1465175 G>T , rs4657618 C>T , rs1417806 A>C , rs856077 C>A , and rs1417805 T>G in the 5’-flanking region , rs2276404 A>G , rs856064 C>T , and rs12727764 G>T in intron 1 , rs2106095 T>A , rs12087333 A>G , rs74122213 T>C , and rs12061401 C>T in intron 6 , and rs1633262 T>C and rs1772414 A>G in intron 7 of the gene . The tagging SNPs were genotyped in blood-derived genomic DNA using Sequenom’s massARRAY system ( Sequenom , Hamburg , Germany ) or TaqMan allelic discrimination assays ( Applied Biosystems , Foster City , CA , USA ) . RNAseq read counts were tested by 2 statistical evaluations ( DEseq [35] ) , and edgeR [36] of the R package in combination with an error correction by Benjamini-Hochberg . All comparisons were considered significantly different at P<0 . 05 and a log2 FC >│0 . 6│and │1 . 5│ , respectively . Differences in morphometric analysis , qRT-PCRs and BrdU-Assays were tested with the non-parametric Kruskal-Wallis test and corrected for multiple comparisons by Dunn’s . For human association studies skewed data were loge-transformed in order to approximate normal distribution . Multiple linear regression analysis ( standard least squares method ) was performed with insulin release as outcome variable , the SNP genotype ( in the additive inheritance model ) as independent variable , and gender , age , and insulin sensitivity as confounding variables . To account for multiple testing ( 25 SNPs tested in parallel ) , a Bonferroni-corrected P-value <0 . 002 was considered statistically significant . For these analyses , JMP 10 . 0 ( SAS Institute , Cary , NC , USA ) was used . All animals were kept in accordance with the NIH guidelines for the care and use of laboratory animals . All experiments were approved by the Ethics Committee of the State Ministry of Agriculture , Nutrition and Forestry ( State of Brandenburg , Germany ) . The approval numbers were: V3-2347-31-2011 and 2347-26-2014 .
|
Complex genetic determinants contribute to an inherent susceptibility of type 2 diabetes , characterized by insulin resistance , a dysfunction and loss of insulin-producing beta-cells . We compared the islet expression profile and the genome of two obese mouse strains that react differently when receiving a caloric enriched diet . One mouse ( B6-ob/ob ) is able to compensate by increasing the beta-cell mass , whereas the other ( NZO ) develops hyperglycemia due to beta-cells loss . Focusing on differentially expressed genes that are located in susceptibility locus for diabetes and obesity on chromosome 1 we found 6 genes to be only expressed in islets of the diabetes-resistant mouse and one to be exclusively present in islets of the diabetes-prone mouse . Among these , the overexpression of 3 genes ( Lefty1 , Apoa2 , and Pcp4l1 ) increased and that of Ifi202b decreased the division of primary islet cells . In summary , our data provide new insights into genes inducing or inhibiting islet size and thereby participate in the pathogenesis of type 2 diabetes .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Material",
"and",
"Methods"
] |
[] |
2015
|
Identification of Four Mouse Diabetes Candidate Genes Altering β-Cell Proliferation
|
Orbiliomycetes is one of the earliest diverging branches of the filamentous ascomycetes . The class contains nematode-trapping fungi that form unique infection structures , called traps , to capture and kill free-living nematodes . The traps have evolved differently along several lineages and include adhesive traps ( knobs , nets or branches ) and constricting rings . We show , by genome sequencing of the knob-forming species Monacrosporium haptotylum and comparison with the net-forming species Arthrobotrys oligospora , that two genomic mechanisms are likely to have been important for the adaptation to parasitism in these fungi . Firstly , the expansion of protein domain families and the large number of species-specific genes indicated that gene duplication followed by functional diversification had a major role in the evolution of the nematode-trapping fungi . Gene expression indicated that many of these genes are important for pathogenicity . Secondly , gene expression of orthologs between the two fungi during infection indicated that differential regulation was an important mechanism for the evolution of parasitism in nematode-trapping fungi . Many of the highly expressed and highly upregulated M . haptotylum transcripts during the early stages of nematode infection were species-specific and encoded small secreted proteins ( SSPs ) that were affected by repeat-induced point mutations ( RIP ) . An active RIP mechanism was revealed by lack of repeats , dinucleotide bias in repeats and genes , low proportion of recent gene duplicates , and reduction of recent gene family expansions . The high expression and rapid divergence of SSPs indicate a striking similarity in the infection mechanisms of nematode-trapping fungi and plant and insect pathogens from the crown groups of the filamentous ascomycetes ( Pezizomycotina ) . The patterns of gene family expansions in the nematode-trapping fungi were more similar to plant pathogens than to insect and animal pathogens . The observation of RIP activity in the Orbiliomycetes suggested that this mechanism was present early in the evolution of the filamentous ascomycetes .
Ascomycota is the largest phylum of kingdom Fungi and includes approximately 33 , 000 described species [1] . The phylum is divided into three monophyletic subphyla: Taphrinomycotina , Saccharomycotina and Pezizomycotina [2] . Pezizomycotina is the largest subphylum and includes the vast majority of filamentous , fruit-body-producing species . Molecular phylogeny resolves Orbiliomycetes and Pezizomycetes as the early-diverging lineages of the Pezizomycotina , with the remaining seven classes sampled forming a well-supported crown clade [2] . The Orbiliomycetes consists of a single order ( Orbiliales ) and one family ( Orbiliaceae ) . This family is best known for containing nematode-trapping fungi [3] . These soil-living fungi capture and kill nematodes using specialized infection structures [4] , [5] , which are morphologically distinct structures called traps . The remarkable morphological adaptations and the dramatic infection process of the nematode-trapping fungi have fascinated mycologists for centuries . Another reason for the interest in the nematode-trapping fungi has been their ability to act as biocontrol agents against parasitic nematodes [5] . The morphology of the nematode trap differs depending on the species , and the major types group according to molecular phylogeny data [6]–[8] . The traps can be divided into four major types: adhesive nets , adhesive knobs , adhesive branches and constricting rings [9] . Species that form constricting rings are monophyletic and found near the base of the tree of nematode-trapping fungi . Among the species that form adhesive traps , those that form adhesive knobs and adhesive nets are a sister clade separated from those that form adhesive branches [6] . Species with adhesive traps capture nematodes using extracellular polymers that accumulate at the site of infection [10] , whereas those with constricting rings ensnare the nematode by rapid swelling of the ring cells . In both adhesive and constricting-ring types , the cuticles of the captured nematode are penetrated and an infection bulb is formed inside the nematode . At the time of penetration , the nematode is paralyzed . Subsequently , the nematode is killed , fungal hyphae grow inside it , and fungal enzymes degrade its tissues . Finally , the nutrients are taken up and translocated to new mycelia that grow out from the digested nematode [11] ( Figure 1 ) . Although the morphology of traps varies extensively , the nematode-trapping fungi are generalists and they can infect many different nematode species [5] . Over the last few years , comparative analyses of genome sequences have provided novel insights into the evolution of the diverse lifestyles of the filamentous ascomycetes . The species include those feeding on dead organic matter only ( saprophytes ) [12]–[14] , plant pathogens [15] , [16] , human pathogens [17] , [18] and insect pathogens [19] . All these species belong to the crown clades of the Pezizomycotina . Recently , the first genome of a nematode-trapping fungus of the Orbiliomycetes was sequenced [20] . The species , Arthrobotrys oligospora forms adhesive networks . Comparative genomics of A . oligospora with 10 other fungal genomes revealed several genes that were shared specifically between A . oligospora and pathogenic fungi . In addition , large gene families related to pathogenicity were identified in the A . oligospora genome , including the subtilisin , cellulase , cellobiohydrolase and pectin esterase families [20] . In addition , the regulation of proteins during trap formation was studied . A majority of the upregulated proteins were classified as being involved in translation , posttranslational modification , amino acid metabolism , carbohydrate metabolism , energy conversion , cell wall and membrane biogenesis [20] . To gain further insights into the evolution of parasitism in the nematode-trapping fungi of the Orbiliomycetes , in this study we have sequenced the genome of a knob-forming species , Monacrosporium haptotylum . A recent study using five protein-coding genes estimated the split between species that form adhesive nets and those that form adhesive knobs to have occurred 198–208 million years ago [21] . However , the taxonomic assignment of the fossils and the identification of traps has been questioned [22] and therefore the evolutionary history of the nematode-trapping fungi remains uncertain . Together with A . oligospora , M . haptotylum is the nematode-trapping fungus in which the infection mechanism has been studied in most detail [23] . A major advantage of using M . haptotylum in infection experiments is that the trap cells ( knobs ) can be isolated from a mycelium growing in liquid cultures [24] . The isolated traps are functionally intact , that is , they can capture and kill nematodes , including Caenorhabditis briggsae ( Figure 1 ) . Accordingly , the system provides unique opportunities to identify genes that are differentially expressed in the trap cells and in the fungus during the various stages of infection [25] , [26] . Our comparative genomics studies of M . haptotylum and A . oligospora showed that two genomic mechanisms are likely to be involved in the adaptation to pathogenicity in nematode-trapping fungi of the Orbiliomycetes . Firstly , gene duplications including gene family expansions indicate that the formation of novel genes was an important mechanism during the evolution of parasitism . Secondly , gene expression of orthologs indicates that differential gene expression between the two nematode-trapping fungi was another important adaptation to parasitism . Lineage- and species-specific genes are significantly shorter than genes shared by other fungi ( core genes ) . Many of these short genes encode secretory proteins and are small secreted proteins ( SSPs ) , which are likely to contain proteins that directly interact with the host during infection . The SSPs have undergone rapid divergence and many are orphans , that is , they lack known homologs and do not contain any Pfam domains . Expression of many of the SSPs was highly upregulated and expressed during nematode infection . We propose that the evolution of the infection-expressed SSPs involved tandem duplications followed by a rapid divergence governed by a repeat-induced point mutation ( RIP ) mechanism . The RIP mechanism has been demonstrated to play a central role in the genome evolution of several ascomycetes in the crown clades of the Pezizomycotina [12] , [15] , [19] , [27] . In addition , RIP has been identified in A . oligospora [20] and M . haptotylum ( this study ) , both belonging to the Orbiliomycetes . The presence of a RIP mechanism in the Orbiliomycetes suggests that RIP was present early in the evolution of the filamentous ascomycetes ( Pezizomycotina ) . The rapid divergence of SSPs coupled with differential expression of these genes indicates a striking similarity in the evolution of the infection mechanisms of the nematode-trapping fungi from the basal lineages and of plant pathogenic and insect pathogenic fungi from the crown clades of the filamentous ascomycetes [15] , [16] , [19] . An analysis of the expansion of gene families along the lineages of pathogenic fungi of the Pezizomycotina revealed that patterns of expansion in the nematode-trapping fungi are most similar to that in plant pathogenic fungi .
The genome of M . haptotylum was sequenced to 28× coverage by 454 pyrosequencing ( Table S1 ) . Based on these data , the genome size of M . haptotylum was estimated to be 40 . 4 Mb and the number of protein-coding genes to be 10 , 959 . The completeness of the sequenced genome and prediction of open reading frames were validated by analyses of transcriptome sequences . In total , 99% of the 422 , 883 pyrosequencing reads and 83–99% of previously generated expressed sequence tag ( EST ) sequences [25] , [26] were unambiguously mapped to the genome ( Table S2 ) . The genome assembly is therefore likely to cover the vast majority of the genes in the genome of M . haptotylum . Furthermore , analyses of RNASeq data showed that almost all of the predicted protein-coding genes ( 10 , 899 out of 10 , 959 ) were expressed by the fungus either in the saprophytic or the parasitic stage . The estimated genome size and number of protein-coding genes in M . haptotylum are almost identical to the equivalent numbers for A . oligospora ( Table 1 ) and similar to those of other ascomycetes ( Table S3 ) . In total , 149 putative tRNA genes were identified in the genome of M . haptotylum ( Table S4 ) , which is similar to the number in A . oligospora ( Table 1 ) , in insect pathogenic fungi [19] and in several other fungi ( http://lowelab . ucsc . edu/GtRNAdb ) . The distribution of proteins into different EuKaryotic Orthologous Groups ( KOG ) categories was similar in M . haptotylum and A . oligospora ( Figure S1 ) . The M . haptotylum genome contained 271 genes per Mb and the average number of exons per gene was 3 . 3 , which is similar to that of the A . oligospora genome . Transposable elements ( TEs ) were identified in M . haptotylum and A . oligospora and annotated as retrotransposons ( Class I ) or DNA transposons ( Class II ) ( Table S5 ) . The genome of M . haptotylum contains approximately twice as many TEs as the genome of A . oligospora . In both organisms two-thirds of the TEs belong to Class I . For most TEs the numbers located in the genes were similar in M . haptotylum and A . oligospora . The mariner and mariner ant1 TE families were more abundant in the genes of M . haptotylum than in those of A . oligospora ( Figure S2 ) . Duplicated genes were identified based on ortholog family assignment using orthoMCL [28] . Tandem duplications of the genes in families were identified in M . haptotylum: two or more genes adjacent in the genome belonging to the same family were considered to be tandem duplicated . In total , 147 duplicated pairs consisting of 272 genes ( 2 . 5% of all genes ) were identified as being located in tandem positions . A permutation test ( 1 , 000 permutations ) was performed by random reordering of the genes in the genome; this showed that tandemly duplicated genes were significantly more common than expected by chance ( P<0 . 001 ) . Only a few of the tandemly duplicated genes , 6 out of 147 , were located near ( <10 kb ) a transposon , indicating that the duplications were not a consequence of transposon activity . The proportion of genes encoding secreted proteins in M . haptotylum was estimated to be 15 . 2% ( 1 , 666 proteins ) , which is similar to that predicted in A . oligospora ( Table 1 ) . The proportions of secreted proteins in the two nematode-trapping fungi was comparable to that predicted for the insect pathogen Metarhizium anisopliae ( 13 . 2%; 1 , 394 proteins ) but higher than in the saprotroph Neurospora crassa ( 10 . 58%; 1 , 042 proteins ) and in the animal pathogen Aspergillus fumigatus ( 10 . 3%; 996 proteins ) . A BLAST analysis was conducted against the pathogen–host interaction protein database ( PHI-base ) [29] , which contain a collection of experimentally verified pathogenicity , virulence and effector genes from fungi , oomycetes and bacteria and enables computational identification of candidate pathogenicity genes . We identified 1 , 161 proteins in M . haptotylum ( 10 . 6% of all genes ) and 1 , 132 proteins in A . oligospora ( 9 . 9% ) that are similar to proteins in the PHI-base database , particularly from plant pathogens . The most abundant PHI-base genes in M . haptotylum and A . oligospora encoded transporters and proteins involved in signaling , oxidation , transcription regulation and metabolism ( Table S6 ) . The putative PHI-base proteins of M . haptotylum contained 555 Pfam domains , of which the most common are shown in Table S7 . A phylogenomic analysis using all single copy orthologs from 16 fungal species resolved the 14 ascomycetes into the three previously identified subphyla: Taphrinomycotina ( Schizosaccharomyces pombe ) , Saccharomycotina ( Candida albicans , Ashbya gossypii and Saccharomyces cerevisiae ) and Pezizomycotina [2] ( Figure 2A ) . Taphrinomycotina is resolved as the earliest diverging clade . Also in agreement with previous studies [2] , the nematode-trapping fungi of the Orbiliomycetes represent the earliest diverging clade of the Pezizomycotina , with the remaining species sampled forming a well-supported crown clade . Taking the date of the split between ascomycetes and basidiomycetes to be 500–650 million years ago ( MYA ) [30] , the date of the split between the Orbiliomycetes and crown clades were estimated to 400–520 MYA . The lineages of A . oligospora and M . haptotylum diverged about 14–18 MYA . Using the same calibration , the divergence between the entomopathogenic fungi M . anisopliae and Metarhizium acridum was estimated to 34–44 MYA , which is almost identical to the divergence time reported by Gao et al . [19] ( 33–43 MYA ) . The finding that the two nematode-trapping fungi are more closely related to each other than the two entomopathogens are to each other is unexpected , because the amino acid identity between pairs of orthologs for the nematode-trapping fungi ( 78 . 5% ) was lower than for the entomopathogenic fungi ( 91 . 5% ) [19] . A likely explanation for these results is that that the rate of amino acid substitutions is higher in the nematode-trapping fungi lineage than in the lineage of M . anisopliae and M . acridum . The estimation of the divergence time between A . oligospora and M . haptotylum differed substantially from the a previous study using five genes [21] . However , taxonomic assignment of one of the fossils used in the study has been questioned [22] . Lineage-specific gene expansion has been shown to be one of the most important means of adaptation in eukaryotes [31] . To study the gene family expansions and contractions in the genomes of M . haptotylum and A . oligospora , an analysis of gene family evolution comparing 16 genomes was performed ( Figure 2A ) . The software CAFE ( Computational Analysis of gene Family Evolution ) uses a maximum likelihood model to study gene family evolution while taking into account the phylogenetic relationships between the species . In total , 13 , 402 gene families , identified , using orthoMCL , were analyzed and the ancestral family sizes estimated with CAFE . In the branch leading to M . haptotylum , 848 gene families changed in size . Out of these , 326 ( 38 . 4% ) of the gene families were expanded and 522 ( 61 . 6% ) contracted ( Figure 2A; Figure S3 ) . Out of 806 gene families that changed in the branch leading to A . oligospora 362 ( 44 . 9% ) were expanded and 444 ( 55 . 1% ) were contracted families . The lowest proportion of expanded gene families among the filamentous ascomycetes was found in N . crassa branch ( 9% ) and the highest in M . anisopliae ( 52 . 6% ) . The proportion of expanded gene families of the nematode-trapping fungi was similar to M . oryzae ( 38 . 0% ) and higher than M . acridum ( 26 . 9% ) ( Figure 2A; Figure S3 ) . A principal component analysis ( PCA ) of the expanded orthoMCL gene families in filamentous fungal pathogens grouped the species according to which host they infect ( Figure 3A ) . The first axis ( explaining 49% of the variability ) separated the nematode-trapping fungi from those using other hosts including plants , insects or humans . All these species were from the crown clades of the Pezizomycotina . The second ( 21% ) and third axis ( 13% ) separated the species in this group into plant , insect and human pathogens . A statistical test was used to identify the gene families that contributed to the separation of the lifestyles of the fungi ( Figure 3B ) . The nematode-trapping fungi and plant pathogens shared the largest number of expanded gene families ( 28 families ) . The nematode-trapping fungi and insect pathogens shared fewer numbers of expanded gene families and the nematode and animal pathogens the least ( 9 and 2 families , respectively ) . Only two proteins encoded by genes in the gene families ( one shared with insect pathogens and one shared with plant pathogens ) above matched proteins in PHI-base [29] ( Table S8 ) . In addition , a separate analysis using Pfam domains was used to investigate protein domain families with functional annotation . Pfam domain families ( in contrast to orthoMCL gene families ) group proteins based on conserved functional protein domains . In total , 3 , 124 protein domain families ( containing 7 , 455 proteins ) in M . haptotylum and 3 , 782 protein domain families ( containing 7 , 555 proteins ) in A . oligospora were identified . 42 of the Pfam families were identified to be significantly larger or smaller in M . haptotylum than in other ascomycetes ( Table S9 ) . The 25 expanded protein domain families in M . haptotylum contained several peptidases , plant cell wall degrading enzymes and virulence factors of plant pathogenic fungi ( Figure 4 ) . Extracellular proteins and proteins involved in protein-protein interactions were also found among the significantly overrepresented gene families . Furthermore , a large fraction ( 19 out of 25 ) of the expanded Pfam families contained members that displayed significant similarities to proteins in the PHI-base ( Figure 4 ) . A comparison of the CAFE gene families and the expanded Pfam domain families revealed strong similarities . In total , 80 out of the 326 expanded orthoMCL gene families in M . haptotylum contained expanded Pfam domains ( 22 of the 25 expanded Pfam domains were detected ) . In A . oligospora , 68 of the 362 expanded gene families contained expanded Pfam domains ( 19 of 25 Pfam domains ) . Together , the expanded gene families in the two branches of M . haptotylum and A . oligospora match all 25 expanded Pfam domains , indicating that the expanded Pfam domain families are important for the evolution of the two nematode-trapping species . The genomes of both M . haptotylum and A . oligospora contained a low number of repetitive elements ( Table S10 ) . RIP generates mutations in repeat regions of a genome and has been reported in several fungi [12] , [15] , [19] , [27] , including A . oligospora [20] . In A . oligospora , the RIP index was calculated using the RIP indices and RIP index scan . We performed a genome-wide RIP index analysis of M . haptotylum and also , for comparison , of A . oligospora , to test whether the RIP mechanism had contributed to the low percentage of repetitive elements . Two different RIP indices were calculated [32] . The TpA/ApT index measures the products of the RIP mutations; a higher value suggests a stronger RIP response . This index corrects for false positives in AT-biased sequences and is suitable for detecting genes that have TpA point mutations due to RIP activity . It has successfully been used to identify RIP-affected genes that are located close to TEs [33] . The RIP index , ( CpA+TpG ) / ( ApC+GpT ) , which estimates the depletion of RIP targets in genes , is an alternative index commonly used [32] . The TpA/ApT index was calculated to be 1 . 12 in M . haptotylum and 1 . 28 in A . oligospora ( ≥0 . 89 indicates RIP mutations ) ( Figure S4 ) ; the ( CpA+TpG ) / ( ApC+GpT ) index was calculated to be 0 . 93 in M . haptotylum and 0 . 83 in A . oligospora ( ≤1 . 03 indicates RIP mutations ) . Both indices therefore indicate RIP activity in both genomes . A decrease in CpA dinucleotide abundance was detected in A . oligospora but not in M . haptotylum , indicating that the RIP mechanism may have different dinucleotide preferences between the two species ( Figure S5 ) . The RIP mechanism is activated when the fungus undergoes sexual reproduction [12] . Ascomycete fungi may be heterothallic ( outcrossing ) or homothallic ( self-crossing ) [34] . Two types of compatible MAT genes ( also called idiomorphs ) are required for successful crossing . The MAT1-1 idiomorph contains an alpha domain and the MAT1-2 idiomorph contains a HMG-box domain . Only one protein with weak match to a MAT_alpha1 domain in the Pfam database ( E-value 0 . 027 ) was detected in M . haptotylum ( H072_9577 ) and no such match was found in A . oligospora . The M . haptotylum protein H072_9577 matched A . oligospora hypothetical protein ( EGX48882; E-value 1E-15 ) using BLASTP . None of these proteins displayed significant sequence similarities to MAT1-1 proteins in the NCBI nr database ( BLASTP search ) . Consequently , no MAT1-1 idiomorph was identified in the genomes of M . haptotylum and A . oligospora . The MAT1-2 idiomorph was identified by searching for the HMG_box domain using Pfam database . In M . haptotylum 11 proteins were identified . Two of these ( H072_9933 and H072_9576 ) had MAT1-2 homologs in a BLASTP search . H072_9676 matched a hypothetical protein in A . oligospora ( EGX48881; E-value 2E-34 ) and a MAT1-2_1 protein from Peyronellaea pinodella ( AER26933; E-value 1E-09 ) . One homolog to H072_9933 was MAT1-2 protein variant 2 from Cercospora beticola ( AFH56912 , E-value 5E-09 ) . In A . oligospora 10 proteins containing the HMG_box domain were identified and 4 of these proteins ( EGX51833 , EGX50317 , EGX49798 and EGX49654 ) have homologs to Podospora anserina hypothetical S mat+ ( MAT1-2 idiomorph ) proteins in the NCBI database . In summary , in the sequenced strains of M . haptotylum and A . oligospora only one of the idiomorphs ( MAT1-2 containing the HMG_box domain ) was identified suggesting that both the M . haptotylum and A . oligospora species are heterothallic . Alternatively , the MAT1-1 idiomorph could be located in a region of the genomes that have not been fully sequenced . An active RIP mechanism would be expected to act on gene duplications and thereby gene families by causing rapid mutations leading to a reduction of the number of expanding families . For example , the strong RIP activity in N . crassa resulted in a very low percentage of expanded gene families [12] . A majority of the gene families in the M . haptotylum and A . oligospora branches are contracted ( 61 . 6% and 55 . 1% of families with altered sizes , respectively ) , indicating that an active RIP mechanism has reduced the number of expanded gene families ( Figure 2A; Figure S3 ) . However , because sequence divergence is not taken into account in the CAFE analysis , the nucleotide divergence between gene duplications was also investigated using Usearch software to identify recently duplicated genes [35] . RIP has been shown to act on duplicate sequences that are longer than 400 bp and at least 80% identical [12] . Among the 11 , 479 genes in A . oligospora , none were both longer than 400 bp and more than 80% identical . In M . haptotylum , we identified eight of of 10 , 959 gene models that were longer than 400 bp and had more than 80% identity . Five of these eight genes encode proteins with the Pfam domains related to transposons . RVP_2 ( retroviral aspartyl protease ) and RVT_1 ( reverse transcriptase ) were detected in three predicted proteins and two proteins contained the Transposase_5 domain . In conclusion , the paucity of highly similar gene duplicates strongly indicates an active RIP mechanism and shows that the expanded gene families contained genes that have diverged to the extent that they are no longer recognized by the RIP mechanism . Comparative genomic analysis showed that the two nematode-trapping fungi contained a large number of lineage-specific and species-specific genes ( Figure 2B ) . In total ∼62% of the genes in the genomes of M . haptotylum and A . oligospora were shared by other fungi; ∼20% of the genes were shared between the two nematode-trapping fungi and up to 16% of the genes were unique in each genome . There was a marked difference in the size and features of genes between the core and specific groups . For example , the average size of the core genes was more than twice as large as that of the species-specific genes ( the length of the encoded protein of the core , lineage-specific and species-specific genes in M . haptotylum were 583 , 455 and 281 amino acids , respectively , and in A . oligospora 571 , 452 and 275 amino acids , respectively ) . Moreover , the TEs were predominantly found among core genes rather than in the lineage-specific and species-specific genes in both species ( Figure S2 ) . A large proportion of the lineage-specific and species-specific genes had no homologs in the NCBI nr database and many lacked Pfam domains ( Figure 5 ) . In M . haptotylum , 71% of the species-specific genes lacked both homologs and Pfam domains , so they can be referred to as orphans [36] . In the A . oligospora genome , 76% of the species-specific proteins were orphans . Moreover , the lineage-specific and species-specific groups were enriched for secreted proteins , including small ( <300 amino acid ) secreted proteins ( SSPs ) ( Figure 5 ) . Analysis also showed that many of the abundant Pfam domains among the lineage-specific and species-specific genes were found among the expanded Pfam domain families ( Table S11 ) . All of the genes in the expanded orthoMCL families shared between the nematode-trapping fungi and plant- , insect- and animal pathogens ( Figure 3B ) were found in the core category . Three of these genes encode SSPs and all belonged to expanded orthoMCL families shared between nematode-trapping fungi and plant pathogenic fungi from the Pezizomycotina crown group . Analysis of KOG terms of the core , lineage-specific and species-specific genes showed that the biological functions were unequally distributed among these groups . The lineage-specific and species-specific groups were enriched for genes involved in transcription , cytoskeleton , cell wall/membrane/envelope biogenesis , secretion and signal transduction . In contrast , the core regions contained homologs of genes involved in metabolism and energy production ( Figure S5 ) . L1 larvae of the nematode C . briggsae were added to the infection structures ( knobs ) of M . haptotylum and the infection was followed by microscopy . After 4 hours of infection , approximately 98% of the added nematodes were paralyzed . This phase of the infection occurs at the time when the fungus has penetrated the nematode cuticle [37] . RNASeq analysis showed that ∼15% ( 1 , 653 ) of the significantly upregulated genes in M . haptotylum were more than two-fold up- or down-regulated during the penetration of C . briggsae ( Figure 6A ) . The distribution of these genes into the core , lineage-specific and species-specific categories were similar to that of the genome . However , a significant enrichment of species-specific genes was found in the cohort of the ten-fold upregulated transcripts . Species-specific genes that are highly upregulated as well as highly expressed during infection are likely to be involved in the adaptation to parasitism in M . haptotylum . Therefore , we identified the most highly upregulated and most highly expressed genes in M . haptotylum . Of the 10% most highly expressed genes during infection ( 1 , 069 genes ) , 117 were more than ten-fold upregulated during the infection . In total , 38 of these 117 infection-regulated genes were specific for the M . haptotylum lineage , 15 were unique for the M . haptotylum and A . oligospora lineage and 64 common to other fungi ( the core set ) . Furthermore , the cohort of the highly expressed infection-regulated genes was greatly enriched with secreted proteins ( Figure 6B ) . Secretion signals were predicted in 75 protein sequences , i . e . 64% of the infection-related genes , which is considerably higher than the proportion of secreted proteins in the whole proteome ( 15% ) ( Figure 6B ) . In total 36 of the 75 secreted proteins were short ( <300 amino acids ) and are thus considered to be SSPs . These proteins are rich in cysteine residues , with 15 of the 36 SSPs containing at least five cysteines , indicating that the secondary structure is likely to be highly stable in the environment outside of the fungal cell . Members of several expanded Pfam domain families were also enriched among the highly upregulated and expressed genes . In total , 36 of the 117 regulated genes ( 31% ) were from expanded Pfam domain families , which is a significantly higher fraction than found in the whole proteome ( 8% ) . The expanded Pfam domain families were primarily encoded by the core set of genes and they contained the DUF3129 , WSC , tyrosinase , mucin and peptidase S8 families ( Table 2 ) . Twenty-eight of the highly infection-regulated genes showed sequence similarities to proteins in the PHI-base , including the members of the DUF3129 Pfam domain family that encode the gas1 proteins of the rice blast fungus Magnaporthe oryzae [38] , an extracellular cutinase ( PBC1 ) from a plant pathogen [39] , a tetraspanin homolog from Colletotrichum lindemuthianum [40] , and RBT4 from C . albicans [41] . The function of the protein encoded by RBT4 is not known , but it contains a CAP ( Cysteine-rich secretory proteins , Antigen 5 and Pathogenesis-related 1 protein ) domain . None of the genes in the expanded orthoMCL families that were shared with the plant , insect or human pathogens were found among the highly upregulated and expressed transcripts during nematode infection . The transcriptome of M . haptotylum expressed during the early stage of infection of C . briggsae was compared with that expressed by A . oligospora during infection of C . briggsae . The initiation of the infection process of A . oligospora was slightly slower and less synchronized than that of M . haptotylum because the traps are fewer and hence more dispersed . Thus , to probe similar stages of the infection , the 4-hour samples of M . haptotylum were compared with the 6-hour and 10-hour samples of A . oligospora . Analysis of the sequences of the 10% most highly expressed genes in M . haptotylum ( which included 1 , 069 genes ) showed that these were slightly enriched with secreted proteins ( not statistically significant ) . Among them , 200 proteins ( 19% ) were predicted to be secreted proteins of which 84 were SSPs . The proportion of secreted proteins in the genome is 15% . Such an enrichment of secreted proteins was not observed in the transcriptome of A . oligospora . Seventy-two secreted proteins were identified among the 10% most highly expressed genes ( in total 1 , 145 genes ) of which 10 were SSPs . Genes from expanded Pfam domain families were highly expressed in both fungi , including the subtilisin ( peptidase S8 ) , tyrosinase , CFEM , DUF3129 ( gas1 ) , WSC and GLEYA families ( Table 2 ) . A more detailed analysis of the regulation of genes shared between the two nematode-trapping fungi was performed by comparing the expression levels of ortholog pairs , which represent genes that have most likely evolved from a common ancestral gene . In total , 2 , 599 ( 32% ) of the 8 , 121 ortholog pairs showed significant difference in fold change ( q-value<0 . 01 ) between the two fungi ( Figure 7A ) . Notably , the orthologs displaying the largest differences in expression values , that is , those that were more than 10-fold up- or down-regulated in the pairwise comparison between M . haptotylum and A . oligospora , were enriched with genes encoding secreted proteins , including many lineage-specific genes and SSPs ( Figure 7B and 7C ) . Several expanded Pfam domain families were also enriched among the differentially expressed genes ( Table 2 ) . Three of the expanded orthoMCL families ( Figure 3B ) were also found among the differentially expressed genes . Two of these families contained Pfam domains including NPP1 ( necrosis inducing protein ) and Polyketide_cyc2 ( polyketide cylcases/dehydrases ) . The plant pathogenic fungus Ustilago maydis and its close relative Sporisorium reilianum contain large clusters of genes encoding SSPs [42] , [43] . Despite the fact that many fungal genomes have now been published , little is known about the clustering of SSPs in other fungi . To investigate whether the SSPs of M . haptotylum were located in such gene clusters , we used the same criteria as Kämper et al . [42] to identify gene clusters of secreted proteins; groups of at least three adjacent genes encoding secreted proteins or groups containing more than three genes with at most one gene encoding a non-secreted protein in between . In total , 121 such gene clusters containing 453 secreted proteins were identified in M . haptotylum compared with only 12 containing 79 secreted proteins in U . maydis . Hence , 27 . 2% ( 453/1 , 666 ) of the genes encoding secreted proteins in M . haptotylum were located in gene clusters , and 18 . 6% ( 79/426 ) of genes encoding secreted proteins in U . maydis were located in gene clusters . However , gene clusters are substantially larger in U . maydis , ranging from 3 to 26 genes , whereas the gene clusters of M . haptotylum consist of 3–11 genes . We further investigated how many of the secreted proteins were SSPs . Unfortunately , the total number of SSPs was not reported in the genome paper of U . maydis [42] . In M . haptotylum 27 . 6% of all SSPs ( 192 out of the 695 in the whole genome ) were localized in 103 of the 121 gene clusters . In U . maydis , 42% ( 33 out of 79 ) of the secreted genes located in gene clusters were >10 fold upregulated in tumor tissue ( in planta ) compared with axenic culture ( without the host ) as measured by DNA array [42] . A significantly smaller fraction ( 8 . 2% , 34 out of 453 ) of the secreted genes that were >10 fold upregulated in M . haptotylum during early infection were located in clusters . Sixteen of the 34 secreted proteins in M . haptotylum were SSPs , and they were located in 13 clusters . The cluster containing the largest number of upregulated SSPs was cluster 74 ( Figure 8 ) . This cluster contained two species-specific SSPs and three secreted core proteins . Four of the five genes in the cluster encoded proteins belonging to the three expanded Pfam domain families , WSC , mucin and DUF3129 ( gas1 ) . The relationships between the SSPs and other secreted proteins located anywhere in the genome of M . haptotylum were examined using sequence homology clustering . The homologous sequences were not required to be part of any gene cluster . The genome contained 1 , 666 secreted proteins , including 695 SSPs . Using data from an all-against-all similarity search of the secreted proteins , 623 of the proteins were grouped into clusters containing three or more members ( Figure 9 ) . These clusters contained 181 SSPs . Out of the 181 clustered SSPs , 85 proteins were core , 60 lineage-specific and 36 species-specific proteins . Considering the total number of SSPs in these fractions , 35 . 2% of the SSPs in the core , 27 . 4% of the lineage-specific ones and only 15 . 3% of the species-specific ones were found in sequence homology clusters . The low numbers indicated that most SSPs do not form homology clusters , that is , they do not belong to any large group of paralogs . Notably , 23 of the 36 SSPs that were unique in M . haptotylum ( i . e . , they are found in the species-specific category ) were orphans . No large clusters of orphan SSP paralogs were detected either . In fact , the largest cluster of orphan SSPs consisted of only five proteins . These proteins contained 12 conserved cysteine residues , whereas most other positions showed large sequence divergence , suggesting rapid evolution among the orphan SSPs ( Figure S6 ) . Taken together , the results from the homology cluster analysis showed that sequences encoding the SSPs are highly divergent and rapidly evolving in M . haptotylum and A . oligospora . To investigate whether the RIP mechanism caused the rapid divergence of SSPs in M . haptotylum , a RIP analysis of the genes was performed separately from the RIP analysis of repeat-rich regions in the genome . The fraction of SSPs affected by RIP was calculated and compared with the fraction of all genes being affected by RIP . Analysis of the TpA/ApT index revealed that 38 . 2% of all M . haptotylum genes ( 4 , 181/10 , 959 ) have a value that is indicative of RIP activity . Using the same index , a significantly larger fraction , 76 . 7% of the genes encoding SSPs ( 533/695 ) , were affected by a RIP mechanism . The TpA/ApT index was used to identify RIP-affected genes in the vicinity of TEs [33] . The ( CpA+TpG ) / ( ApC+GpT ) index estimating the depletion of RIP targets in the genes indicated that 31 . 0% of all genes , and 33 . 5% of the SSPs were affected by RIP . The lower percentage of RIP-affected genes is probably a result of the differential preference of target dinucleotides in M . haptotylum ( Figure S4 ) , where CpA is less preferred , thereby lowering the index values of the ( CpA+TpG ) / ( ApC+GpT ) . In addition to generating rapid sequence divergence , the RIP mechanism may also introduce premature stop codons by mutating CpG into TpA . Thus , such a mechanism may lead to the formation of SSPs from genes encoding longer secreted proteins . The point mutations generated by the RIP mechanism may form the stop codons TAA and TAG , but not TGA because RIP forms TA mutations and this dinucleotide is absent in the TGA stop codon . Hence , RIP mutations would be expected to give an increased frequency of TAA and/or TAG stop codons than of TGA stop codons . Among the SSPs , the ratio of genes with the TAA versus the TGA stop codons was 2 . 3 ( 358/155 ) and the ratio of TAG/TGA was 1 . 1 ( 177/155 ) . The ratio of TAA/TGA and TAG/TGA in the genes ( in total 6 , 784 ) with no indication of RIP activity was 1 . 3 ( 2 , 631/1 , 994 ) and 1 . 1 ( 2 , 131/1 , 994 ) , respectively . Accordingly , the RIP mechanism has most likely contributed to the shortening of secreted proteins into SSPs by introducing the TAA stop codon .
The phylogenomic analyses presented in this study confirms previous gene-based phylogeny in placing the nematode-trapping fungi in the Orbiliomycetes as a basal branch among the filamentous growing ascomycetes ( i . e . Pezizomycotina ) [2] , [44] . The nematode-trapping fungi divergence from the other Pezizomycotina species were estimated to 400–520 MYA , which is similar to previous estimates of 419 MYA [21] . However , despite the similar divergence time of the Orbiliaceae and the Pezizomycotina , the estimation of the divergence time in our study between A . oligospora and M . haptotylum differed substantially from that of Yang et al . [21] . These authors estimated that the estimated divergence between species with adhesive knobs and those with adhesive nets occurred 198–208 MYA . The divergence of the nematode-trapping fungi by Yang et al . relied on fewer genes ( five ) for the phylogenetic reconstruction , compared with the 602 genes used in our study . More importantly , the fossil records used for the dating are different between the two studies . We used the split between ascomycetes and basidiomycetes , which has been estimated at 500–650 MYA [30]; Yang et al . [8] used two fossil records of carnivorous fungi , dated to 100 MYA [22] and 24 MYA [45] , respectively . Notably , the interpretation of the fossil record from 24 MYA has been questioned because of uncertainties in the identification of the trap structures and the assignment of the taxa [22] . In conclusion , our phylogenetic analysis using genome-wide analysis highlights the importance of reconstruction of phylogenetic trees using a large number of genes from fully sequenced genomes . On the genomic level , there are basically three compatible mechanisms that may account for the multiple emergence and adaptations to parasitic growth in fungi [46] . First , parasitism could result from the formation of novel genes , which could have a specific role during host infection and could be acquired by gene duplication or horizontal gene transfer . Second , it could result from differences in the regulation of gene expression . Third , it could result from gene loss and deletions . The comparative genomics and transcriptomics of the closely related M . haptotylum and A . oligospora presented here provide evidence that the first two mechanisms are of major importance for the adaptation to parasitism in nematode-trapping fungi of the Orbiliomycetes . The identified increase in protein domain family sizes and large numbers of unique genes , together with the result that many of these genes were highly expressed and regulated during infection , suggest that gene duplications followed by functional diversification has been an important mechanism underlying the evolution of parasitism in M . haptotylum and A . oligospora . Moreover , the observed differential expression of orthologs in the two fungi during the early stages of infection supports the hypothesis that gene expression in nematode-trapping fungi has evolved in response to interactions with the nematode host . The RIP mechanism has been shown to play a central role in the genome evolution of several filamentous ascomycetes belonging to the crown clade of Pezizomycotina [12] , [15] , [19] , [27] . Signs of RIP has been reported in the A . oligospora genome [20] . In this study , we have examined in detail the evidence for RIP and its impact on genome evolution in nematode-trapping fungi from the Orbiliomycetes . Five observations support an active RIP mechanism in these fungi . First , the genomes of the sampled nematode-trapping fungi have a lack of repeat regions . Second , the genomes have a bias in the dinucleotide frequencies consistent with point mutations generated by RIP [32] . Third , the genomes of the nematode-trapping fungi have a low proportion of closely related gene duplicates . Our analyses showed that the genes in the expanded gene families in M . haptotylum and A . oligospora have diverged to such an extent that their sequences are no longer recognized by the RIP mechanism . Fourth , the nematode-trapping fungi have reduced number of expanded gene families in the terminal branches . The impact of an active RIP mechanism on the pattern of gene family expansion can be revealed by comparing gene family evolution in M . anisopliae , which lacks RIP , and M . acridum , which has RIP [19] . Based on these differences in RIP activity , it can be expected that M . anisopliae should contain more expanded gene families than contracted , as compared with M . acridum . Indeed , the CAFE analysis detects these differences in gene family evolution between the two Metarhizium species . The proportion of expanded gene families in M . haptotylum and A . oligospora branches were more similar to that of M . oryzae and higher than that of M . acridum , indicating that the RIP mechanism in the two nematode-trapping fungi is similar in strength to that observed in M . oryzae [27] , but not as strong as detected in N . crassa [12] . Fifth , the strength of the RIP mechanism may be partly related to how often a particular species undergoes a sexual cycle [15] . Sexual stages have been detected only in a few species of nematode-trapping fungi like A . oligospora [47] . For the first time , we report on MAT genes in nematode-trapping fungi . The fact that the genomes of A . oligospora and M . haptotylum only contained one of the MAT idiomorphs strongly suggests that both species are heterothallic . The observation of an active RIP mechanism in the Orbiliomycetes , suggest that RIP evolved early in the evolution of filamentous ascomycetes , and that this mechanism has subsequently been lost in certain lineages such as those leading to M . anisopliae . The most striking example of gene duplications and rapid diversification of infection-expressed genes in the genomes of M . haptotylum and A . oligospora were found among the SSPs . The first SSP that was functionally characterized in detail was Pep1 in Ustilago maydis [48] . Pep1 is a secreted effector protein that suppresses the plant defense responses . The importance of SSPs for successful infection of the host has been reported for many fungi belonging to the Pezizomycotina crown group , including the insect pathogen Beauveria bassiana [49] and the plant pathogens Blumeria graminis f . sp . hordei [50] as well as fungi from other taxonomical groups , such as Melampsora larici-populina ( Basidiomycota ) and Puccinia graminis ( Basidiomycota ) [51] . Despite the rapid diversification of SSPs , expanded gene families that were shared between nematode-trapping fungi and pathogens from the Pezizomycotina crown group contained genes encoding SSPs . All these genes encoding SSPs belonged to families that were shared with plant pathogens . Interestingly , it appears that expansion of SSPs related to pathogenicity is an evolutionary strategy used primarily by insect and plant pathogens , whereas the genomes of animal pathogens ( with the exception of entomopathogens ) are not enriched for genes encoding SSPs [52] . Fungal gene clusters of genes encoding secreted proteins have been reported for U . maydis and its close relative Sporisorium reilianum [42] , [43] but have not been demonstrated in other pathogenic fungal genomes [15] , [53]–[55] . A large number of gene clusters was revealed in M . haptotylum . However , the size of the clusters ( number of genes ) and the proportion of genes encoding SSPs was lower than in U . maydis [42] . A large percentage of the SSP-encoding genes in U . maydis were co-regulated and highly expressed during infection . In contrast , most of the highly upregulated genes encoding SSPs in M . haptotylum were not located in the gene clusters . Hence , while the gene clustering appear to be important for secreted proteins in general , the organization of SSPs into gene clusters may be less important for their function during infection in M . haptotylum . One noteworthy exception was , however , gene cluster 74 . The genes in this gene cluster were all highly upregulated during infection and included virulence genes associated with adhesion ( mucin and WSC ) as well as DUF3129 ( gas1 ) , and hence the cluster provides an excellent candidate for future deletion experiment in M . haptotylum similar to the gene cluster deletions performed in U . maydis [42] . Two major mechanisms have been proposed for the origin of SSPs in fungal genomes [42] , [56] . Firstly , TEs may increase the frequency of genome rearrangements and thereby increase the rate of gene duplication . Transposons were shown to be very frequent in the genomes of Leptosphaeria maculans [57] , Mycosphaerella fijiensis and Cladosporium fulvum [33] , leading to a dramatically larger genome size . Enrichment of SSPs in repeat-rich regions was detected in L . maculans but not in M . fijiensis or C . fulvum , indicating that the prolific TEs may under certain conditions lead to increased numbers of SSPs . However , the paucity of transposons and the low abundance of repetitive elements suggest that TEs did not play a major role in the evolution of SSPs in the nematode-trapping fungi . The second mechanism leading to gene duplications is unequal crossover during meiosis [58] , [59] . We detected frequent tandem duplications of genes in the genome of M . haptotylum and a proportion of these genes were SSPs . Accordingly , we propose that unequal crossover is a major mechanism for the origin of SSP coding genes in M . haptotylum . In the genome of L . maculans , RIP activity has been shown to cause rapid divergence of SSP effectors located in close vicinity to TE-rich regions [57] . Despite the fact that such regions were not found in M . haptotylum , the genes encoding SSPs showed evidence of RIP activity , as revealed by a biased TpA/ApT index . The majority of the SSP sequences that were highly expressed and highly upregulated during infection have this RIP signature . In addition , we have shown that RIP can generate premature stop codons and thereby generate novel SSP coding genes from longer secreted proteins . Taking these results together , we propose the following model for the origin and diversification of SSPs in nematode-trapping fungi . Gene duplications through unequal crossover generated a large number of SSPs . Following such duplications , the SSPs underwent rapid diversification through the RIP mechanism . The lack of SSP paralogs in M . haptotylum can be explained by a relatively slow rate of gene duplication in combination with a rapid divergence of the duplicated genes through RIP . In contrast , fungi such as M . larici-populina with large families of SSP paralogs have a fast rate of gene duplications mediated through transposons and no detected RIP activity [51] . The proposed mechanism of origin and diversification of SSPs is general and may apply to any species with an active RIP mechanism . Evolution of SSPs by gene duplication without the involvement of TEs in combination with their diversification by RIP has previously not been described in pathogenic fungi . Several of the expanded Pfam domain families that were highly expressed during nematode infection have been identified in a previous study of M . haptotylum [26] . They include the subtilisins , the CFEM and the DUF3129 family ( gas1 ) . Many of the families containing upregulated genes contain genes that have been shown to be involved in the pathogenicity of animal and plant pathogenic fungi [29] . Aspartyl proteases are expressed by many pathogenic ascomycetes during infection . For example , C . albicans has at least 10 aspartyl proteases , which contribute to the infection by degrading the host cell surface , facilitating adhesion and degrading the host tissues for nutrition [60] . Tyrosinases are involved in the synthesis of melanin pigments [61] . The production of melanin has been shown to be essential in microbial pathogenesis , as it provides protection against host defense mechanisms and contributes to virulence in many animal and plant pathogenic fungi [62]–[65] . Tyrosinases can also oxidize protein- and peptide-bound tyrosyl residues , resulting in the formation of inter- and intramolecular crosslinks between peptides , proteins and carbohydrates [66] . It has been speculated that such enzymes could be involved in the adhesion mechanism of nematode-trapping fungi by cross-linking and stabilizing the proteins and polymers present in the extracellular layer mediating the binding between the trap and cuticle [10] . Other putative components of the extracellular adhesins might be found among the WSC and GLEYA Pfam domain families . Several genes encode proteins with features of known fungal adhesins , including signal peptide tandem repeats and predicted O-glycosylation sites [67] [68] . Despite the fact that there was a strong phylogenetic signal in the expansion of the orthoMCL gene families , our analysis showed that several of the expanded families in nematode-trapping fungi found in the basal clade of Pezizomycotina were also expanded in the plant , insect and animal parasitic fungi positioned in the crown clades of the subphylum . Some of these shared genes were slightly regulated in the knobs and mycelium of M . haptotylum and one encoded an SSP ( Table S8 ) . However , none of them were found among the highly upregulated and highly expressed transcripts during the early stages of nematode infection , which suggest that they are not of major importance for the interaction with the nematode hosts . It has been observed that nematode-trapping fungi display an extensive plasticity in the morphology and function of infection structures . For example , within one single species , namely A . oligospora , not only adhesive nets are formed but also appressoria in the rhizosphere of agricultural crops and hyphal coils around hyphae of other fungi [69]–[71] . It remains to be determined whether the genes in families shared between the nematode-trapping fungi and the other parasitic fungi are expressed and regulated in the interaction and colonization of non-nematode hosts , including plants . In conclusion , we have used comparative genomics and transcriptomics to decipher the genomic mechanisms leading to the evolution of parasitism in nematode-trapping fungi . We have identified two genomic mechanisms that are likely to have been of major importance during the evolution of parasitism in nematode-trapping fungi: the formation of novel genes through gene duplication and the differential regulation of existing genes identified by differential expression of orthologous genes . In M . haptotylum , we have identified enrichment of SSPs that were highly and differentially expressed during the infection . We propose an evolutionary mechanism for the duplication and divergence of these SSPs . The high expression of SSPs and other secreted proteins revealed a remarkable similarity in the infection mechanisms of nematode-trapping fungi , and plant and insect pathogenic fungi . In addition , the gene family analysis revealed that nematode-trapping fungi shared more expanded protein families with the plant pathogenic fungi compared to other pathogenic fungi such as insect and other animal pathogens .
Monacrosporium haptotylum ( strain CBS 200 . 50 ) was maintained on corn meal agar 1∶10 . Mycelium was grown in liquid medium ( soya peptone 0 . 5% w/v ) and incubated at room temperature on a shaker at 200 rpm for 7 days . The mycelium was harvested by filtering and then ground in liquid nitrogen . Genomic DNA was extracted using the Qiagen Plant Maxi Kit according to the manufacturer's instructions . DNA was precipitated with ethanol and dissolved in TE buffer . The whole genome of M . haptotylum was sequenced with 454 pyrosequencing technology using a titanium shotgun protocol ( XLR70 ) at KTH Stockholm and paired end sequencing of 3 kb insert libraries at the DNA Sequencing Facility at Lund University . The reads were assembled using the Newbler gsAssembler 2 . 3 software program ( Roche/454 Life Sciences ) . Gene models were predicted with the ab initio predictor GeneMark-ES [72] . Gene models that translated to peptides shorter than 48 amino acids were removed . The program Pfamscan was downloaded from the Sanger Centre FTP site ( ftp://ftp . sanger . ac . uk/pub/databases/Pfam/Tools/ ) together with Pfam databases and hidden Markov model ( HMM ) libraries [73] . The Pfamscan tool was locally installed , and the predicted gene models were scanned using the Pfamscan . pl with an E-value threshold of 0 . 05 . Predicted protein sequences were also mapped to the KOG classification system [74] , [75] . Transfer RNAs were predicted by the Aragorn [76] and tRNASCAN-SE [77] programs using default settings . For comparison of the tRNAs with the M . haptotylum we also predicted the tRNA from the A . oligospora genome . Prediction of signal peptide sequences were performed using the SignalP 3 . 0 [78] . The genomes of A . oligospora , M . anisopliae , A . fumigatus and N . crassa were reanalyzed to identify signal peptides . Virulence-related genes were identified by BLASTP [79] similarity searches against the PHI-base database version 3 . 2 [29] using a cutoff of <1E-10 . Core , lineage-specific and species-specific proteins were identified by BLASTP ( cutoff <1E-10 ) similarity search against M . haptotylum , A . oligospora and other genomes ( Figure 2B ) using low complexity filtering . Low complexity proteins were not classified into core lineage- or species-specific proteins . Orphans were sequences that have no matches in the Pfam database and no homologs to sequences in other organisms as revealed by BLASTP searches against the NCBI nr database ( cutoff <1E-10 ) . Detection of MAT genes in M . haptotylum was performed by identifying proteins containing the Pfam domains MAT_alpha1 and HMG-box . Homologs were identified using BLASTP against NCBI nr database and the results were manually investigated . Percentage nucleotide identity was calculated using Usearch 6 . 037 [35] with cluster_fast option and an identity of 80% . Results from BLASTP ( cutoff <1E-10 ) were used as input for clustering the sequences with orthoMCL version 2 . 02 [28] . The orthoMCL clusters were used for identifying gene families . Tandem duplicated genes in these families were manually identified . A custom Perl script was used to perform 1 , 000 permutations of the gene order to generate random sets with the same gene family distribution . For the phylogenomic analysis , only families containing exactly one gene copy for each of the 16 genomes were used , because families with paralogs may hinder correct phylogenetic inference . In total , 602 families containing 9 , 632 orthologs were identified and the proteins aligned using ClustalW [80] . The sequences were concatenated and poorly aligned regions of each alignment were removed using Gblocks [81] . The trimmed alignment was subsequently used for phylogenetic reconstruction using maximum likelihood method with the Dayhoff amino acid substitution model implemented in PhyML 3 . 0 [82] with 1 , 000 bootstrap replicates at the BioPortal computer cluster , Oslo University . BEAST 1 . 7 . 0 [83] was used to estimate the divergence time between A . oligospora and M . haptotylum using a Yule tree prior and lognormal relaxed clock rates . Divergence times was calculated by using the estimated splitting time between the Ascomycota and Basidiomycota branches ( 500 and 650 MYA ) as a calibration point [30] . An ultrametric tree based on the PhyML tree was generated using Mesquite 2 . 75 [84] . Gene family expansions and contractions were estimated with the CAFE ( Computational Analysis of gene Family Evolution ) software version 3 . 0 [85] using the ultrametric tree and the orthoMCL gene families as input . The tree and expansion/contraction data were displayed using the iTOL web tool [86] . To rescale the data prior PCA and cluster analysis , orthoMCL families were log2 ( gene counts+1 ) transformed . The PCA plot with variance filtering 0 . 2 and heat maps using group categories ( F-test ) with variance filtering <0 . 2 and q-value <0 . 05 ( false discovery rate , adjusted for multiple testing ) were generated using the Qlucore Explorer version 2 . 2 ( Qlucore AB , Lund , Sweden ) . Repetitive elements in the genome assembly were analyzed using Repeat Masker ( version 3 . 2 . 8 ) using cross match an implementation of the Smith Waterman-Gotoh algorithm [87] , Tandem Repeats Finder [88] and Repeat Scout [89] . RIP index was determined with the software RIPCAL by reference against the non-repetitive control families [32] . Stop codons for all genes were calculated . Putative TEs were identified by the Transposon-PSI [90] a program that performs tBLASTn searches using a set of position-specific scoring matrices ( PSSMs ) specific for different TE families . The genome positions of the transposons were evaluated for identification of clusters of transposons computationally as well as visually using IGV viewer version 2 . 2 [91] . Clusters were defined as at least five transposons located on the same contig in the genome . To validate the prediction of gene models in M . haptotylum , the transcriptome expressed by the fungus when grown on liquid medium was sequenced using the 454 technology . The mycelium was grown in aerated liquid medium as previously described [24] . The mycelium was quickly collected by filtering and directly dropped into a clean mortar filled with liquid N2 and homogenized using a pestle . The resulting powder was collected into 50 ml Falcon tubes and stored at −80°C until use . Total RNA was isolated using the RNeasy Plant Mini Kit ( Qiagen ) according to the manufacturer's instructions , and using the RLC buffer . Total RNA was eluted in either 60 or 100 µl of H2O and stored at −20°C until use . For quality and concentration assessments all samples were analyzed using a 2100 Bioanalyzer and the RNA 6000 Nano kits ( Agilent ) . After total RNA purification , mRNA was isolated from each sample using approximately 100 µg and the absolutely mRNA Purification Kit ( Agilent ) . The purified mRNA was then used as starting material in the cDNA Library Preparation protocol ( GS FLX Titanium Series ) provided by 454/Roche . In total the sequencing generated 422 , 883 reads that were mapped to the genome using gsMapper 2 . 6 ( Roche ) . In order to compare the transcriptome with the gene models and to avoid mismatches due to reads located in the untranslated regions ( not predicted by GeneMark ) , the reads from the pyrosequencing were assembled using gsAssembler 2 . 6 and subsequently mapped to the predicted gene models using BLASTN [79] . In addition , previously generated EST sequences [25] , [26] were mapped to the genes as well as the genome using BLASTN . The transcriptome expressed during infection of nematodes by M . haptotylum and A . oligospora was examined using the nematode C . briggsae ( strain A16 ) as a host . Eggs were obtained by treating C . briggsae with 1% NaOCl in 580 mM NaOH for 5 minutes and washed with water . Hatched L1 larvae were collected the next day . Knobs of M . haptotylum were isolated and were incubated at room temperature with the L1 larvae on water agar plates as previously described [26] . The infection was followed under a light microscope and infected nematodes were collected after 4 h of incubation . Duplicate samples of knobs and knobs infecting C . briggsae were analyzed . Infection experiments with A . oligospora ( strain ATCC 24927 ) were performed by growing the fungus on dialysis membranes placed on the surface of low nutrient mineral salt ( LNM ) medium . Traps ( nets ) were induced by adding Panagrellus redivivus nematodes [92] . After several days when the nematodes were completely digested and the traps formed , approximately 75–100 L1 larvae of C . briggsae were added to the dialysis membranes . The infection was followed in a light microscope and fungal mycelia , traps and infected nematodes were collected after 6 and 10 h of infection . Total RNA was extracted and sent for sequencing at GATC Biotech AG , Konstanz , Germany . After reverse transcription into double-stranded cDNA for tag preparation according to the massively parallel signature sequencing protocol [93] it was sequenced using HiSeq2000 ( Illumina Inc . ) in single read mode with a length of 50 bp . More than 87 million reads were sequenced for the A . oligospora samples and 82 million reads for the M . haptotylum samples . The reads were mapped to the corresponding M . haptotylum and A . oligospora genome as well as to C . briggsae using the Burrows Wheeler Aligner ( BWA ) software with default settings [94] . The transcript abundances were normalized and significantly differentially expressed genes ( q-value ) were identified using the R package DESeq [95] . Two replicates were used for each fungus; for M . haptotylum two samples after 4 hours of infection and for A . oligospora after 6 and 10 hours of infection ( 1 sample each ) . Although replicates for each time point in A . oligospora would be preferred , using two different time points as replicates , as done in this study , increase the stringency of the analysis since any variation between A . oligospora time points will increase the variance within the sample and thereby decrease the number of significantly differentialy expressed orthologs between the two species . This approach is similar to the approach suggested by in the DESeq manual . Bidirectional best BLAST hits ( BLASTP cutoff 1E-10 ) were used to identify pairs of proteins ( orthologs ) for comparative transcriptome analyses . Gene clusters were identified in the genome of M . haptotylum using a procedure described by Kämper et al . [42] . A homology-based clustering of secreted proteins was performed by using an all-against-all BLASTP [79] similarity search ( cutoff <1E-10 ) . The network visualization program Biolayout Express 3D version 2 . 2 [96] was used to analyze the sequence homology clusters containing at least three secreted proteins . The whole genome sequencing and transcriptome data of M . haptotylum have been deposited at DDBJ/EMBL/GenBank under Bioproject PRJNA186729 and the A . oligospora transcriptome data under BioProject PRJNA196395 . This Whole Genome Shotgun project has been deposited at DDBJ/EMBL/GenBank under the accession AQGS00000000 . The version described in this paper is the first version , AQGS01000000 .
|
Nematode-trapping fungi are a group of soil-living fungi found worldwide . They capture and kill nematodes and are used as biocontrol agents against parasitic nematodes . The infection structures differ morphologically . Their traps can be classified into several main groups , including adhesive knobs , adhesive nets , adhesive branches and constricting rings . We have sequenced the genome of Monacrosporium haptotylum , a knob-forming fungus , and compared it with the genome of Arthrobotrys oligospora , a net-forming fungus . Comparative and functional genomics analyses of the predicted functions of the two species' genes provide new insights into how the nematode-trapping lifestyle has evolved . Two genomic mechanisms are likely to have been responsible for the adaptation to this lifestyle: the formation of new genes through gene duplications , including gene family expansions , and the differential gene expression of orthologous genes in the two fungi . We identified a large number of genes that were found only in these two species ( either shared by both or found just in one or other genome ) ; many of these were highly expressed and differentially regulated during nematode infection . Our results suggest that the function and expression patterns of these genes have evolved in response to interactions with nematode hosts .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
Genomic Mechanisms Accounting for the Adaptation to Parasitism in Nematode-Trapping Fungi
|
Epidermal growth factor receptor variant III ( EGFRvIII ) has been associated with glioma stemness , but the direct molecular mechanism linking the two is largely unknown . Here , we show that EGFRvIII induces the expression and secretion of pigment epithelium-derived factor ( PEDF ) via activation of signal transducer and activator of transcription 3 ( STAT3 ) , thereby promoting self-renewal and tumor progression of glioma stem cells ( GSCs ) . Mechanistically , PEDF sustained GSC self-renewal by Notch1 cleavage , and the generated intracellular domain of Notch1 ( NICD ) induced the expression of Sox2 through interaction with its promoter region . Furthermore , a subpopulation with high levels of PEDF was capable of infiltration along corpus callosum . Inhibition of PEDF diminished GSC self-renewal and increased survival of orthotopic tumor-bearing mice . Together , these data indicate the novel role of PEDF as a key regulator of GSC and suggest clinical implications .
Glioblastoma multiforme ( GBM ) is the most aggressive malignant primary brain tumor [1] . Despite multimodal treatment with surgery , radiotherapy , and chemotherapy , the prognosis of GBM is poor , with a median overall survival of 14 mo and 2-y survival rates of less than 10% [2] . Failure of GBM treatment is attributed in part to the widespread infiltration of tumor cells into the normal brain parenchyma , leading to inevitable tumor recurrence , as well as GBM’s resistance to standard therapeutics [3 , 4] . Emerging evidence suggests that glioma stem cells ( GSCs ) might contribute to multiple aspects of GBM tumor biology , including the initiation , progression , diffusive infiltration , recurrence , and drug resistance of glioma [5 , 6] . The xenograft models of GSCs recapitulate clinical features of glioma infiltration , such as migration along white-matter tracts , perivascular spread , and subpial growth [7–10] . GSCs isolated from human tumors show remarkable similarities to neural stem cells ( NSCs ) as GSCs express markers for neural stem/progenitors , such as Nestin and Sox2 , and harness the ability to grow as nonadherent spheres when cultured in serum-free conditions containing the defined growth factors [7 , 11] . Upon serum induction , such GSCs differentiate into cells of neuronal or glial lineages and lose stemness as well as tumorigenicity [12–14] . Similarly , transient exposure of GSCs to bone morphogenetic protein 4 ( BMP4 ) , a well-known differentiation factor , abolishes the tumor initiating and infiltrating potential [15–17] . Moreover , primary GBM cells that are enriched with GSCs , but not the traditional glioma lines grown in standard serum-containing culture conditions , closely mirror the genotype of parental tumors and yield tumors with a highly infiltrative phenotype when orthotopically implanted into immunodeficient mice [7] . These studies suggest that tumor initiation and the infiltrative phenotype of glioma cells are associated with stemness . EGFRvIII , a frequently occurring mutation in primary glioblastoma , results in a protein that is unable to bind any known ligand . Although controversial , EGFRvIII expression in patients has been associated with poor prognosis as well as resistance to radiotherapy and chemotherapy [18 , 19] . Despite the loss of ligand-binding ability , EGFRvIII is known to transmit a low level of constitutive signaling leading to the activation of pro-oncogenic signaling molecules such as AKT , extracellular signal-regulated kinases ( ERK ) , and STATs in GBM and breast cancers [20–24] . Intriguingly , expression of EGFRvIII positively correlates with the expression of stem/progenitor markers , including Nestin , Sox2 , and CD133 , and is associated with an enhanced ability to self-renew and initiate tumor [25] . As EGFR signaling is one of the most well-known therapeutic targets and autocrine signaling has increasingly been implicated in the regulation of stem cell self-renewal and tumorigenicity of various malignancies , including gliomas [26–29] , we tested the possibility that autocrine signaling in GSCs plays a part in the regulation of the self-renewal property of EGFRvIII+ infiltrative GSCs . Here , we show that EGFRvIII contributes to the self-renewal and tumor-initiating ability of GSCs in part via inducing PEDF , an autocrine factor that has been shown to be expressed in the NSC niche .
To investigate the possible role of EGFRvIII in the regulation of GSC stemness , we first classified EGFRvIII+ or EGFRvIII- cells based on the results from semiquantitative reverse transcription polymerase chain reaction ( RT-PCR ) . EGFRvIII transcript was detected in ten out of 13 GSCs ( Fig 1A ) . In two of such EGFRvIII+ cells , CSC2 and X01 , we examined whether EGFRvIII expression was modulated during differentiation of GSCs . Upon the induction of differentiation by serum , CSC2 and X01 cells revealed down-regulation of Sox2 and Nestin , markers for undifferentiated cells , with concurrent increase of glial fibrillary acidic protein ( GFAP ) , a marker for differentiation ( Fig 1B–1E , S1A and S1B Fig ) . Importantly , the level of EGFRvIII was gradually reduced after serum treatment and finally became undetectable by day 9 ( Fig 1B ) . In sharp contrast , expression levels of the EGFR wild type ( EGFR-WT ) in these two cells were up-regulated over time ( Fig 1B ) . These results suggest that EGFRvIII might be associated with GSC maintenance . To investigate the functional role of EGFRvIII in the maintenance of GSCs , we selectively inhibited EGFRvIII by expressing small interfering RNA ( siRNA ) against EGFRvIII ( siEGFRvIII ) and assessed for changes in the expression of stemness markers and the ability to form spheres . EGFRvIII knockdown significantly reduced the levels of Sox2 and Nestin while increasing GFAP level in CSC2 cells ( Fig 1F–1H and S1C Fig ) . However , knockdown of EGFR-WT by specific siRNA only had minimal effects on the expression levels of Nestin and Sox2 ( Fig 2F ) . By performing limiting dilution assays ( LDA ) , we confirmed that knocking down EGFRvIII inhibited the ability to form spheres ( Fig 1I ) . When we overexpressed EGFRvIII in X02 cells , which do not normally harbor EGFRvIII ( Fig 1A ) , the levels of Sox2 and Nestin increased , whereas that of GFAP decreased ( Fig 1J–1L and S1D Fig ) . Moreover , the ability to form spheres of X02 cells markedly increased , as assessed by LDA ( Fig 1M ) . Collectively , these results support the notion that EGFRvIII regulates the ability of GSCs to self-renew . When we collected conditioned media ( CM ) from GSC cultures grown in serum-free conditions ( serum-free GSC-CM ) and added the CM to CSC2 cultures , sphere formation of CSC2 cells was enhanced . By contrast , sphere formation was drastically prevented when CM from CSC2 cultures grown in the presence of serum ( serum-CM ) was added ( Fig 2A ) . These results suggest that soluble factors in serum-free GSC CM , secreted by GSCs under the control of EGFRvIII , might potentiate sphere formation and glioma stemness . To identify secreted factors that might regulate EGFRvIII-dependent glioma stemness , we used LC-MS/MS to compare the secreted proteins between serum-free GSC-CM and serum-CM of CSC2 cells and between serum-free GSC-CM from CSC2 ( EGFRvIII+ GSC ) and Ex Vivo ( EGFRvIII- GSC ) ( Fig 2B ) . We extracted commonly enriched proteins in serum-free GSC-CM but not in others . Among those proteins , PEDF was the only secretory protein . Based on these results , we chose PEDF as the strongest candidate ligand for the maintenance of GSCs ( S1 Table ) . PEDF has been identified as a secretory protein which is implicated as a niche factor of NSCs in the subventricular zone ( SVZ ) [30] . Using three different GSCs ( CSC2 , X01 , and X02 ) , we confirmed that serum induced the expression of an astrocytic differentiation marker GFAP but decreased the levels of PEDF and NSC markers , Nestin and Sox2 ( Fig 2C ) . The fact that PEDF-silenced GSC-CM lost sphere forming ability in CSC2 cells ( Fig 2D ) further raises the possibility of PEDF as a stemness factor . When EGFRvIII was depleted in CSC2 and X01 cells , PEDF expression was greatly reduced ( Fig 2E ) , suggesting PEDF as a possible downstream factor of EGFRvIII signaling . Conversely , depletion or overexpression of EGFR-WT in CSC2 and X01 cells did not affect PEDF expression ( Fig 2F and 2G ) . Moreover , we found that EGFRvIII , but not EGFR-WT , increased the expression levels of PEDF and the markers for NSC stemness in X02 ( Fig 2H ) . These results strongly support the possibility of PEDF as downstream mediator of EGFRvIII-induced stemness of GSCs . STAT3 signaling has been shown to play a crucial role in controlling the stemness of GSCs induced by EGFRvIII [31 , 32] . To further elucidate the functional involvement of STAT3 in EGFRvIII-induced PEDF expression , we modulated EGFR expression in GSCs and analyzed the activation of STAT3 . Silencing of EGFRvIII decreased total STAT3 expression and blocked STAT3 activation and PEDF expression ( Fig 2E ) , whereas silencing of EGFR-WT did not ( Fig 2F ) . When STAT3 signaling was blocked by either a small-molecule inhibitor of STAT3 activation ( Stattic , 5 μM ) or a STAT3-specific siRNA , we observed decreased expression of PEDF ( Fig 2I–2K ) . To verify possible involvement of other downstream signaling of EGFRvIII , we further examined AKT and ERK activation after EGFRvIII silencing ( S2A Fig ) and overexpression ( S2B Fig ) . Although there were slight effects of siEGFRvIII on AKT and ERK1/2 phosphorylation , EGFRvIII overexpression could not induce AKT and ERK1/2 phosphorylation in GSCs . Furthermore , inhibitor treatment for ERK and AKT activation did not show any change in PEDF expression ( S2C Fig ) . Therefore , we conclude that EGFRvIII/STAT3 signaling might be a major pathway for PEDF expression in GSCs . Taken together , these data suggest that PEDF is a soluble factor secreted by GSCs and that PEDF expression and subsequent secretion is regulated by EGFRvIII through STAT3 signaling . To determine the involvement of PEDF in maintaining glioma stemness , we treated X02 cells with recombinant PEDF in serum-free GSC medium , performed sphere-forming assays , and examined the levels of stem cell markers . Our results revealed that recombinant PEDF significantly and dose-dependently increased sphere formation ( S3A Fig ) , increased NSC markers expression , and decreased the level of GFAP in these X02 cells ( Fig 3A and 3B ) . In X01 cells cultured in serum-free GSC medium , withdrawal of growth factors ( EGF and basic fibroblast growth factor [bFGF] ) from the medium caused a reduction in sphere formation ( Fig 3C ) . In these cells , the levels of PEDF and Nestin and Sox2 were decreased , while GFAP expression was elevated ( Fig 3D ) . Importantly , addition of recombinant PEDF into the serum-free GSC medium lacking EGF and bFGF restored sphere formation ( Fig 3C ) and induced Nestin and Sox2 expression in X01 , X04 , and X06 cells ( Fig 3D and S3B Fig ) . These results indicate that extracellular PEDF plays an important role in the induction and/or the maintenance of the self-renewal property of GSCs . Next , we altered the levels of PEDF in various GSCs and assessed their sphere-forming ability and expression of stemness markers . In CSC2 and X01 cells , which express high levels of PEDF under normal culture conditions ( see Fig 2 ) , PEDF knockdown prevented sphere formation and reduced the levels of Nestin and Sox2 while increasing GFAP expression ( Fig 4A–4D , S4A–S4D Fig ) . Since we designed short hairpin RNA of PEDF ( shPEDF ) to target the 3ʹUTR region , we conducted a rescue experiment by overexpressing a PEDF construct that does not contain the 3ʹUTR . As we expected , overexpression of PEDF rescued the glioma stemness and sphere-forming ability of PEDF-silenced GSCs ( Fig 4A–4D ) . Conversely , in X02 cells , PEDF overexpression promoted sphere formation and induced the expression of Nestin and Sox2 , while GFAP expression was decreased ( Fig 4E and 4F ) . As seen in Fig 1J–1M , overexpression of EGFRvIII in X02 cells promoted sphere formation and induced the expression of Nestin and Sox2 . Importantly , we found that knocking down PEDF in X02 cells overexpressing EGFRvIII completely prevented sphere formation , down-regulated the NSC markers ( Nestin and Sox2 ) , and up-regulated GFAP ( Fig 4G and 4H ) . These results further support the notion that PEDF promotes self-renewal of GSCs and that PEDF function is regulated by the EGFRvIII-STAT3 axis . PEDF was originally identified as an antiangiogenic factor [33] . Subsequent studies have shown pleotropic effects mediated by PEDF , suggesting that PEDF is regulated in cell context-dependent manners . A previous study has suggested that PEDF plays a role in the neurovascular niches to control stem-cell maintenance through activation of Notch signaling [30] . We also found that treatment of GSCs with recombinant PEDF led to the generation of the cleavage product of Notch-1 , Notch-1 intracellular domain ( NICD ) ( S5A Fig ) , accompanied by up-regulation of Notch target genes , such as Hes1 and Hey1 ( S5B Fig ) . Similarly , overexpression and knockdown of PEDF resulted in the accumulation and depletion of NICD , respectively ( Fig 5A and 5B ) . Importantly , pharmacological inhibition of the γ-secretase activity by N-[N- ( 3 , 5-difluorophenacetyl ) -l-alanyl]-S-phenylglycine t-butyl ester ( DAPT ) essentially abolished the effects of PEDF on NICD generation ( Fig 5C ) and sphere formation ( Fig 5D ) . These results suggest that PEDF activates Notch signaling , which is required for self-renewal in GSCs . To further establish the link between PEDF signaling and GSC stemness , we examined the downstream effectors of NICD in GSCs . A previous study has shown that NICD induces the expression of Sox2 in NSCs [34] . In GSCs , we detected an increase in Sox2 expression in response to PEDF , an effect that was blocked by γ-secretase inhibition ( Fig 5C ) . As Sox2 promoter contains two putative CBF1-binding sites ( TGGGAA ) in the -1 kb region of the transcriptional start site [35] and the mRNA level of Sox2 correlated with PEDF expression ( Fig 5E ) , we performed chromatin immunoprecipitation ( ChIP ) experiments to examine whether Sox2 gene was directly regulated by NICD . As shown in Fig 5F , antibodies against NICD were able to immunoprecipitate specific regions of the Sox2 gene , covering the -1 . 007 and -0 . 894 kb region from the transcriptional start site ( TSS ) . Next , we investigated whether Sox2 overexpression in GSCs could restore the reduced ability to self-renew induced by PEDF depletion . We found that the changes in Nestin and GFAP expression induced by PEDF depletion were reversed by Sox2 overexpression ( Fig 5G ) . The ability to form spheres was also enhanced by Sox2 overexpression ( Fig 5H ) , indicating that Sox2 is a direct target of Notch and that Sox2 regulates self-renewal of GSCs . Similarly , overexpression of Sox2 in X02 ( EGFRvIII-/PEDFlow ) cells significantly changed the expression levels of Nestin and GFAP ( S5C Fig ) and increased sphere-forming ability ( S5D Fig ) . As our data indicate the association between EGFRvIII , STAT3 , and PEDF , we examined the protein expression of these molecules in 13 primary glioma cells . Among the 13 GSCs in which we examined the level of EGFRvIII ( see Fig 1 ) , we confirmed that those ten GSCs expressed EGFRvIII proteins by immunoblot analysis ( Fig 6A ) . We found that the level of PEDF was higher in CSC2 , X01 , X03 , X04 , X06 , X08 , and X09 cells as compared to that in X02 , Ex Vivo , and 528NS cells . Among the 13 GSCs , three GSCs , MD30 , 1123NS and 83NS , expressed EGFRvIII , but there was no STAT3 phosphorylation . Consistent with the notion that STAT3 signaling is crucial for PEDF expression , these cells did not express PEDF . EGFRvIII expression was also confirmed by using EGFRvIII specific antibody ( Fig 6A ) . To assess the tumorigenic role of PEDF in GSCs , we orthotopically injected 13 different GSC lines into nude mice . As shown in Fig 6B , EGFRvIII+/PEDFhigh GSCs ( CSC2 , X01 , X03 , X04 , X06 , X08 , and X09 ) formed brain tumors and exhibited highly infiltrative phenotypes ( Fig 6B and S6 Fig ) . By contrast , X02 cells lacking EGFRvIII failed to induce brain tumor formation within 2 mo . These infiltrating tumor cells showed activation of EGFRvIII/STAT3 signaling and expression of stemness markers ( Nestin and Sox2 ) in vivo ( Fig 6B ) . In contrast , Ex Vivo and 528NS cells induced the formation of tumors , but these tumors were noninfiltrative ( Fig 6B ) . Importantly , the EGFRvIII+/PEDFlow GSCs ( MD30 , 1123NS , and 83NS ) induced tumor formation , but these tumors again were noninfiltrative , highlighting that only EGFRvIII+/PEDFhigh cells are capable of inducing infiltration ( Fig 6B ) . However , Ki-67 and Nestin expression did not significantly differ among the groups of GSCs , indicating that both groups exhibited similar mitotic activity and glioma stemness ( S7A and S7B Fig ) . Next , we orthotopically implanted GFP-labeled X01 control or X01-PEDF-KD GSCs into nude mice . Control GSCs showed an infiltration phenotype , displaying invasion through the corpus callosum . In contrast , depletion of PEDF in GSCs significantly inhibited infiltration ( Fig 7A ) . We then compared survival rates of mice by intracranial injection . Depletion of PEDF in GSCs significantly increased mice survival rate compared to the control GSCs ( 104 cells ) , and this finding was more evident when smaller number of cells were injected ( 103 cells ) ( Fig 7B and 7C ) . H&E staining showed that control GSCs formed tumor mass at the injected site and infiltrated to the other half of the brain through the corpus callosum ( Fig 7D and 7E ) . Depletion of PEDF significantly decreased expression levels of Sox2 , NICD , and Hes1 and caused GSCs to form tumor mass that was more restricted to the injection site ( Fig 7D ) . To further analyze the potential gain of function , we overexpressed PEDF in noninfiltrating GSCs ( MD30 , 1123NS , 83NS , and Ex Vivo ) . Overexpression of PEDF increased tumor size and infiltrative phenotype ( Fig 7F and S8 Fig ) . However , infiltrative phenotype was only observed in Ex Vivo . PEDF-overexpressing Ex Vivo showed local infiltration from injection site ( Fig 7Fb and 7Fd ) , myelin-associated infiltration in the corpus callosum ( Fig 7Ff ) , and tumor growth at a distance from the injection site ( Fig 7Fh ) compared with control ( Fig 7Fa , 7Fc , 7Fe and 7Fg ) . Since several molecules were previously suggested as candidate receptors for PEDF , we confirmed the presence of those molecules in our GSCs [36–38] . As shown in S9A Fig , PNPLA2 , LRP6 , and PLXDC1 were expressed in all of the tested GSCs , whereas PLXDC2 was not . Specific siRNA knockdown of the individual PEDF receptors failed to affect the stemness , differentiation marker expression , and sphere formation ability of our GSCs ( S9B–S9D Fig ) . These results suggest another receptor for PEDF might regulate the stemness and tumor progression of GSCs . To address the functional association between EGFRvIII and PEDF in vivo , we orthotopically injected X02 cells overexpressing EGFRvIII into nude mice . Injection of X02 cells overexpressing EGFRvIII significantly reduced survival rate as compared to mice injected with control X02 cells . Knocking down PEDF in X02 cells overexpressing EGFRvIII significantly increased survival rates similar to X02 control ( Fig 7G ) . These results suggest that EGFRvIII/PEDF signaling plays an important role in the tumorigenicity and infiltration of GSCs . Given the role of PEDF in controlling the ability of GSCs to self-renew and infiltrate , we analyzed the possible relationship between PEDF expression and prognosis in glioma patients using the REMBRANDT ( REpository for Molecular BRAin Neoplasia DaTa ) dataset . In all glioma , we found that 3-fold down-regulation of PEDF expression ( 81 out of 254 patients ) correlated with a better survival rate ( Fig 8A; p < 0 . 001 ) , and low expression of PEDF showed increased survival rate in patients with GBMs and astrocytoma ( Fig 8B and 8C; p < 0 . 05 ) . Furthermore , in glioblastoma patient samples , we found that the levels of EGFRvIII , PEDF , p-STAT3 , and NICD proteins were highly correlated ( Fig 8D and 8E ) . In conclusion , we reveal that EGFRvIII induces the secretion of PEDF , which , in turn , activates Notch signaling that elevates Sox2 expression , leading to self-renewal and infiltration of GSCs . These results implicate that EGFRvIII-PEDF-Notch-Sox2 signaling is a crucial regulatory node that contributes to the tumor propagation and invasion , and they additionally provide a basis for future therapeutic approaches against highly infiltrative gliomas ( Fig 8F ) .
Human gliomas are notorious for their diffuse infiltration into normal brain tissues [39] . Many reports have suggested that GSCs are involved in this process [7 , 9 , 40] , but the cellular identity of the infiltrating glioma cells and the mechanisms responsible for maintaining their stemness remain unclear . Here , we demonstrate that EGFRvIII+/PEDFhigh GSCs are responsible for glioma infiltration , and that PEDF , an autocrine factor secreted by GSCs , promotes self-renewal and tumorigenic and infiltrative properties of GSCs . Moreover , we show that PEDF maintains glioma stemness and self-renewal ability by activating Notch/Sox2 signaling , and silencing of PEDF decreases the infiltration of GSCs and increases the survival of tumor-bearing mice . These results collectively suggest that inhibition of the PEDF signaling pathway may provide a novel therapeutic strategy for combating the infiltration of GSCs . EGFRvIII , the most frequently occurring EGFR mutation in primary glioblastoma , encodes a protein product that constitutively signals regardless of EGF ligand . Previous studies showed that 24%–67% of GBM samples harbored EGFRvIII mutations , whereas this mutation was absent from normal tissues [41 , 42] . EGFRvIII signaling has been shown to be important in driving tumor progression and is often correlated with poor prognosis [43] . Glioma cells expressing EGFRvIII were found to recruit EGFR-WT-expressing cells and accelerate glioma tumorigenicity by up-regulating the expression levels of cytokines ( e . g . , interleukin 6 [IL6] and/or leukemia inhibitory factor ) [44] . EGFRvIII expression has been reported to be sporadic or focal to the cancer area , and EGFRvIII+ cells have been shown to drive tumorigenesis by regulating other cell types [45] . Moreover , recent studies showed that EGFRvIII expression correlates with GSC markers ( e . g . , CD133 ) and confers resistance to EGFR-targeted therapy [25 , 46] . These previous findings collectively suggest that EGFRvIII might have functional significance in the regulation of glioma stemness , but the molecular mechanisms linking the EGFRvIII and stemness or infiltration are not clearly understood . In the present study , we found that EGFRvIII expression correlates with GSC stemness during differentiation . Moreover , the loss of EGFRvIII was associated with losses of self-renewal and GSC markers , whereas EGFRvIII overexpression increased tumor formation and decreased survival rates in an orthotopic xenograft model . Our results strongly support the notion that EGFRvIII expression can drive glioma progression and infiltration by increasing GSCs . Recently , EGFRvIII vaccination has been suggested as a promising therapeutic option for EGFRvIII-mutation-bearing patients , with vaccinated patients showing significant increases in overall and progression-free survival ( PFS ) [47] . The remaining challenges will include efforts to increase response rate and overcome acquired resistance . Interestingly , samples of recurring tumors did not show any significant expression of EGFRvIII , suggesting the existence of a possible resistance mechanism against EGFRvIII vaccination . In the present paper , we demonstrate that PEDF signaling acts as a downstream regulator of EGFRvIII-induced glioma stemness and that PEDF expression can functionally replace EGFRvIII . In the future , it will be exciting to address whether PEDF expression contributes to resistance against the EGFRvIII vaccination . Like NSCs , it has been suggested that the GSCs reside in a niche microenvironment including perivascular and hypoxic locations and remain in a stem cell-like status [48–50] . The niche microenvironment generates extrinsic factors that maintain stemness and direct stem cell behavior [51–53] . Various soluble factors that are involved in neuronal development and known to regulate NSC self-renewal ( e . g . , Sonic hedgehog , Wingless-type proteins , and fibroblast growth factor ) [51 , 54] have also been implicated as a critical inducers of glioma stemness and tumorigenicity [6] . In the adult brain , NSC migration is limited: neuroblasts migrate to the olfactory bulb , and two stemness regions ( the SVZ and dentate gyrus [DG] of the hippocampus ) are involved in very limited migrations of adult NSCs [55–57] . In contrast , many studies have found that GSCs show extensive infiltrations into the corpus callosum , cortex , subpial space , and meninges without any extracellular stimulus in orthotopic xenografts [7 , 9 , 40] . This suggests that GSCs might have the ability to maintain self-renewal and stemness in a non-niche environment . In this study , we identify PEDF as an autocrine factor from GSCs that is expressed in response to EGFRvIII and STAT3 phosphorylation . Silencing of PEDF clearly decreased infiltrating GSCs along the corpus callosum . Multiple functions of PEDF have been discovered: it has been demonstrated [33 , 58 , 59] to be a survival factor against oxidative stress [60] , a regulator of immune-cell migration [61] , and a possible self-renewal factor in NSCs [30] . Furthermore , functions of PEDF in tumor were known to include acting as an antitumorigenic agent by blocking angiogenesis [62] . In contrast to the known function of PEDF in tumor cells , we herein demonstrated that PEDF is secreted from GSCs and promotes self-renewal activity as an autocrine factor in infiltrating GSCs . Furthermore , we manifested that knockdown of PEDF expression did not changed microvessels in in vivo tumor tissue ( S10 Fig ) . Therefore , our results suggest PEDF as a novel tumorigenic factor that could act as a key regulator of self-niche for the infiltrating GSCs . There are many similarities in the growth characteristics and gene expression profiles of NSCs and GSCs , suggesting that similar signaling pathways could be required for their survival and growth . Notch signaling is known to promote the self-renewal ability of NSCs and to inhibit differentiation of NSCs [63] . The Notch signaling cascade was elevated in GSCs and found to regulate GSC self-renewal and tumorigenicity [64 , 65] . Importantly , Notch signaling promotes the radioresistance of GSCs , whereas inhibition of Notch signaling depletes GSCs and tumorigenicity [66 , 67] . Our present results suggest that PEDF regulates Notch signaling and is involved in GSC self-renewal . Furthermore , we observed a very tight positive correlation between the protein expression levels of PEDF and NICD in our GBM patient samples , suggesting that Notch signaling may be important to the progression of glioblastoma . Activated NICD directly regulates transcription of the Sox2 gene , which suggests that ( at least in GSCs ) PEDF promotes Notch cleavage and strengthens its transcriptional effects . Sox2 was previously shown to be a direct target of Notch and to regulate the self-renewal and maintenance of NSCs [34] . It is the first report , to our knowledge , to show that PEDF regulates the Notch-Sox2 signaling axis involved in GSC self-renewal and maintenance . Sox2 is an essential driver of stem-like populations in multiple malignancies , and recent papers have suggested that Sox2 is a member of a core set of neurodevelopmental transcription factors ( TFs ) that are essential for GBM propagation and can reprogram differentiated GBM cells into “induced” GSCs [68] . In our xenograft model , the infiltrative phenotypes of the GSCs were strongly correlated with EGFRvIII expression . EGFRvIII is known to drive glioma infiltration in EGFRvIII overexpressing mice , wherein tumor cells infiltrate along white matter tracts ( e . g . , the corpus callosum ) and the perivascular space [69] . Consistent with these results , we found that EGFRvIII+ GSCs showed high-level infiltration along the corpus callosum and the perivascular space . However , not all of the EGFRvIII+ GSCs were found to be infiltrative in the present work . Three out of ten GSC lines expressed EGFRvIII but failed to form any infiltrative glioma . These GSCs did not express PEDF , suggesting that it might be a key downstream factor for EGFRvIII-dependent glioma infiltration . Consistent with this notion , overexpression of EGFRvIII in EGFRvIII- GSCs increased tumor formation and decreased survival rates in our xenograft mode , and this phenomenon was rescued by down-regulation of PEDF . Furthermore , overexpression of PEDF in noninfiltrative EGFRvIII- GSCs ( Ex Vivo ) confers infiltrative nature in the brain parenchyma . Collectively , these data indicate that EGFRvIII-regulated PEDF increases the tumorigenicity and self-renewal capacity of infiltrative GSCs . We speculate that the differences in PEDF expression between the generated EGFRvIII+ cells might arise from variations in downstream signaling . STAT3 has been suggested to be one of the most important downstream signaling partners of EGFRvIII in tumorigenesis , and many studies have implicated EGFRvIII-STAT3 signaling in the progression of glioblastoma [31 , 32] . Here , we found that seven of the ten EGFRvIII+ GSCs ( CSC2 , X01 , X03 , X04 , X06 , X08 , and X09 ) activated STAT3 and further promoted the self-renewal capacity of GSCs through autocrine secretion of PEDF . In contrast , the remaining three EGFRvIII+ GSCs ( MD30 , 1123NS and 83NS ) did not show STAT3 activation or glioma infiltration . These results suggest that EGFRvIII-induced PEDF expression is mediated by the activation of STAT3 and that EGFRvIII/STAT3/PEDF signaling regulates the self-renewal of infiltrative GSCs . One of the most important features of CSCs is the tumorigenic potential that arises from a small number of cells . Fewer than 100 GSCs reportedly initiated tumor formation in an in vivo xenograft model representing an original patient phenotype , whereas non-GSCs ( even at 105 cells ) failed to cause tumor formation in the same mouse model [70] . In the present study , we show that silencing of PEDF significantly increased survival and decreased GSC infiltration in a mouse model . The requirement of PEDF for tumorigenesis was greater in 103 cell-injected xenografts compared to those injected with 104 cells . These results indicate that PEDF specifically regulates the ability of a small number of GSCs to initiate tumors in the mouse brain . Together , our findings strongly suggest that PEDF may be a crucial therapeutic target , as well as an indicator for the tumorigenicity of infiltrative GSCs . As expected , these results are consistent with REMBRANDT glioma patient survival data . Since grade of glioma infiltration starts from low grade [World Health Organization ( WHO ) grade II] , the level of PEDF correlated from grade II glioma to grade IV GBM . Taken together , these findings suggest that PEDF may be an indicator of infiltrative GSCs and a prognostic marker of low grade glioma and that it could be a crucial therapeutic target for the future treatment of glioma . In conclusion , we herein demonstrate that EGFRvIII promotes PEDF secretion , thereby activating Notch signaling and triggering the regulation of Sox2 expression . Our observations suggest that EGFRvIII-induced self-niche formation regulates the self-renewal and infiltrative ability of GSCs and offer PEDF as a candidate therapeutic target for infiltrating glioma . Further studies will be needed to identify the relevant functional domains of PEDF and determine its extracellular binding partner ( s ) during the regulation of GSCs .
The work with animals reported in this study was conducted in accordance with protocols approved by the Institutional Animal Care and Use Committee at the National Cancer Center , Republic of Korea . 293T cells were maintained in Dulbecco’s modified Eagle’s medium ( DMEM ) supplemented with 10% fetal bovine serum ( HyClone ) . All GSCs were cultured in DMEM/F-12 supplemented with B27 ( Invitrogen ) , EGF ( 10 ng/ml , R&D Systems ) , and bFGF ( 5 ng/ml , R&D Systems ) . Differentiation of GSCs cells was maintained in DMEM/F-12 supplemented with 10% fetal bovine serum . 293T in 100-mm plates were transfected with 6 . 67 μg of pLenti6/V5-PEDF , pLL-EGFR-WT , pLL-EGFRvIII , pLL3 . 7-shPEDF1 , pLL3 . 7-shPEDF2 vector , 3 . 33 μg of VSV-G plasmid DNA , and 5 ug of packaging viral CMV delta 8 . 9 plasmid using Lipofectamine 2000 ( Invitrogen ) . The medium was changed 6 h after transfection . The medium containing lentivirus was harvested at 48 h after transfection . Viral particles were concentrated and purified using a Lenti-X concentrator . Cells were infected with lentivirus in the presence of 6 μg/ml polybrene . Small interference RNA against human EGFRvIII , EGFRvIII , EGFR-WT , STAT3 , PNPLA2 , PLXDC1 , LRP6 , and negative control siRNA ( Bioneer ) were transfected in GSCs using Lipofectamine 2000 ( Invitrogen ) . Nucleotide sequences used for target-specific siRNA or shRNA are shown in the following: anti-EGFRvIII siRNA , 5ʹ-CUGGAGGAAAAGAAAGGUAAU-3′ [71]; anti-EGFR-WT siRNA , 5′-GGAAAUAUGUACUACGAAA-3′; anti-STAT3 siRNA , 5′-GCUCCAGCAUCUGCUGCUUC-3′; anti-PNPLA2 siRNA , 5′-GUUCAUUGAGGUAUCUAAAUU-3′; anti-PLXDC1 siRNA , 5′-GUCUUGUAACCAUGAAACAUU-3′; anti-LRP6 siRNA , 5′-GCAGAUAUCAGACGAAUUU-3′; shPEDF1 , 5′-GGTTTCAATGCATACAATAAA-3′; and shPEDF2 , 5′-CGAGTTCATTCATGACATAGA -3′ . Semiquantitative RT-PCR was performed to determine mRNA levels . Total RNA was isolated from cells using TRIzol reagent ( Invitrogen ) according to the manufacturer’s instructions . Total RNA ( 1 μg ) was used as a template to synthesize cDNA using M-MLV reverse transcriptase ( Invitrogen ) . The PCR primers are shown in the following: PEDF , sense 5′-AACCTTACAGGGGCAGCCTT-3′ and antisense 5′-TGAGGGACACAGACACAGGG-3′; GFAP , sense 5′-TCTCTCGGAGTATCTGGGAACTG-3′ and antisense 5′-TTCCCTTTCCTGTCTGAGTCTCA-3′; Nestin , sense 5′-CCAGAGACTTCAGGGTTTC-3′ and antisense 5′-AGAGTGTTCAGCATTATGCC-3′; Sox2 , sense 5′-AACCCCAAGATGCACAACTC-3′ and antisense 5′-CGGGGCCGGTATTTATAATC-3′; EGFRvIII , sense 5′-ATGCGACCCTCCGGGACG-3′ and antisense 5′-ATCTGTCACCACATAATTACCT-3′; EGFR-WT , sense 5′-AACTGTGAGGTGGTCCTTGG-3′ and antisense 5′-AGCTCCTTCAGTCCGGTTTT-3′; and GAPDH , sense 5′-GGAGTCCACTGGCGTCTTCAC-3′ and antisense 5′-GAGGCATTGCTGATGATCTTGAGG-3′ . The PCR products were analyzed on the 1% agarose gel . Serum-free GSC cultured CSC2 , Ex Vivo , and serum-cultured CSC2 were washed three times with PBS to remove all growth factors , supplements , and serum residues . After 24 h incubation with DMEM/F12 , conditioned media for each experimental CM were collected . Floating cells and cellular debris were removed by centrifugation ( 1 , 300 rpm , 5 min , at 4°C ) . Acetone precipitated protein were run on SDS-PAGE gel ( NuPAGE Novex 4%–12% Bis-Tris gel , Invitrogen , Carlsbad , California ) , followed by staining with Colloidal Blue staining kit ( Invitrogen ) . SDS-PAGE gel was sliced into eight pieces for in-gel tryptic digestion , according to the manufacturer's instructions using in-gel tryptic digestion kit ( Thermo Fisher Scientific , Rockford , Illinois ) . Briefly , the excised gels were destained , reduced by TCEP ( Tris[2-carboxyethyl] phosphine ) , and alkylated by idoacetamide ( IAA ) . The alkylated gel pieces were dehydrated in 100% ACN and digested with MS grade trypsin in 25 mM NH4CO3 for 12 h at 30°C . Digested peptides were evaporated from the liquid using vacuum concentrator and cleaned up using C18 spin columns ( Thermo Fisher Scientific ) for MS analysis . The tryptic digested peptides were analyzed by a Q Exactive hybrid quadrupole-orbitrap mass spectrometer ( Thermo Fisher Scientific ) coupled with an Ultimate 3000 RSLCnano system ( Thermo Fisher Scientific ) . The tryptic peptides were loaded onto trap column ( 100 μm x 2 cm ) packed with Acclaim PepMap100 C18 resin in which loaded peptides were eluted with a linear gradient from 5% to 30% solvent B ( 0 . 1% formic acid in ACN ) for 120 min at a flow rate 300 nL/min . The eluted peptides , separated by the analytical column ( 75 μm x 15 cm ) , were sprayed into nano-ESI source with an electrospray voltage of 2 . 4 kV . The Q Exactive Orbitrap mass analyzer was operated in a top 10 data-dependent method . Full MS scans were acquired over the range m/z 300–2000 with mass resolution of 70 , 000 ( at m/z 200 ) . The AGC target value was 1 . 00E + 06 . The ten most intense peaks with charge state ≥2 were fragmented in the higher-energy collisional dissociation ( HCD ) collision cell with normalized collision energy of 25% , and tandem mass spectra were acquired in the Orbitrap mass analyzer with a mass resolution of 17 , 500 at m/z 200 . For in vitro limiting dilution assay , GSCs with decreasing numbers of cells ( 200 , 100 , 50 , 25 , 12 , 6 , and 1 ) or ( 100 , 50 , 25 , and 5 ) per well plated in 96-well plates containing DMEM/F-12 with B27 , EGF ( 10 ng/ml ) , and bFGF ( 5 ng/ml ) were used . Extreme limiting dilution analysis was performed using software available at http://bioinf . wehi . edu . au/software/elda/ . Sphere formation assays were also performed with 1 , 000 cells per well plate in 12 well plates and incubated in a humidified atmosphere with 5% CO2 at 37°C . Fourteen days later , plates were examined for sphere formation using an inverted microscope . The spheres with diameter >100 μm were then counted . Protein was extracted with RIPA buffer with complete protease inhibitors ( Roche ) , separated by electrophoresis , transferred to PVDF Membrane ( Millipore ) , and blocked with 5% skim milk ( BD ) . The primary antibodies , EGFR ( 1005 ) ( Santa Cruz ) , p-EGFR ( Tyr 1173 ) ( Santa Cruz ) , Sox2 ( R&D systems ) , Nestin ( BD ) , GFAP ( ImmunO ) , NICD ( Cell signaling ) , Hes1 ( Millipore ) , Hey1 ( abcam ) , PEDF ( Upstate ) , and β-actin ( Santa Cruz ) were incubated overnight at 4°C . Immunoreactive bands were visualized using peroxidase-labeled affinity purified secondary antibodies ( KPL ) and the detection reagent Amersham ECL prime western blotting detection reagent ( GE Healthcare ) . Approximately 4x106 cells ( X02-con , X02-PEDF ) were used per ChIP reaction after crosslinking with 1% formaldehyde for 10 min at room temperature . ChIP was performed with NICD antibody . The associated DNA after purification was subjected to qRT-PCR to detect the probable in vivo binding of NICD protein to specific DNA sequences within the Sox2 promoter . The primer sequences are as follows: forward primer 5′-CTGGAGTCCTGGGAACTCTG-3′ and reverse primer 5′-TCTACTGTCTGCCCCCACTC-3′ . Antibody against IgG was used as a nonspecific control . Cells were orthotopically transplanted following washing and resuspension in PBS . Cells were injected stereotactically into the left striatum of 6-wk-old female Balb/c nude mice . The injection coordinates were 2 . 2 mm to the left of the midline and 0 . 2 mm posterior to the bregma at a depth of 3 . 5 mm . The brain of each mouse was harvested and fixed in 4% paraformaldehyde . To allow observation of histologic features , mice were anesthetized with isoflurane and euthanized by transcardial perfusion with 10 ml of PBS , followed by 10 ml of 4% paraformaldehyde solution . The brains were removed , fixed with 4% paraformaldehyde for 24 h at 4°C , and stained with hematoxylin ( DaKo ) and 0 . 25% eosin ( Merck ) . For immunohistochemical staining of neural stem cell markers ( Nestin , Abcam ) , after the antigen retrieval process with citrate buffer ( pH 6 . 0 ) and endogenous peroxidase blocking with 3% hydrogen peroxide , tissue sections were incubated in 1% BSA blocking solution ( vol/vol ) for 0 . 5 h at room temperature and then in primary antibody overnight at 4°C in a humidified chamber . To decrease nonspecific Nestin signals in mouse tissue , we used the Mouse on Mouse Elite Peroxidase kit ( Vector Laboratories ) and developed samples using 3 , 3ʹ-diaminobenzidine ( DAB , Vector Laboratories ) as chromogen . For immunocytochemistry , GSCs were seeded in bovine fibronectin ( 10 μg/ml in PBS ) coated chamber slide with the complete medium of GSCs . After 24 h of incubation , cells were fixed with 4% paraformaldehyde for 20 min at 4°C and washed three times with PBS at room temperature . Cells were then incubated in blocking solution ( 5% BSA and 0 . 5% Triton X-100 in PBS ) for 1 h at room temperature . Cells were stained with primary antibodies in blocking solution ( 1:100 ) for 2 h at 4°C and washed three times with PBS . Staining was visualized using Alexa Fluor 488 goat antirabbit and Alexa Fluor 594 goat antimouse ( Invitrogen ) secondary antibodies ( 1:1000 ) in dark condition for 1 h at 4°C and washed three times with PBS . Nuclei were stained using 4ʹ , 6-diamidino-2-phenylindole ( DAPI ) ( contained mounting solution ) , and stained cells were viewed under a confocal laser scanning microscope . MRI analysis was performed and images were acquired using a 7 . 0 T magnet ( BioSpin , Bruker , Germany ) . After localizer imaging on three orthogonal axes , T2-weighted images of the entire mouse brain were acquired using a Rapid Acquisition with Refocused Echoes ( RARE ) sequence with TR and TE set to 2500 and 35 ms , respectively . Other parameters used were a 2-cm field of view and a 256 x 256 matrix in four averages , resulting in a total scan time of 4 min . Snap-frozen brain tumor tissues were pulverized in liquid nitrogen frozen mortar and extracted with RIPA buffer with complete protease inhibitors ( Roche ) . Patients’ survival data grouped by PEDF expression levels in all glioma , GBM , and astrocytoma were obtained from the REMBRANT database of the National Cancer Institute . ( REMBRANDT data portal will be retired on or after June 1 , 2015 . All data currently hosted in REMBRANDT , including microarray gene expression , copy number , and clinical data , has been migrated to the Georgetown Database of Cancer [GDOC] , a knowledge discovery platform that will allow continued support for the community's efforts to mine these data . ) Kaplan-Meier survival plots were analyzed by Statistical Package for the Social Sciences software version 12 . 0 ( SPSS , Chicago , Illinois , United States ) . Results of the multidataset experiments were compared by analysis of variance using the Statistical Package for the Social Sciences software version 12 . 0 ( SPSS , Chicago , Illinois , US ) . Results of the two-dataset experiments were compared using the two-tailed Student’s t test . The level of statistical significance stated in the text was based on the p-values . * p < 0 . 05 or ** p < 0 . 01 was considered statistically significant .
|
Malignant gliomas are among the most lethal types of cancer , due in part to the stem-cell-like characteristics and invasive properties of the brain tumor cells . However , little is known about the underlying molecular mechanisms that govern such processes . Here , we identify pigment epithelium-derived factor ( PEDF ) as a critical factor controlling stemness and tumor progression in glioma stem cells . We found that PEDF is secreted from glioblastoma expressing EGFRvIII , a frequently occurring mutation in primary glioblastoma that yields a permanently activated epidermal growth factor receptor . We delineate an EGFRvIII-STAT3-PEDF signaling axis as a signature profile of highly malignant gliomas , which promotes self-renewal of glioma stem cells . Our results demonstrate a previously unprecedented function of PEDF and implicate potential therapeutic approaches against malignant gliomas .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Pigment Epithelium-Derived Factor (PEDF) Expression Induced by EGFRvIII Promotes Self-renewal and Tumor Progression of Glioma Stem Cells
|
Estrogens act by binding to estrogen receptors alpha and beta ( ERα , ERβ ) , ligand-dependent transcription factors that play crucial roles in sex differentiation , tumor growth and cardiovascular physiology . Estrogens also activate the G protein-coupled estrogen receptor ( GPER ) , however the function of GPER in vivo is less well understood . Here we find that GPER is required for normal heart rate in zebrafish embryos . Acute exposure to estrogens increased heart rate in wildtype and in ERα and ERβ mutant embryos but not in GPER mutants . GPER mutant embryos exhibited reduced basal heart rate , while heart rate was normal in ERα and ERβ mutants . We detected gper transcript in discrete regions of the brain and pituitary but not in the heart , suggesting that GPER acts centrally to regulate heart rate . In the pituitary , we observed gper expression in cells that regulate levels of thyroid hormone triiodothyronine ( T3 ) , a hormone known to increase heart rate . Compared to wild type , GPER mutants had reduced levels of T3 and estrogens , suggesting pituitary abnormalities . Exposure to exogenous T3 , but not estradiol , rescued the reduced heart rate phenotype in gper mutant embryos , demonstrating that T3 acts downstream of GPER to regulate heart rate . Using genetic and mass spectrometry approaches , we find that GPER regulates maternal estrogen levels , which are required for normal embryonic heart rate . Our results demonstrate that estradiol plays a previously unappreciated role in the acute modulation of heart rate during zebrafish embryonic development and suggest that GPER regulates embryonic heart rate by altering maternal estrogen levels and embryonic T3 levels .
Zebrafish are an established model for human cardiovascular development and function [1] with conserved estrogen signaling [2–4] . While studying the function of ERα ( esr1 ) in zebrafish embryonic heart valves [5 , 6] , we serendipitously observed that estrogen receptor modulators caused acute changes in heart rate . Estrogens bind two classes of receptors: nuclear hormone receptors ( ERα , ERβ ) that are ligand-dependent transcription factors [7] , and the G protein-coupled estrogen receptor ( GPER , also known as GPR30 ) , an integral membrane protein [8 , 9] . It has been difficult to tease apart to what degree ERα and/or ERβ are involved in regulating GPER function in vivo . The observations that ERα can directly activate G proteins in cultured cells [10–13] and that GPER coimmunoprecipitated with ERα in tumor cells [14] has been used to argue that either GPER is dispensable for estrogen-dependent signaling or that GPER mediates interactions between ERα and G proteins [15] . Studies using GPER-deficient mice implicate GPER in ventricular hypertrophy [16] , regulation of blood pressure and vascular tone [17 , 18] and atherosclerosis progression [19] , but whether nuclear ER signaling is required for GPER function in these contexts is unknown . Additionally , these studies examined GPER function in adult animals , while the role of GPER during embryonic development is not well understood . Here we use zebrafish embryos , an established model of human development , to reveal a new function for GPER during cardiovascular development . Estrogen signaling often differs between males and females . However , zebrafish embryos and larvae are bipotential hermaphrodites that have not begun to sexually differentiate before approximately 10 days post fertilization ( dpf ) [20] , meaning that estrogen levels are uniform between age-matched embryos . Additionally , zebrafish embryos develop outside of the mother and not within a confined space , such as the uterus . Therefore , zebrafish embryos are not subject to local estrogen concentration gradients , as has been reported to occur in rodents depending upon their position in utero and their proximity to embryos of the same or opposite sex [21 , 22] . These developmental traits make zebrafish a powerful model to study how sex hormone signaling influences the formation and function of non-gonadal tissues . Using complementary genetic and pharmacologic approaches , we sought to characterize how estradiol regulates heart rate and determine to what extent each estrogen receptor mediates estradiol-dependent changes in heart rate in zebrafish embryos .
We exposed 49 hour post fertilization ( hpf ) embryos to 17β-estradiol ( E2 ) and assayed heart rate following one hour exposure . We found that E2 exposure caused an approximately 20% increase in heart rate ( Fig 1 , mean difference in heart rate between estradiol and vehicle exposed embryos 26 . 51 ± 10 . 63 ( standard deviation ) beats per minute ( bpm ) ; see S1 Table for heart rate values for each embryo in this and subsequent experiments ) . Exposure to progesterone , a structurally similar steroid sex hormone , had no effect on heart rate ( Fig 1 , mean difference in heart rate 1 . 57 ± 1 . 71 bpm ) , suggesting that the effects on heart rate were specific to estrogens . Our results are consistent with previous results in cultured cells demonstrating that progesterone binds GPER with less than 0 . 01% binding affinity compared to E2 [23] . However , we cannot exclude the possibility that GPER could respond to higher concentrations of progesterone in vivo . To explore whether heart rate was influenced by nuclear estrogen receptor or GPER signaling pathways , we employed a pharmacological approach . We exposed embryos to ICI182 , 780 ( fulvestrant ) , a well-characterized ERα and ERβ antagonist [24] that also acts as a GPER agonist [8] . Following one hour exposure to ICI182 , 780 , heart rate was significantly increased ( Fig 1 , mean difference in heart rate 25 . 17 ± 9 . 24 bpm ) . This effect was blocked by co-administration of G36 , a specific GPER antagonist [25] ( Fig 1 , mean difference in heart rate 2 . 26 ± 4 . 09 bpm ) , suggesting that estradiol increases heart rate via GPER . We also exposed embryos to G1 , a specific GPER agonist with no detectable agonist activity against nuclear estrogen receptors [26] , and found that heart rate increased significantly ( Fig 1 , mean difference in heart rate 26 . 57 ± 15 . 89 bpm ) . This effect was partially blocked by co-administration of G36 ( Fig 1 , mean difference in heart rate 10 . 95 ± 3 . 35 bpm ) . Together , our pharmacological results suggest that GPER regulates heart rate acutely . Curiously , the GPER inhibitor G36 by itself did not change heart rate significantly ( Fig 1 , mean difference in heart rate 5 . 00 ± 2 . 26 bpm ) . Additionally , while G36 partially blocked ICI182 , 780- and G1-dependent increases in heart rate , G36 failed to inhibit estradiol-dependent increase in heart rate ( Fig 1 , mean difference in heart rate 39 . 75 ± 2 . 87 bpm ) . This could be due to off-target effects , a limitation of pharmacologic approaches , or due to differences in how G36 , which was designed to inhibit human GPER , binds to zebrafish GPER . To definitively test the hypothesis that estradiol regulates heart rate via GPER , we generated gper mutant embryos , exposed them to estrogen receptor modulators and assayed heart rate . Using CRISPR-Cas technology [27] , we generated embryos with a 133 basepair deletion in the gper open reading frame ( Fig 2A and 2B; S1 Fig ) . Embryos were viable and grossly normal , allowing us to measure heart rate ( Fig 2C and 2D ) . We exposed homozygous maternal zygotic gper mutant embryos ( MZgper-/- ) to estradiol , to ICI182 , 780 or to G1 and found no increase in heart rate compared to embryos exposed to vehicle ( Fig 2E , mean difference in heart rate estradiol versus vehicle 0 . 42 ± 11 . 47 , ICI versus vehicle -9 . 65 ± 12 . 41 , G1 versus vehicle -0 . 38 ± 0 . 33 ) . Our results demonstrate that estradiol increases heart rate in a GPER-dependent manner . Note that zygotic gper mutants exhibited increased heart rate in response to estradiol ( S1 Fig , mean difference in heart rate 29 . 11 ± 6 . 16 bpm ) , indicating that GPER is maternally deposited into oocytes and expressed in embryos . This is consistent with previously published results that detected gper transcript in zebrafish embryos at 1 hpf , suggesting the presence of maternally loaded gper mRNA [28] . To test whether endogenous estrogens regulate heart rate during embryonic development , we examined basal heart rate in GPER mutant embryos reared in untreated water , reasoning that if heart rate was reduced , then that would suggest that endogenous estradiol regulates heart rate via GPER . We compared heart rate in wildtype versus MZgper-/- embryos at 50 hpf and found that MZgper-/- embryos had reduced heart rate compared to wildtype ( Fig 2F , mean difference in heart rate between wildtype and mutant -30 . 80 ± 7 . 07 bpm ) . These results demonstrate that GPER is required for normal basal heart rate in embryos and strongly suggest that endogenous estrogens influence heart rate via GPER . Whether GPER acts as an autonomous estrogen receptor in vivo is controversial . Previous reports suggest that GPER activity might require interaction with nuclear estrogen receptors at the membrane or that estrogens activate GPER indirectly , by binding to nuclear receptors in the cytosol that then activate downstream proteins , including GPER [15 , 29] . To determine whether nuclear estrogen receptors influence heart rate , we generated zebrafish with loss-of-function mutations in each nuclear estrogen receptor gene: esr1 ( ERα ) , esr2a ( ERβ1 ) and esr2b ( ERβ2 ) ( S2–S4 Figs ) . All mutant embryos were viable and grossly normal , allowing us to measure heart rate ( S2–S4 Figs ) . To test whether estradiol increases heart rate via nuclear estrogen receptors , we exposed 49 hpf esr1-/- , esr2a-/- and esr2b-/- embryos to estradiol , G1 or vehicle for one hour and assayed heart rate . Following estradiol exposure , heart rate was increased in all mutants compared to vehicle control ( Fig 3A , mean difference in heart rate between estradiol and vehicle 25 . 04 ± 9 . 83 bpm for esr1-/- , 37 . 23 ± 13 . 27 bpm for esr2a-/- , 32 . 48 ± 3 . 33 bpm for esr2b-/-; mean difference in heart rate between G1 and vehicle 23 . 26 ± 2 . 68 for esr1-/- , 31 . 63 ± 13 . 54 for esr2a-/- , 38 . 38 ± 18 . 50 for esr2b-/- ) , similar to what we observed when wildtype embryos were exposed to estradiol ( Fig 1 ) . These results demonstrate that nuclear estrogen receptors are not necessary for estradiol-dependent increase in heart rate . To test whether endogenous estrogens regulate heart rate via nuclear estrogen receptors , we bred heterozygous fish to generate embryos homozygous for mutations in either esr1 , esr2a or esr2b genes and assayed heart rate in 50 hpf embryos . We observed no significant difference in basal heart rate between homozygotes , heterozygotes or wild type siblings within the same clutch ( Fig 3B , mean difference in heart rate between homozygote and wildtype -4 . 34 ± 2 . 73 bpm for esr1 , -0 . 46 ± 6 . 50 for esr2a , 0 . 63 ± 2 . 87 for esr2b; between heterozygote and wildtype -3 . 34 ± 2 . 05 for esr1 , -0 . 91 ± 2 . 64 for esr2a , -0 . 67 ± 2 . 79 for esr2b ) . To test for maternal effects , we bred homozygous mutant male and females to each other to generate maternal zygotic esr1 and esr2a mutant embryos ( MZesr1 , MZesr2a ) . Heart rate increased following acute exposure to E2 and G1 as above ( S5 Fig; MZesr1 mean difference in heart rate E2 versus vehicle 18 . 69 ± 11 . 67 , G1 versus vehicle 9 . 86 ± 0 . 57; MZesr2a E2 versus vehicle 11 . 73 ± 1 . 51 , G1 versus vehicle 14 . 20 ± 2 . 29 ) . Due to a fertility defect in esr2b -/- females , we were unable to generate MZesr2b embryos . These results suggest that nuclear estrogen receptors are not required for the establishment of normal basal heart rate in embryos . It is possible that the mutations generated in each nuclear estrogen receptor gene do not cause loss of functional estrogen receptor proteins . To exclude this possibility and show that esr mutants exhibit loss of functional ER proteins , we generated esr mutants on the Tg ( 5xERE:GFP ) c262/c262 transgenic background , where green fluorescent protein ( GFP ) expression occurs in cells with activated nuclear estrogen receptors [5] ( referred to as 5xERE:GFP ) . Previous studies using whole mount in situ hybridization demonstrated that esr1 is expressed in embryonic heart valves while esr2b is expressed in the liver [6] , therefore we hypothesized that mutants would fail to upregulate GFP in tissues where the relevant receptor is normally expressed . We exposed 2–3 day post fertilization ( dpf ) 5xERE:GFP , 5xERE:GFP;esr1-/- , 5xERE:GFP;esr2a-/- and 5xERE:GFP;esr2b-/- embryos to 100 ng/ml estradiol overnight and assayed fluorescence . Consistent with esr gene expression patterns , 5xERE:GFP;esr1-/- larvae exhibited fluorescence in the liver but not in the heart ( S2 Fig ) , whereas 5xERE:GFP;esr2b-/- larvae exhibited fluorescence in the heart but not in the liver ( S4 Fig ) . esr2a transcript was not detected at these embryonic and larval stages [6] and , as expected , we saw no change in fluorescence between 5xERE:GFP and 5xERE:GFP;esr2a-/- ( S3 Fig ) . We conclude that the zebrafish nuclear estrogen receptor mutants lack estrogen receptor function . Deleterious mutations can induce genetic compensation [30] , however results from the 5xERE:GFP esr mutants suggest that compensatory expression of esr genes is not occurring . For example , it is possible that in the esr1 mutant there is compensatory upregulation of esr2a and/or esr2b that masks a heart rate phenotype . If esr2a or esr2b were upregulated in esr1 mutants , then we would expect to see fluorescence in the heart in 5xERE:GFP;esr1-/- embryos . Instead , we observed no fluorescence in the hearts of 5xERE:GFP;esr1-/- embryos ( S2 Fig ) . Similarly , we observed no ectopic fluorescence in 5xERE:GFP;esr2b-/- embryos ( S4 Fig ) , suggesting that esr genes are not compensating for one another in the multiple zebrafish esr mutants . To further test whether nuclear estrogen receptor signaling is influenced by GPER , we generated gper mutants on the 5xERE:GFP transgenic background and asked whether estradiol exposure reduced nuclear estrogen receptor activity in mutants compared to wildtype . Following overnight exposure to estradiol , 3 dpf 5xERE:GFP and 5xERE:GFP;MZgper-/- larvae exhibited similar fluorescence ( S6 Fig ) . These results demonstrate that nuclear estrogen receptor transcriptional activity does not require GPER and support the hypothesis that GPER acts as an autonomous estrogen receptor in vivo . Heart rate can be modulated by cardiomyocytes in the heart , or by cells in the central nervous system . Neurons may directly innervate the heart to modulate heart rate and/or regulate the release of humoral factors , such as thyroid hormone , that bind to receptors in cardiomyocytes and regulate heart rate [31] . To determine whether GPER regulates heart rate tissue autonomously , we performed whole mount in situ hybridization to test whether gper transcripts are expressed in 50 hpf zebrafish embryo hearts . We did not detect transcript in the heart or in the vasculature . In contrast , we detected gper mRNA in three discrete anatomic areas of the brain: the preoptic and olfactory areas and in the ventral hypothalamus-pituitary ( Fig 4A–4C ) . Thus , gper localization is consistent with the hypothesis that at 50 hpf , GPER acts in the brain , and not through cells in the heart , to regulate heart rate . At later stages of development and adulthood , it is possible that gper is expressed in the heart and regulates cardiac function . Genetic evidence using esr mutants suggests that GPER acts independently of nuclear estrogen receptors to regulate heart rate ( Fig 3 ) . To further test the hypothesis that GPER acts as an autonomous estrogen receptor in vivo , we asked whether GPER and nuclear estrogen receptors are expressed in the same cells in the brain , reasoning that if GPER and nuclear estrogen receptors fail to colocalize , this would support the idea that GPER acts as an autonomous estrogen receptor in vivo . We exposed 1 dpf 5xERE:GFP embryos to 100 ng/ml estradiol overnight . At 48 hpf , we fixed the embryos and used two color fluorescent in situ hybridization to detect gfp and gper transcripts simultaneously . Since all three nuclear estrogen receptor genes activate the 5xERE:GFP transgene , detecting gfp allows us to monitor activity of all three estrogen receptors using a single RNA probe . In the olfactory and preoptic areas , we found no colocalization between gfp and gper ( Fig 4D and 4E ) . In the ventral hypothalamus , we found a cluster of cells at the midline expressing gper but not gfp . Surrounding this region of gper-positive cells was a bilaterally symmetric ‘U’-shaped labeling pattern of cells expressing both gper and gfp ( Fig 4F ) . These results demonstrate that GPER and nuclear estrogen receptors are expressed in unique and overlapping cells in the brain , supporting the hypothesis that GPER can act independently of nuclear estrogen receptors in vivo . At 2 dpf , the ventral hypothalamus and pituitary are contiguous , therefore gper could be expressed in both locations . To determine whether gper is expressed in the pituitary , we performed whole-mount two color fluorescent in situ hybridization to detect gper and pituitary cell markers: thyroid stimulating hormone ( tshb ) , expressed in thyrotropes in the middle zone of the pituitary , and proopiomelanocortin ( pomca ) , expressed in corticotropes and melanotropes in the anterior and posterior zones of the pituitary , respectively [32 , 33] . We identified a subpopulation of tshb-positive cells that also expressed gper ( Fig 4G ) , suggesting that GPER functions in thyrotropic pituitary cells . In the anterior zone , we identified cells expressing both gper and pomca ( Fig 4H ) . In contrast , we detected no gper transcript in cells in the posterior zone of the pituitary ( Fig 4H” ) . We conclude that gper is expressed in the embryonic pituitary , including in thyrotropes and corticotropes . The pituitary regulates systemic levels of many hormones , such as gonadotropins and thyroid hormones . Because gper is expressed in the pituitary , we wondered whether GPER is required for pituitary development or function . To test this idea , we measured levels of hormones regulated by the pituitary: the principle endogenous estrogens estrone ( E1 ) , estradiol ( E2 ) and estriol ( E3 ) , and thyroid hormone triiodothyronine ( T3 ) . Pituitary gonadotropins stimulate the ovary to synthesize estrogens , which are released into systemic circulation and thought to be deposited into oocyte yolks and absorbed by the embryo [34] . Thyrotrope cells in the pituitary regulate levels of thyroid hormone T3 , which increases heart rate in mammals [35–45] . Using a new mass spectrometry assay to measure circulating estrogen levels in zebrafish blood plasma , we found that adult MZgper female zebrafish had lower levels of E2 and E1 compared to wild-type females ( Table 1 , wild-type female mean 32 ng/ml E2 , 487 pg/ml E1; MZgper female <25 pg/ml E2 and E1 , below limit of detection; E3 below limit of detection in all samples ) . Among wild-type fish , we observed reduced levels of E1 and E2 in males compared to females ( Table 1 ) , consistent with the idea that females have higher circulating levels of estrogens than males and supporting the validity of the assay . Using an enzyme linked immunosorbent assay to detect T3 , we observed that MZgper embryos showed a 50% reduction in total T3 levels compared to wild-type embryos ( Fig 5A ) . We conclude that gper mutants have reduced levels of E1 , E2 and T3 , consistent with an abnormal pituitary . To determine the order in which E2 and T3 act in the GPER signaling pathway , we performed pharmacologic rescue experiments . We depleted maternal estrogens in the presence or absence of gper by generating maternal ( Mgper ) or maternal zygotic ( MZgper ) gper mutant embryos and asked whether exogenous E2 or T3 could rescue the heart rate phenotype . Mgper embryos , derived from breeding gper-/- females with wild-type males , have reduced deposition of maternal estrogens but express wild-type gper . MZgper embryos , derived from breeding gper -/- males with gper -/- females , have reduced deposition of maternal estrogens and lack functional gper . Both Mgper and MZgper embryos have reduced basal heart rate compared to wild type ( S7 Fig; difference between wild type and Mgper -24 . 5 ± 3 . 42 bpm , between wild type and MZgper -25 . 4 ± 1 . 62 ) . We found that E2 increased heart rate only when wild-type gper was present , in Mgper but not in MZgper embryos ( Fig 5B; Mgper difference between E2 and vehicle 17 . 94 ± 7 . 05 , MZgper difference between E2 and vehicle 3 . 44 ± 3 . 35 ) . In contrast , T3 increased heart rate in the absence of both E2 and GPER , in MZgper embryos ( Fig 5C , mean difference in MZgper heart rate between T3 and vehicle 23 . 60 ± 4 . 98 bpm ) . Therefore , we conclude that T3 acts downstream of E2 and GPER . To confirm that T3 increases heart rate in zebrafish as it does in mammals , we exposed wild-type embryos to T3 and found a mean 13% increase in heart rate compared to embryos exposed to vehicle ( Fig 5C; mean difference in heart rate 22 . 86 ± 3 . 43 bpm ) . We also confirmed that heart rate increased in Mgper embryos following G1 exposure ( S7 Fig; mean difference between G1 and vehicle 28 . 63 ± 8 . 09 ) , demonstrating that Mgper embryos express zygotic wild-type gper . We conclude that adult gper mutant females have lower levels of circulating estrogens than wild-type females and consequently deposit less estrogens into embryos . Our results support a model whereby GPER , likely acting in the pituitary , regulates estrogen levels in adult females , allowing sufficient maternal deposition of estrogens into oocytes . As embryos develop , the maternally deposited estrogens activate GPER , leading to increased T3 levels and proper heart rate ( Fig 6 ) .
Here we provide evidence that estrogens signal through a non-canonical estrogen receptor , the G protein-coupled estrogen receptor ( GPER ) , to regulate heart rate in zebrafish embryos by regulating levels of maternally deposited estrogens and by altering levels of thyroid hormone T3 in embryos . Our results also support the hypothesis that GPER acts as an autonomous estrogen receptor in vivo . Previous reports using cultured cells demonstrated that fluorescently labeled or isotopic estradiol specifically binds membranes from cells expressing GPER [8 , 9] . Additionally , estradiol exposure increased cyclic AMP and calcium levels in HEK293 and COS7 cells in a GPER-dependent manner [8 , 9] , while estradiol exposure increased phosphoinositide 3-kinase activity in SKBR3 breast cancer cell line in a GPER-dependent manner [8] . However , because these studies utilized cells that either express artificially high levels of GPER or are tumorigenic , the findings do not address whether GPER acts as an estrogen receptor in vivo under normal physiologic conditions . Our genetic and pharmacologic results strongly suggest that GPER is an estrogen receptor in vivo . If estradiol was binding to ERα or ERβ , and these receptors activated GPER , then we would expect to see no increase in heart rate in esr1 , esr2a or esr2b mutants following exposure to estradiol . Instead , all esr mutants responded normally to estradiol and G1 exposure ( Fig 3 , S5 Fig ) , suggesting that ER and GPER signaling pathways are distinct in this context . Consistent with these results , we found gper transcript expressed in cells in the brain that lack nuclear estrogen receptor activity ( Fig 4 ) , further supporting the hypothesis that GPER responds to estrogens independently of nuclear estrogen receptors in vivo . Studying the influence of estrogens on heart rate in zebrafish embryos is a powerful in vivo system where GPER activity is dissociated from classical nuclear estrogen receptor signaling . Between 2 and 5 dpf , zebrafish heart rate normally increases [46 , 47] . Our results support the hypothesis that endogenous estradiol regulates this increase in heart rate . The finding that Mgper and MZgper embryos have lower basal heart rate compared to wild type implicates endogenous estradiol . Additionally , a recent HPLC analysis of endogenous estradiol concentration in zebrafish embryos found that estradiol concentrations increased from 137 pg/embryo at 48 hpf to 170 pg/embryo at 72 hpf [48] . However , whether this increase in embryonic estradiol is due to increased synthesis by embryos or increased release of estrogens from the yolk is not known . Our genetic and mass spectrometry results support the hypothesis that at 48 hpf , maternally derived estrogens are required for normal heart rate . Data from Mgper embryos suggests that , in the absence of maternal deposition of estrogens , basal heart rate is reduced . Since wild-type gper is present in Mgper embryos , the reduced basal heart rate in Mgper embryos is due to reduced maternal deposition of estrogens ( leading to reduced activation of wild-type GPER ) and not due to mutant gper ( Fig 6 ) . We cannot exclude the possibility that reduced maternal estrogen levels indirectly caused reduced GPER activity in embryos in an estrogen-independent manner , though this seems unlikely . The identity of the physiologic estrogen ( or estrogens ) that regulates heart rate is not known . Our mass spectrometry results implicate E2 , the most potent and abundant endogenous estrogen in humans [49] and many vertebrate species . However , E1 levels were also reduced in gper mutant females compared to wild-type females ( Table 1 ) , thus E1 may also contribute to regulation of heart rate . Many other endogenous estrogens exist , such as 27-hydroxycholesterol and estetrol [50 , 51] , and it will be interesting to see to what degree they are present in embryos , regulate embryonic heart rate and activate GPER . There are several mechanisms by which GPER activity in the brain could regulate heart rate , for example by modulating sympathetic and parasympathetic nerve activity or by regulating the release of humoral factors , such as thyroid hormone . Expression of gper transcript in thyrotropic cells in the pituitary ( Fig 4 ) and decreased levels of total T3 in gper mutants ( Fig 5 ) support the latter hypothesis . There are three primary mechanisms by which GPER could promote the increase of total T3: 1 ) by increasing levels of thyroid stimulating hormone , leading to release of thyroxine ( T4 ) from the thyroid , which is then converted to T3 , 2 ) by increasing the conversion of T4 to T3 , or 3 ) by blocking the conversion of T3 into inactive metabolites , such as 3 , 5-Diiodo-L-thyronine ( T2 ) and reverse T3 ( RT3; 3 , 3’5’-triiodothyronine ) . gper expression in tshb-positive cells in the pituitary supports the first hypothesis . In humans , thyroid stimulating hormone is thought to be required for the differentiation of the thyroid [52] . Curiously , zebrafish mutant embryos that lack thyrotrope progenitor cells and tsh gene expression still produce thyroid follicle cells and T4 [53] . This suggests that thyrotropes may not be required to produce T3 , however total T3 was not measured in this study . It is possible that the localized production of T4 is sufficient to stimulate thyroid development , while total T3 is reduced . In support of this , a majority of the cartilage in the pharynx is missing in thyrotrope deficient zebrafish [53] , which suggests that even though the thyroid begins to differentiate , tissues adjacent to the thyroid do not , presumably due to reduced total T4 in circulation . While TSH signaling may not be required for the development of the thyroid , it may be required for secondary functions , including proper regulation of heart rate . We cannot exclude the possibility that GPER activity leads to increased expression or activity of deiodinase enzymes that convert T4 to T3 , or that GPER activity reduces the expression or activity of enzymes that metabolize T3 . All four deiodinases genes ( dio1 , dio2 , dio3a and dio3b ) are expressed in zebrafish as early as 24 hpf [54] and are therefore available to convert thyroid hormones at 48 hpf , when we observe changes in heart rate . Interestingly , at 24 hpf dio2 , the enzyme that converts T4 to T3 , is expressed in the pituitary [55] in addition to its expression in the thyroid . dio3a , which inactivates T3 by conversion to RT3 and T2 , is also expressed in the brain and thyroid at 24 hpf [55] . Previous work suggests that GPER influences neurotransmitter release and cAMP levels [56] . cAMP was shown to increase deiodinase activity in the brain leading to increased T3 levels [57–59] . Therefore , GPER activity could trigger neuronal activity that leads to increased activity of deiodinases and increased production of T3 , independently of TSH , to regulate heart rate [31] . More generally , it is not known to what extent estrogens influence thyroid hormone signaling . Our work suggests that endogenous estrogens influence T3 levels . Previous work suggests that environmental estrogens may also influence thyroid hormone signaling . Exposure to the plasticizer diethylhexyl phthalate ( DEHP ) increased total T3 levels in zebrafish larvae and upregulated thyroid signaling genes thyroglobulin ( tg ) , transthyretin ( ttr ) , and dio2 [60] . DEHP exhibits estrogen-like activity , although the receptor by which it acts has not been determined . DEHP can inhibit tamoxifen-induced apoptosis and also induce cell proliferation in GPER positive MCF-7 cells , but not in GPER negative MDA-MB-231 cells [61] , suggesting that DEHP can activate GPER . One possibility is that DEHP increases T3 levels in zebrafish larvae via GPER activation . Similarly , chronic exposure to perfluorooctanesulphonic acid ( PFOS ) , a surfactant that enhances the effects of estradiol [62] , increased total T3 levels in juvenile zebrafish and upregulated thyroid signaling genes including thyroid hormone receptor β , the sodium/iodide symporter slc5a5 , dio1 and dio2 [63] . We speculate that like estradiol , the environmental endocrine disruptors DEHP and PFOS modulate T3 levels by activating GPER . This raises the important consideration that diverse environmental estrogens could alter thyroid signaling and thus cardiac function . While our results illuminate GPER signaling in the context of embryonic heart rate , it is not clear to what extent GPER influences heart rate at later stages of development . At larval , juvenile and adult stages it is difficult to assess heart rate without immobilizing or anesthetizing zebrafish , manipulations that themselves may influence heart rate . In adult mice with mutations in GPER , there was no significant difference in basal heart rate between mutant and wild type of either sex [16 , 17 , 64] . It is possible that GPER regulates heart rate in embryos but not in adults . Additionally , heart rate in GPER mutant mice was assayed using general anesthesia , which is known to depress heart rate compared to conscious mice [65] . Anesthesia may mask the effect of GPER on basal heart rate that we observe in conscious animals . We also cannot exclude the possibility that the effects of GPER on heart rate are specific for zebrafish . The fact that adult gper mutant females have reduced circulating E2 and E1 suggests that gper mutants will exhibit additional phenotypes in juvenile and adult stages of life . It is possible that female gper mutants have reduced fertility or become infertile at an earlier age compared to wild type , phenotypes that the current study was not powered to detect . Additionally , reduced estrogen levels may indicate deficits in locomotor activity , mating behavior or aggression , behaviors known to be influenced by estrogens [66–71] . Considering that GPER deficient zebrafish embryos have reduced T3 levels , it will be interesting to examine whether this deficiency exists at later developmental stages and whether GPER mutant adults have growth and metabolic defects consistent with reduced total T3 . In MZgper embryos , we observed no gross morphological detects up to 2 dpf , while mutant adults are viable and fertile . Zebrafish mutants with ~70–90% reduced total T3 levels due to genetic ablation of dio2 exhibit delayed swim bladder inflation , altered locomotor activity through 7 dpf , delayed fertility , reduced number of eggs , and reduction in viable fertilized eggs [72] . In GPER deficient fish , the reduction in T3 levels is less drastic and we anticipate seeing less severe phenotypes as a result of the more modest decrease in T3 . The zebrafish estrogen receptor mutants we developed enable experiments to rapidly and conclusively identify the causative estrogen receptor associated with any estrogen signaling phenotype , as demonstrated with the estradiol-dependent increase in heart rate reported here . This has significant implications for studies of estrogenic environmental endocrine disruptors , which are frequently tested on zebrafish to identify effects on embryonic development , organ formation and function [73] . Zebrafish estrogen receptor mutants can now be used to determine whether such effects are specific for estrogen receptors and to identify the precise receptor target . Our results also establish a need to consider the impact on cardiac function when considering the toxicity of estrogenic environmental endocrine disruptors .
Zebrafish were raised at 28 . 5°C on a 14-h light , 10-h dark cycle in the UAB Zebrafish Research Facility in a recirculating water system ( Aquaneering , Inc . , San Diego , CA ) . Wildtype zebrafish were AB strain [74] and all mutant and transgenic lines were generated on the AB strain . To visualize nuclear estrogen receptor activity , transgenic line Tg ( 5xERE:GFP ) c262/c262 was used for all studies unless otherwise mentioned [5] . All procedures were approved by the UAB Institutional Animal Care and Use Committee . Embryos were collected during 10 minute intervals to ensure precise developmental timing within a group . Embryos were placed in Petri dishes containing E3B ( 60X E3B: 17 . 2g NaCl , 0 . 76g KCl , 2 . 9g CaCl2-2H2O , 2 . 39g MgSO4 dissolved in 1 liter Milli-Q water; diluted to 1X in 9 liter Milli-Q water plus 100 μL 0 . 02% methylene blue ) and placed in an incubator at 28 . 5°C on a 14-h light , 10-h dark cycle . At 24 hours post fertilization ( hpf ) , embryos were incubated in E3B containing 200 μM 1-phenyl 2-thiourea ( PTU ) to inhibit pigment production [74] . Between 24 and 48 hpf , embryos were manually dechorionated and randomly divided into control and experimental treatment groups in 60mm Petri dishes and kept at 28 . 5°C until 49 hpf . At 49 hpf , embryos were incubated in E3B with estrogen receptor modulator ( s ) at 28 . 5°C for 1 hour . Estrogen receptor modulator treatments consisted of: 3 . 67 μM E2 ( 17β-estradiol; Sigma E8875 , purity ≥ 98% ) , 10 μM ICI182 , 780 ( fulvestrant; Sigma I4409 , purity >98% ) , 1 μM G1 ( Azano AZ0001301 , purity ≥ 98% ) , 1 μM G36 ( Azano , AZ-0001303 , purity ≥ 98% ) , 1 μM progesterone ( Sigma P0130 , purity ≥ 99% ) , 5 nM 3 , 3′ , 5-Triiodo-L-thyronine ( T3; Sigma T2877 , purity ≥ 95% ) , vehicle ( 0 . 1% dimethylsulfoxide ( DMSO ) , Fisher D128-500; purity ≥ 99 . 9% or 0 . 1% methanol , Fisher A411-4 ) . All chemical stocks were made in 100% DMSO at 1000x and diluted in E3B embryo media to final concentration at the time of treatment , except for T3 which was made in 100% methanol for chemical stocks but was diluted in E3B media as above . Exposure concentrations were chosen based on previous evidence of efficacy . 3 . 67 μM E2 and 10 μM ICI were previously shown to be effective at activating and inhibiting nuclear estrogen receptor activity in zebrafish embryos [5 , 6] . The doses of G1 and G36 were selected based on reports that chronic exposure to G1 elicits concentration-dependent effects on gross morphology and gene expression , which are blocked by G36 [28 , 75] ( note that since we exposed zebrafish to G1 and G36 for only 1 hour , we did not observe defects in gross morphology ) . We chose 1 μM progesterone because it is non-toxic following 1 hour exposure and because previous studies have demonstrated that <1 μM concentrations are effective at regulating progestin-dependent gene expression in zebrafish embryos and larvae [76–79] . The T3 concentration used was previously shown to be effective at inducing thyroid hormone-dependent gene expression and premature differentiation of pectoral fins in zebrafish [54 , 80] . For rescue experiments ( ICI182 , 780 + G36 ) , final DMSO concentration was 0 . 2% . There was no difference in heart rate between embryos incubated in 0 . 1% or 0 . 2% DMSO ( not shown ) . All vehicle controls shown in figures are 0 . 1% DMSO , except where indicated . All embryos were reared at 28 . 5°C and heart rate was measured at room temperature . Following one hour incubation in treatment compounds at 28 . 5°C , heart rate ( beats per minute , bpm ) was calculated by counting the number of heart beats in fifteen seconds and multiplying that number by four . Prior to measurements , each dish was removed from the incubator and placed under the microscope light for 4 minutes at room temperature , allowing embryos to acclimate to the light and eliminate any effect of the startle response . At the time of heart rate measurement , water temperature was 25°C ( assayed with Fisher Scientific digital thermometer with stainless steel probe , catalogue number 15-077-9D ) . Control groups were counted first and last to ensure that the overall heart rate did not increase during the duration of counting due to natural increases in heart rate during development . All heart rates were measured on a Zeiss Stemi 2000 dissecting microscope with a halogen transmitted light base ( Carl Zeiss Microimaging , Thornwood , NJ ) . Absolute heart rates in wildtype , untreated ( or vehicle control treated ) embryos at 48–50 hpf were frequently 140–150 bpm , consistent with previous studies [46 , 81–83] . However , there is variability in heart rate , despite controlling for conditions such as developmental stage and water temperature , due to factors we do not yet understand . For example , on some days , wildtype heart rate was approximately 130 bpm ( see S1 Table ) . However , the change in heart rate between treated and untreated embryos , or between wildtype and mutant embryos , was always consistent , irrespective of the absolute heart rate values recorded on the day we performed the experiment ( see S1 Table for absolute heart rate data ) . Therefore , we report both mean difference in heart rate between groups ( not normalized ) and fold difference in heart rate compared to control ( normalized ) . We performed a rigorous statistical analysis on the non-normalized data to determine whether differences in heart rate were due to clutch and day variability and found that differences in heart rate were due to the effects of treatment or genotype ( see Experimental Design and Data Analysis ) . Plasmids pT7-gRNA and pT3TS-nCas9n were obtained from Addgene ( numbers 46759 , 46757 ) [27] . pT7-gRNA was digested simultaneously with BsmBI , BglII and SalI for one hour at 37 oC followed by one hour at 55 oC . To generate esr2a , esr2b and gper gRNAs , oligonucleotides containing target site sequences ( Table 2 ) were synthesized by Invitrogen . Oligos were hybridized to each other using NEBuffer3 restriction enzyme buffer ( New England Biolabs ) to generate double stranded target DNA and annealed into digested pT7-gRNA using Quick T4 DNA Ligase ( New England Biolabs ) as previously described [27] . Guide RNAs were synthesized using the MegaShortScript T7 Kit ( Life Technologies ) using the relevant modified pT7-gRNA vector linearized with BamHI as a template . Guide RNA was purified using the RNA clean & concentrator kit ( Zymo Research ) . To generate esr1 guide RNA , target-specific oligonucleotides containing the SP6 ( 5’-ATTTAGGTGACACTATA ) promoter sequence , a 20 basepair target site without the PAM , and a complementary region were annealed to a constant oligonucleotide encoding the reverse-complement of the tracrRNA tail as described [84] . This oligo was used as a template for in vitro transcription using the MegaShortScript Sp6 Kit ( LifeTechnologies ) . To generate Cas9 mRNA , the pT3TS-nCas9n plasmid was linearized with XbaI and transcribed using the mMessage mMachine T3 kit ( Life Technologies ) and purified using RNA clean & concentrator kit ( Zymo Research ) . RNA concentration was quantified using a Nanodrop spectrophotometer ( Nanodrop ND-1000 , ThermoFisher ) . One-cell-stage embryos were injected using glass needles pulled on a Sutter Instruments Fleming/Brown Micropipette Puller , model P-97 and a regulated air-pressure micro-injector ( Harvard Apparatus , NY , PL1–90 ) . Each embryo was injected with a 1 nl solution of 150 ng/μl of Cas9 mRNA , 50 ng/μl of gRNA and 0 . 1% phenol red . Mixtures were injected into the yolk of each embryo . Approximately 100 injected embryos per gRNA were raised to adulthood and crossed to wild-type fish ( either AB or Tg5xERE:GFPc262 ) to generate F1 embryos . F1 offspring with heritable mutations were sequenced to identify loss of function mutations . Individual embryos or tail biopsies from individual adults were placed in 100 μL ELB ( 10 mM Tris pH 8 . 3 , 50 mM KCl , 0 . 3% Tween 20 ) with 1 μL proteinase K ( 800 U/ml , NEB ) in 96 well plates , one sample per well . Samples were incubated at 55°C for 2 hours ( embryos ) or 8 hours ( tail clips ) to extract genomic DNA . To inactivate Proteinase K , plates were incubated at 98°C for 10 minutes and stored at -20°C . PCR and melting curve analysis was performed as described [85] . PCR reactions contained 1 μl of LC Green Plus Melting Dye ( BioFire Diagnostics ) , 1 μl of Ex Taq Buffer , 0 . 8 μl of dNTP Mixture ( 2 . 5 mM each ) , 1 μl of each primer ( 5 μM ) , 0 . 05 μl of Ex Taq ( Takara Bio Inc ) , 1 μl of genomic DNA , and water up to 10 μl . PCR was performed in a Bio-Rad C1000 Touch thermal cycler , using black/white 96 well plates ( Bio-Rad HSP9665 ) . PCR reaction protocol was 98°C for 1 min , then 34 cycles of 98°C for 10 sec , 60°C for 20 sec , and 72°C for 20 sec , followed by 72°C for 1 min . After the final step , the plate was heated to 95°C for 20 sec and then rapidly cooled to 4°C . Melting curves were generated with either a LightScanner HR 96 ( Idaho Technology ) over a 70–95°C range and analyzed with LightScanner Instrument and Analysis Software ( V . 2 . 0 . 0 . 1331 , Idaho Technology , Inc , Salt Lake City , UT ) , or with a Bio-Rad CFX96 Real-Time System over a 70–95°C range and analyzed with Bio-Rad CFX Manager 3 . 1 software . All mutations were confirmed by TA cloning and sequencing . Live zebrafish embryos and larvae were visualized using a Nikon MULTIZOOM AZ100 equipped with epi-fluorescence and an Andor Clara digital camera unless otherwise noted . To validate mutants with 5xERE reporter activity , larvae were treated overnight with 100 ng/mL estradiol beginning at 2–3 dpf . Following overnight treatment , larvae were washed in E3B , anesthetized with 0 . 04% tricaine and imaged in Petri dish containing E3B . For S1H–S1K Fig , larvae were mounted in bridged coverslips in E3B with 0 . 04% tricaine [74] . Images were captured on a Zeiss Axio Observer . Z1 fluorescent microscope equipped with an Axio HRm camera and Zen Blue 2011 software ( Carl Zeiss Microscopy , Oberkochen , Germany ) . Adjustments , cropping and layout were performed using Photoshop CC and InDesign CC ( Adobe Systems Inc . , San Jose , CA ) . For synthesis of RNA probes , full-length gper open reading frame was amplified by PCR from genomic DNA extracted from 3 dpf larvae ( gper coding region is within a single exon and therefore the open reading frame sequence is identical in genomic and cDNA ) using primers 5’-ATGGAGGAGCAGACTACCAATGTG-3’ and 5’-CTACACCTCAGACTCACTCCTGACAG-3’ . For tshb probe , a 252bp product was amplified by PCR from cDNA ( prepared from total RNA from 5 dpf AB larvae using RETROscript reverse transcription kit ( ThermoFisher Scientific ) with oligo ( dT ) primers ) using primers 5’ GAGTTGGTGGGTCCTCGTTT 3’ and 5’ TGCTTGGGCGTAGTTGTTCT 3’ . Each product was then TA cloned into pCR2 . 1 vector ( Invitrogen ) . amhc and gfp probes were used as described [5 , 86] . All clones were verified by sequencing . Digoxigenin-labeled antisense RNA and FITC-labeled antisense RNA were transcribed using T7 and T3 polymerase , respectively , as previously described [5] . Colorimetric whole-mount in situ hybridization was performed on zebrafish embryos and larvae as described previously , using 5% dextran in the hybridization buffer [87 , 88] . Following colorimetric in situ hybridization , embryos were sequentially cleared in glycerol ( 25% , 50% , 75% in phosphate buffered saline ) , mounted in 4% low-melting temperature agarose , and imaged using a Zeiss Axio Observer . Z1 microscope with Zeiss Axio MRc5 camera and Zen Blue 2011 software . Fluorescent in situ hybridization ( FISH ) was performed as previously described [88] with the following modifications: After rehydration , Proteinase K treatment was extended to 35 minutes . Following hybridization , embryos were washed in 2xSSC prior to being placed in PBT . Embryos were blocked in 2% Roche blocking reagent in 100 mM Maleic acid , 150 mM NaCl , pH 7 . 5 [89] . For double labeling , following development of anti-DIG-POD antibody , reaction was inactivated in 100 mM glycine pH 2 for 10 minutes then incubated in anti-FITC antibody . Following florescent in situ hybridization , embryos were cleared in 50% glycerol , mounted on a bridged coverslip and imaged using a Nikon A1/R scanning confocal microscope with Nikon Advanced Elements software . T3 levels were measured using enzyme-linked immunosorbent assay as previously described [90] , with minor modifications , using T3 ELISA Kit ( IBL America IB19107 ) . Briefly , 50 embryos were pooled in 50 μl of PBS and pulsed sonicated intermittently for 5 minutess , alternating 5 second sonication and 5 seconds on ice , then vortexed intermittently for 10 minutes , alternating 30 seconds vortexing and 30 seconds on ice . Samples where then centrifuged for 10 minutes at 15 , 000g at 4°C . Supernatant was collected and diluted 1:8 in PBS . 50 μl was used per reaction in accordance with the manufacturer’s instructions . Each sample was tested in duplicate and the mean of duplicates were compared statistically . Plasma from adult zebrafish was collected as described [91] with the following modifications . Zebrafish were anesthetized in 0 . 04 mg/mL tricaine then patted dry . Standard length [92] was measured with digital calipers . The caudal fin was completely severed with a razor blade and discarded , then each fish was transferred into a perforated 0 . 5 mL microcentrifuge tube placed inside a 1 . 5 mL microcentrifuge tube containing 10 μL heparin ( 5 mg/mL in water; Sigma-Aldrich #H3393 ) . Fish were centrifuged ( 1000 rpm , 5 minutes , 4°C ) , then the tail was cut above the previous incision and centrifuged with the same parameters . Whole blood collected in the 1 . 5 mL tube was then centrifuged ( 14000 rpm , 15 minutes , 4°C ) to isolate plasma ( supernatant ) . Plasma was stored at -80°C until mass spectrometry analysis . Prior to mass spectrometry , plasma samples were initially subjected to a liquid-liquid extraction using hexane:ethyl acetate , followed by derivatization with dansyl chloride and then the final solid-phase extraction before analysis . Estrone-d4 , Estradiol-d5 , and Estriol-d3 ( Toronto Research Chemicals Inc . , North York , Ontario , Canada ) were used as internal standards . E1 , E2 , E3 and respective internal standards were eluted from a C18 column ( XSelect HSS T3 , 2 . 1x75mm , 2 . 5μm , Waters Corporation , Milford , MA ) using mobile phase gradient where mobile phase A consisted of 0 . 1% formic acid in water and mobile phase B consisted of 0 . 1% formic acid in acetonitrile:isopropanol . Analytes were detected in positive ion mode using multiple reaction monitoring ( Sciex Qtrap 6500+ with IonDrive TurboV Source , Sciex , Foster City , CA ) : E1 , 504 . 1→171 . 1 m/z; E2 , 506 . 1→171 . 1 m/z; E3 , 522 . 1→171 . 1 m/z; E1-d4 , 508 . 1→171 . 1 m/z; E2-d5 , 511 . 1→171 . 1 m/z; and E3-d3 , 525 . 1→171 . 1 m/z . The dynamic range of the assay was 25 to 2500 picograms/mL for each of the target analytes using 5μL of plasma . Heart rate assays were conducted in separate experiments . Each experiment included comparing groups ( treated vs untreated or mutant vs wildtype ) using at least 3 embryos per group with all embryos from the same clutch . All experiments were replicated at least 3 times ( n≥3 ) using different clutches . This is essentially a complete block design with clutch as block . Mean heart rate of individual embryos from a clutch was used for comparing treatment groups ( or mutant groups ) within experiments using two-way ANOVA controlling for clutch effect . The overall treatment effect ( or the genotype effect in some experiments ) was tested using F test . If it was significant , Dunnett’s test was then used to compare each treatment group with the vehicle group or mutant group with the wildtype group . For some individual pairs of comparisons , paired t test was used . Significance level is 0 . 05 . All the analyses were conducted using R ( version 3 . 0 . 2 ) . Graphs were produced using GraphPad Prism 7 . 0c software .
|
Estrogen hormones are important for the formation and function of the nervous , reproductive and cardiovascular systems . Here we report that acute exposure to estrogens increases heart rate , a previously unappreciated function of estrogens . Using zebrafish with mutations in genes that respond to estrogens , we found that heart rate is regulated not by the typical molecules that respond to estrogens–the nuclear estrogen receptors–but rather by a different molecule , the G protein-coupled estrogen receptor . We also show that estrogens increase heart rate by increasing levels of thyroid hormone . Our results reveal a new function for the G protein-coupled estrogen receptor and a new connection between estrogens and thyroid hormone . Environmental compounds that mimic estrogens can be harmful because they can influence gonad function . Our results suggest that endocrine disrupting compounds may also influence cardiac function .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"fish",
"nervous",
"system",
"cardiovascular",
"anatomy",
"vertebrates",
"neuroscience",
"animals",
"hormones",
"animal",
"models",
"osteichthyes",
"developmental",
"biology",
"model",
"organisms",
"heart",
"experimental",
"organism",
"systems",
"embryos",
"estrogens",
"estradiol",
"research",
"and",
"analysis",
"methods",
"cardiology",
"embryology",
"pituitary",
"gland",
"life",
"cycles",
"heart",
"rate",
"biochemistry",
"zebrafish",
"eukaryota",
"neuroanatomy",
"anatomy",
"lipid",
"hormones",
"endocrine",
"system",
"biology",
"and",
"life",
"sciences",
"larvae",
"organisms"
] |
2017
|
G protein-coupled estrogen receptor regulates embryonic heart rate in zebrafish
|
Schistosomiasis has a considerable impact on public health in many tropical and subtropical areas . In the new world , schistosomiasis is caused by the digenetic trematode Schistosoma mansoni . Chemotherapy is the main measure for controlling schistosomiasis , and the current drug of choice for treatment is praziquantel ( PZQ ) . Although PZQ is efficient and safe , its repetitive large-scale use in endemic areas may lead to the selection of resistant strains . Isolates less susceptible to PZQ have been found in the field and selected for in the laboratory . The impact of selecting strains with a decreased susceptibility phenotype on disease dynamics and parasite population genetics is not fully understood . This study addresses the impact of PZQ pressure on the genetics of a laboratory population by analyzing frequency variations of polymorphic genetic markers . Infected mice were treated with increasing PZQ doses until the highest dose of 3×300 mg/Kg was reached . The effect of PZQ treatment on the parasite population was assessed using five polymorphic microsatellite markers . Parasitological and genetic data were compared with those of the untreated control . After six parasite generations submitted to treatment , it was possible to obtain a S . mansoni population with decreased susceptibility to PZQ . In our experiments we also observed that female worms were more susceptible to PZQ than male worms . The selective pressure exerted by PZQ led to decreased genetic variability in S . mansoni and increased endogamy . The understanding of how S . mansoni populations respond to successive drug pressure has important implications on the appearance and maintenance of a PZQ resistance phenotype in endemic regions .
Schistosomiasis is a disease that affects approximately 206 million people in 76 countries . On a global scale , 779 million people are at risk of contracting the disease [1] . Currently , the main approaches for the reduction of schistosomiasis morbidity are the treatment of infected patients and sanitation [2] . The life cycle of the parasite is complex , involving a vertebrate host and a snail vector . Infection occurs when parasites in the infective stage ( cercariae ) are released from aquatic snails and penetrate the skin or mucosa of the vertebrate host upon contact with water . By migrating through the circulatory system , the larvae ( schistosomula ) reach the portal hepatic vein where they mature into adult worms and migrate to the mesenteric veins . The female then deposits eggs in the intestinal vein system . Some eggs pass through the gut lumen and are eliminated with the feces , or urine in the case of S . haematobium , while others are trapped in the host tissues where they produce a granulomatuous immune-inflammatory process that represents the main cause of pathology [3] . Praziquantel ( PZQ ) is the drug used for treatment of human schistosomiasis and is inexpensive and effective against all human schistosome and some cestode species [4] . However , PZQ is less effective against the immature stages of Schistosoma mansoni [5] . PZQ appears to exert multiple effects on schistosomes , damaging the tegument and causing muscle contraction [6] . These morphological alterations are accompanied by an increased exposure of parasite antigens on the worm surface , which in turn appears to render the worms more susceptible to attack by the immune system [7] . The relationship between PZQ and Ca+2 influxes suggests that the sites of action are Ca2+ ion channels in the membrane of tegument and muscle cells [8] , [9] , [10] . Other effects of PZQ in Schistosoma that have been demonstrated include the following: reduction in the level of glutathione ( GSH ) [11]; inhibition of the male excretory system [12]; inhibition of nucleoside uptake [13]; and increase in MDR transporter activity [14] , [15] . However , a complete description of PZQ's mode of action has not yet been accomplished . The availability of only one drug to treat schistosomiasis is a concern , and too little investment has been made toward the development of alternatives [16] . There is some natural variation in the parasite sensitivity to PZQ . The variability of S . mansoni strains regarding susceptibility to schistosomicidal drugs was demonstrated in different endemic regions , such as Senegal , Egypt and Brazil [17] , [18] , [19] , [20] , [21] , [22] , [23] . It is conceivable that in an endemic situation a resistant phenotype may be selected for under frequent drug pressure [18] . Resistant , less susceptible isolates or parasites that survive treatment of S . mansoni in endemic regions may emerge by different mechanisms . The main causes may be high parasitic loads , fast reinfection , the presence of immature worms , inefficient human immune responses , the use of low quality drugs , differences in evaluation , and cure criteria or actual drug resistance [24] , [25] . The first study that showed it is possible to induce S . mansoni less susceptible to PZQ was published by Fallon and Doenhoff [26] . Recently , PZQ resistant populations were also obtained after the intramolluscan parasite phase was exposed to the drug [27] . The genetics of S . mansoni populations are being investigated with the use of highly polymorphic microsatellite markers [28]; the availability of a genetic MAP significantly contributed to these efforts [29] . Polymorphic microsatellites have proven to be extremely useful in describing the genetic structure of field and laboratory isolates [30] , [31] , [32] , [33] , [34] . However , little is known about the genetics of PZQ susceptibility [35] , [36] . Understanding how drug pressure affects the genetic structure of a parasite population in a controlled setting raised questions relevant to the endemic setting related to transmission , chemotherapy effects , and epidemiology of the disease [35] , [36] , [37] . Studies of how Schistosoma populations at endemic sites perform under drug pressure are necessary for determining the most efficient treatment strategy for infected populations and for the design and modeling of control programs [38] , [39] . Because PZQ is the drug of choice being used in the population control of schistosomiasis , studies about the genetics of populations under drug pressure and new drug targets are urgently needed . In this paper , we hypothesize that in a mouse system , as a result of a regimen of exposures to increasing amounts of PZQ , the S . mansoni population will experience a decrease in genetic diversity and susceptibility to the drug .
In vivo studies were conducted in compliance with the guidelines of the Collegiate of Animal Experimentation ( COBEA ) and approved by the Commission on the Ethical Use of Animals ( CEUA-FIOCRUZ ) protocol number L-018/09 approved on Jan . 23 , 2009 . The S . mansoni ( LE strain ) life cycle was maintained using Biomphalaria glabrata snails as intermediate hosts and Swiss mice as definitive hosts according to Pellegrino and Katz [40] . Female albino mice weighing approximately 20 g were infected with 100±10 cercariae of S . mansoni ( LE strain ) by a subcutaneous route . To induce decreased susceptibility , increasing doses of an aqueous suspension of PZQ were administered to infected mice by oral gavage . The protocol used for parasite selection was based on Fallon and Doenhoff [26] with one modification: in the third treatment , mice were given two doses of 200 mg/kg . The dose of 200 mg/kg PZQ was used to ensure that some of the parasites survived drug treatment for subsequent passages until the highest dose of 3×300 mg/kg was reached . The first and second rounds of treatment started with two subsequent doses of 100 mg/kg on days 28 and 35 after infection . The eggs produced by the worms that survived drug treatment were used to infect snails , which , after 35 days , started to eliminate cercariae that were used to infect mice . For the third and fourth rounds of treatment , subsequent doses of 2×200 mg/kg on days 28 and 35 were used . At the fifth treatment round , mice were treated with two doses of 250 mg/kg and after the sixth round , treatment was maintained with three doses of 300 mg/kg on days 28 , 35 and 37 until the treatment rounds were completed . Five weeks after each round of PZQ treatment the mice were portally perfused and the surviving worms recovered were counted . The worm population obtained after seven successive treatment rounds was considered the Selected population ( S ) . For each round of PZQ treatment the same procedure was used in a group of mice infected with cercariae produced by snails from the Non-selected population ( N ) . Adult worms in the Selected and Non-selected groups were retrieved by perfusion of the portal system with 0 . 85% saline solution and 1% heparin [41] . Four groups of 20 mice were used for each cycle . Two groups ( Selected and Non-selected populations ) were left untreated and animals were portally perfused 45 days after infection to obtain an infection control worms count . The other two groups were treated as described above and mice were sacrificed five weeks after treatment . For DNA extraction of individual parasites , worm pairs were separated and digested in extraction buffer ( 50 mM Tris-HCl pH 8 . 0 , 100 mM EDTA , 100 mM NaCl and 0 . 5% SDS ) and 20 µg/ml of proteinase K ( Sigma ) and incubated for 12 hours at 37°C , followed by incubation at 95°C for 5 min . Nucleic acids were precipitated with cold absolute ethanol and resuspended in 30 µl of TE buffer and rehydrated for 12 hours at 4°C . We analyzed five polymorphic microsatellite loci for each population . Allelic and genotypic data were identified for a total of 250 worms . PCR reactions and scoring of the polymorphic microsatellites were conducted as previously described [42] , [43] . Primer pairs and genomic locations are described in Table 1 . Genetic data was analyzed with the Fstat software version 2 . 9 . 3 . 2 [44] . For each locus , allelic and genotypic frequencies , linkage disequilibrium among polymorphic loci , the inbreeding coefficient , and observed and expected heterozygosity values were calculated . Hardy-Weinberg ( HW ) equilibrium was calculated using the Arlequin software [45] . Population structure was estimated for the isolates using the fixation indices Fst [46] , Fis [47] and Rst [48] . The Guo and Thompson test [49] was used to evaluate significance of deviations of the observed and expected heterozygosity between Selected and Non-selected worms and to assess the difference between genetic variability of male and female worms ( P≤0 . 05 ) . Statistical analysis of parasitological data was conducted with the SPSS software version 11 . The mean number of worms recovered after selection or lack of selection by PZQ was compared using the Mann-Whitney and Kruskall-Wallis tests . To evaluate the significance between female and male worm burdens , the Wilcoxon test was used . P-values≤0 . 05 were considered to be statistically significant . The percent reduction of worm burden in each treatment group was calculated according Cioli et al . [23] . Approximately 200 worms each from treated and untreated mice were used to produce the genetic data .
PZQ selection resulted in a striking increase in the number of recovered parasites after treatment with this drug , compared with the number recovered after equivalent treatment of non-selected parasites . Thus , after seven parasite generations of selection using doses of PZQ that increased from 200 mg/kg to 900 mg/kg , we observed a greater than three-fold increase in the degree to which the proportion of worm load in the Non-selected group had been reduced by drug treatment ( 47 . 3% ) , compared with the level of reduction in the Selected group ( 14 . 58% - Table 2 ) . These findings are in agreement with previous observation of laboratory selection reported by Fallon and Doenhoff [26] . Female adult worms were shown to be more susceptible than males to PZQ exposure . We observed that after exposure to 3×300 mg/kg of PZQ , both male and female worms showed a significant decrease in susceptibility . Table 3 shows the results obtained after treatment of infected mice with 3×300 mg/kg doses of PZQ after 11 passages , indicating the increased sensitivity of female adult worms in comparison to male worms . After treatment , the male to female ratio changed from 2 . 5 in the control population to 8 . 7 in the Selected population . A decreased recovery of males was observed by Delgado et al . [50] . It has been observed that tegumental damage induced by PZQ is more apparent in male worms , while the female tegument displays limited damage [51] . Pica-Mattoccia and Cioli [52] observed that immature females are refractory to PZQ treatment , but upon sexual maturation they become sensitive to PZQ . The same researchers showed that in vitro males displayed higher sensitivity to PZQ . These experiments differ from ours in that we assessed PZQ sensitivity in vivo . It has been suggested that the environment in which the parasite is located affects the parasite physiology and drug sensitivity [52] . This is especially relevant for females living in the gynecophoric canal of males , where they may be more protected from PZQ activity . It has also been reported that resistance can lead to decreased reproductive fitness in schistosomes [53] , [38] . Previous laboratory studies show that selection can rapidly change infectivity , virulence phenotypes , and population structure of schistosomes after a few generations [34] . The decrease in the female population was so striking after 11 passages that it resulted in the impossibility to maintain the strain in the laboratory . Our study shows that in a short time PZQ can select against female parasites , and such variations may substantially influence transmission dynamics , thereby decreasing the spread of decreased susceptibility to PZQ . PZQ-mediated selection led to decreased allelic diversity in S . mansoni . The number of alleles found in the Non-selected population decreased in the Selected population from 98 to 42 . The number of genotypes found in the Non-selected population and the Selected population decreased from 265 to 112 . The allele and genotype number for each locus decreased after successive treatments . The Selected population ( S ) showed a lower total number of alleles per genotype locus than that of Non-selected parasites ( Table 4 ) . It was clearly demonstrated that the selection process decreased parasite genetic diversity . The degrees of population structuring and diversity among Selected and Non-selected populations were evaluated by genetic diversity analysis ( Rst ) and Inbreeding Coefficient ( Fis ) . Our results show that there is a significant genetic differentiation between the first and seventh generations of the S population ( Rst = 0 . 240 ) . In contrast , without PZQ pressure , we did not observe a significant differentiation in the Non-selected populations ( Rst = 0 . 080 ) . Genetic diversity data reveal that decreased susceptibility selection contributes to the decrease of genetic variability of parasites selected with PZQ . Next , we measured the level of inbreeding between the Selected and Non-selected populations . According to Wright [47] , a Fis value of 0 . 309 indicates a high level of inbreeding in the Selected population . The Fis index shows that there is a strong increase in endogamy in the Selected population ( 0 . 134 to 0 . 309 ) . Our results with both indexes suggest that the Selected S . mansoni population is less genetically diverse as a result of PZQ pressure . These results agree with a report by Norton and colleagues [36] , which showed that in an endemic setting a single round of mass drug administration to a human population produced a genetic bottleneck on S . mansoni . We did not observe significant differences in the genetic variability of male and female worms ( data not shown ) nor in populations passaged in parallel but not submitted to drug pressure . The analysis of Hardy-Weinberg equilibrium deviation is an important resource for the study of the evolutionary mechanisms that can act upon a given population . The comparisons between the expected and observed heterozygosities in Selected and Non-selected populations show that the observed heterozygosity was lower than expected for all polymorphic loci except SmBr 9 in the N strain ( Table 5 ) . HW equilibrium deviation occurred in the majority of analyzed loci along with the loss of observed heterozygosity in the S and N populations . The analysis of the genetic structure of the N and S populations using the Hardy-Weinberg coefficient agrees with the genetic diversity index and endogamy coefficient and indicates decreased genetic diversity in the population under drug pressure . The maintenance of the S . mansoni LE strain by PZQ-mediated selective pressure contributed to the decrease in variability of the S population . The genetic bottleneck and subsequent loss in genetic diversity were also observed in parasites found at endemic sites that had received PZQ treatment [36] . One explanation for the observed deficiency in heterozygosity in the Selected population could be selective pressures suffered by this strain after seven generations under drug pressure [39] . Taken together , these results may indicate that at endemic sites , there may be a larger pool from which genetically resistant strains may be positively selected under drug pressure . In this paper , we clearly demonstrate the effects of PZQ selection on populations displaying decreased susceptibility to PZQ . We observed a decrease in the genetic diversity of the parasite . In addition , the apparent cost of coping with large doses of PZQ is an imbalance in the male/female ratio . In the laboratory , the consequence was the difficulty of maintaining the strain . Although PZQ may cause changes in the genetic structure of treated populations , our results suggest a slow spread of drug resistance because after 11 rounds of treatment we recovered almost exclusively female worms . Levels of genetic diversity , as opposed to genetic divergence between populations , may itself be an important component of the epidemiology of infection and disease , as well as a key indicator for monitoring the effects of selection imposed by drug treatment ( 36 ) . We speculate that in the field , one possible reason for resistance or decreased susceptibility not being widespread is perhaps its biological cost . The creation of a genetic map constructed using microsatellites , combined with the future development of a SNP map , means that genetic tools will be invaluable in both investigating the genetic profile of populations and identifying markers or genes that convey resistance .
|
Schistosomiasis causes severe health problems and economic losses in transmission areas in the Americas , Africa , and Southeast Asia . In regions where treatment is available , such as Brazil , chemotherapy has led to decreased morbidity of the disease . However , parasite resistance to drugs is of general concern . Drug resistance is known in some species of helminths , and there are reports of decreased susceptibility in schistosomes , although such reports are not widespread . The impact of consecutive drug treatment on parasite populations at transmission sites can be evaluated by the genetic makeup of the population . For example , the decrease in the genetic diversity of these populations upon chemotherapy can be interpreted as the selection of less sensitive parasites . In this study , we used a laboratory mouse model of schistosomiasis that was treated with increasing amounts of praziquantel ( the drug of choice for human treatment ) until we reached the highest dose of 3×300 mg PZQ/kg . Our results clearly indicate that the parasite population showed a decrease in genetic diversity after measuring allelic and genetic diversity . This observation confirms that drug selection occurred in this population . The work highlights the need for genetic monitoring of populations under drug pressure , resistance mechanisms , and the dynamics of parasite populations upon treatment .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"and",
"Discussion"
] |
[] |
2013
|
Praziquantel Treatment Decreases Schistosoma mansoni Genetic Diversity in Experimental Infections
|
Canine malignant melanoma , a significant cause of mortality in domestic dogs , is a powerful comparative model for human melanoma , but little is known about its genetic etiology . We mapped the genomic landscape of canine melanoma through multi-platform analysis of 37 tumors ( 31 mucosal , 3 acral , 2 cutaneous , and 1 uveal ) and 17 matching constitutional samples including long- and short-insert whole genome sequencing , RNA sequencing , array comparative genomic hybridization , single nucleotide polymorphism array , and targeted Sanger sequencing analyses . We identified novel predominantly truncating mutations in the putative tumor suppressor gene PTPRJ in 19% of cases . No BRAF mutations were detected , but activating RAS mutations ( 24% of cases ) occurred in conserved hotspots in all cutaneous and acral and 13% of mucosal subtypes . MDM2 amplifications ( 24% ) and TP53 mutations ( 19% ) were mutually exclusive . Additional low-frequency recurrent alterations were observed amidst low point mutation rates , an absence of ultraviolet light mutational signatures , and an abundance of copy number and structural alterations . Mutations that modulate cell proliferation and cell cycle control were common and highlight therapeutic axes such as MEK and MDM2 inhibition . This mutational landscape resembles that seen in BRAF wild-type and sun-shielded human melanoma subtypes . Overall , these data inform biological comparisons between canine and human melanoma while suggesting actionable targets in both species .
Human melanoma is of increasing clinical concern . It is one of a few cancers with rising incidence , while five-year survival for patients with metastatic disease has until recently remained low ( 15–20% ) due to a dearth of curative systemic therapies [1] . Discovery of frequent activating BRAF mutations in melanoma and treatment with selective inhibitors of this mutant kinase has led to dramatic responses in the setting of metastatic disease [2–4] . However , not all BRAF-mutant melanomas respond to targeted therapy and responses that do occur are often brief and followed by the emergence of drug-resistant disease [5] . Moreover , targeted treatment options in melanoma subtypes without activating BRAF mutations are limited . New treatment paradigms such as immunotherapy , drug combinations , and alternative dosing strategies may circumvent resistance and broaden the scope of precision medicine in melanoma [6–9] , but rapid preclinical study of such regimens requires access to robust models that recapitulate complex tumor features such as intratumoral genomic heterogeneity and tumor-host interactions . Meanwhile , few animal models exist for uncommon molecular or histological melanoma subtypes such as BRAF wild-type ( BRAFwt ) or mucosal melanoma . Naturally-occurring canine cancers are increasingly recognized as meeting a need for complex cancer models that develop gradually amidst interactions with the host stroma and immune system [10–16] . Spontaneous canine malignant melanomas , which are almost universally BRAFwt and for which the mucosal subtype is the most prevalent clinically significant form , may fill a specific gap in models of BRAFwt and rare histological melanoma subtypes [11] . Human mucosal melanoma is an aggressive histological subtype that is predominantly BRAF , RAS , and NF1 wild type ( Triple Wild Type or TWT ) with occasional mutations in KIT or NRAS . It carries a five-year survival rate between 12 . 3% and 35 . 3% [17–26] . Study of this subtype is limited by its low prevalence , accounting for only 1–2% of human melanomas in the United States with as few as 1 , 500 cases per year [27] . On the other hand , canine malignant melanoma accounts for up to 100 , 000 yearly cancer diagnoses in the United States , occurring most commonly in the oral mucosa , but also arising in cutaneous and acral epithelium [28–31] . Canine malignant melanoma is highly prevalent , closely mirrors human melanoma clinically and pathologically , and is extremely aggressive , with median survival for oral cases being a mere 200 days[32–36] . However , little is known about its genetic etiology . It is predominantly BRAFwt with frequent copy number alterations of regions of canine chromosomes ( CFA ) 13 , 17 , 22 , and 30 , alongside frequent MYC amplifications and deletions of CDKN2A . Targeted sequencing studies , though limited , have shown that it infrequently bears alterations in other known drivers of human melanoma [32 , 36–42] . It has been shown that CFA30 aberrations are characteristic of canine oral melanoma and complex copy number profiles on this chromosome homologous to the same profiles on human chromosome ( HSA ) 15 in human mucosal melanoma are suggestive of rearrangements that may drive this melanoma subtype [41] . Despite the very low prevalence of BRAF mutations , immunohistochemistry ( IHC ) has shown that the mitogen-activated protein kinase ( MAPK ) and/or phosphoinositide 3-kinase ( PI3K ) pathways are activated in 52–77% of cases [32 , 36–40] . These data hint at underlying mutations driving these pathways that could guide future biological exploration and therapeutic development in the canine and human diseases . We therefore set out to map the genomic landscape of canine melanoma using a combination of massively parallel whole genome sequencing ( WGS ) , array-based platforms and targeted sequencing to identify somatic changes driving these cancers . Here we report the identification of recurrent inactivating mutations in the candidate tumor suppressor gene PTPRJ in addition to frequent RAS mutations , and mutually-exclusive MDM2 and TP53 alterations . We thereby define the genomic landscape of these cancers and identify similarities between melanoma subtypes across species while highlighting subtype-specific aberrations that may be used to guide future research .
We undertook comprehensive analysis of acquired genetic alterations in a discovery cohort of seven melanomas and matched germlines from six dogs ( two tumors were derived from one dog ) using WGS for detection of subtle sequence alterations alongside long-insert WGS ( LI-WGS , see Materials and Methods ) [43] for sensitive detection of structural variants . We then performed copy number and targeted gene analyses in an additional 27 canine malignant melanoma tumors and three canine malignant melanoma cell lines ( Table 1 ) . Snap-frozen tumors ( all primary tumors except one acral metastasis ) and matching whole blood were collected through an IACUC-approved protocol at the Van Andel Research Institute ( VARI ) from dogs undergoing surgery at 21 specialty veterinary clinics located in 10 states ( see Materials and Methods ) . Diagnosis of melanoma was confirmed by two independent board-certified veterinary pathologists ( an on-site pathologist and BD ) in addition to staining for three melanocytic differentiation markers where tissue was available ( in 26 samples as indicated in S1 Table ) [36 , 44] . Diverse breeds are represented in this cohort with enrichment for Cocker Spaniels and Golden Retrievers ( five dogs of each breed ) , an equal ratio of male and female dogs and a median age at resection of 11 years . Clinicopathologic characteristics for this cohort are described in S1 Table and S1 Fig . For WGS and LI-WGS respectively a median of 38/11-fold sequence coverage and 209/155-fold physical coverage was achieved ( S2 Table ) . Read alignment was performed using the canine reference genome CanFam 3 . 1 and stringent criteria were used to call somatic sequence variants intersecting Seurat v2 . 6 , Strelka v1 . 0 . 13 , and Mutect v1 . 1 . 4 ( Materials and Methods ) . A total of 31 , 053 somatic single nucleotide variants ( SNVs ) and small insertions and deletions ( indels ) were found with a median of 4 , 223 genome-wide SNVs ( range 1 , 880–6 , 342 ) and 316 indels ( range 88–655 ) and a median mutation rate of 2 . 03 mutations per callable haploid megabase ( range 0 . 97–3 . 14 , Table 2 ) . The genome-wide SNV spectrum showed C:G>T:A transitions to be most prevalent , at a median of 27 . 09% of total SNVs followed by T:A>C:G transitions ( median of 21 . 19% ) and C:G>A:T transversions ( median 15 . 74% , S2 Fig ) . Despite the prevalence of C:G>T:A transitions , most occurred in CpG dinucleotides and were not enriched at dipyrimidines ( median 22 . 5% ) . Therefore , a canonical UV signature was not present in any of these cases ( S2 Fig ) [45 , 46] . We additionally looked for TERT promoter mutations , which have been reported in 71% of human cutaneous melanomas and are associated with UV damage [47] , but no somatic mutations were found within one kilobase of the TERT transcription start site . The most common mutation overall was C:G>T:A in GCG trinucleotides ( median 3 . 29% ) followed by C>T in ACG ( median 2 . 6% ) and C>A in TCT ( median 2 . 5% ) ( S2 Fig ) . No evidence of localized hypermutation ( kataegis ) was identified in these tumors [48] . Tumors assessed by whole-genome analysis displayed an abundance of somatic structural variants ( SVs ) and copy number variants ( CNVs ) , with a modest burden of SNVs in coding regions ( Fig 1A and 1B ) . The landscape of somatic mutations in the full cohort of 37 tumors based on multi-platform analysis is shown in Fig 1C . Circos plots depicting somatic alterations in each tumor in the discovery cohort are shown in S3 Fig . Of the genome-wide SNVs described above , a median of 26 nonsynonymous ( ns ) single-base substitutions and indels occurred within coding regions ( nsSNVs , range 14–42 ) with a median nonsynonymous: synonymous mutation ratio of 2 . 3 ( range 1 . 9–3 . 9 ) ( Fig 1B ) . We additionally performed RNA sequencing in this cohort , aligning with STAR2 . 4 [49] , calling SNVs with HaplotypeCaller ( GATK 3 . 3 . 0 ) , and utilizing IGV to manually validate expressed sequence variants ( Materials and Methods ) . Ninety-seven percent of nsSNVs ( all but five ) identified by WGS and genotyped on more than one sequencing platform were confirmed in at least one additional platform ( S3 Table ) . A number of mutations in orthologs of human cancer genes were present in a single tumor each . These genes include: ATF6 , EPAS1 , FAT2 , FAT4 , FOXA3 , FOXO1 , GAB2 , HRAS , KIT , KRAS , MMP21 , NRAS , PBX1 , and XPO1 . Although no recurrent SNVs were seen in the discovery cohort , three genes were mutated in two cases: FAT4 , LRFN2 , and PTPRJ . Of these , only PTRPJ was validated in multiple platforms in both cases . Both cases containing somatic PTPRJ mutations were mucosal ( ND10-166 and ND10-376 ) and both putatively contained two hits . To determine the prevalence of mutations in a panel of genes whose orthologs are known to play a role in human melanomagenesis , as well as the PTPRJ gene mutated in two cases , we performed targeted Sanger sequencing of all protein-coding regions of BAP1 , BRAF , CDK4 , GNA11 , GNAQ , KIT , KRAS , MDM2 , MITF , NF1 , NRAS , PTEN , PTPRJ , and TP53 in the expanded cohort . BRAF , CDK4 , GNAQ , MDM2 , MITF , and NF1 were all found to be universally wild-type whereas putative pathogenic mutations were discovered in BAP1 , GNA11 , KIT , KRAS , NRAS , PTEN , PTPRJ , and TP53 as described below and in S4 Table . Somatic CNVs in the discovery cohort were identified by analysis of short-insert whole genome sequencing ( SI-WGS ) using established methods ( Materials and Methods ) . A median of 27 focal CNVs ( range 4–68 ) , two focal amplifications with a log2 ratio ≥ 2 ( range 0–61 ) , and eight focal deletions with a log2 ratio ≤ 0 . 2 ( range 3–41 ) were identified in the discovery cohort ( Table 2 and S5 Table ) comprising 0%-10% of the genome ( Table 2 ) . CNVs were additionally identified in this cohort utilizing Illumina CanineHD BeadChip Single Nucleotide Polymorphism ( SNP ) arrays and Agilent SurePrint G3 Canine Genome CGH microarrays as previously described [41 , 50] ( Materials and Methods ) with a high platform concordance ( S4 Fig ) . CNV analysis was then expanded to a total of 37 melanomas through SNP arrays in an additional 30 cases in the prevalence cohort ( Table 1 and S5 Table ) . Altered regions were assessed by GISTIC [51] for statistically significant frequency and amplitude ( G-score >1 . 0 and Q<0 . 05 ) . Ten significant regions were identified including losses within CFA 1 , 11 , 15 , and X , as well as gains in CFA10 , 11 , 13 , 30 , and X ( S6 Table ) . Nine of 10 GISTIC regions contained genes and included gains in orthologs of the human cancer genes MDM2 and CDK4 . Additional cancer driver alterations ( homozygous deletions of tumor suppressor genes or focal amplifications of oncogenes ) included CDKN2A homozygous deletion ( 3% ) and KIT focal amplification ( 8% ) ( S7 Table ) . Somatic SVs including translocations , inversions , and duplications , were identified in the discovery cohort , based on calls from Delly v0 . 7 . 6 [52] in LI-WGS ( Materials and Methods ) . Between 9 and 65 predicted SVs were identified in each tumor ( median 34 ) and were predominantly inversions ( Table 2 and S8 Table ) . No recurrent rearrangements were present . Notable alterations in human cancer gene orthologs impacted by SVs in single cases include an ARHGEF12 inversion , a BIRC3 inversion , a CLPTM1L-TERT translocation , a DDIT3 inversion , a MYO5A translocation , and a TCF12 inversion . However , two regions of CFA10 and 30 were found to contain somatic SVs in two or more tumors . CFA10 rearrangements occurred in five of seven cases , four of which bore alterations in the region spanning 10–12 Mb ( also a significant GISTIC region from CNV analysis ) . CFA30 SVs were also present in three tumors with alterations occurring within a region spanning 15–24 Mb ( also encompassing a GISTIC region ) in each case . Complex chromosomal rearrangements reminiscent of chromothripsis were observed in four tumors ( ND09-345 , ND10-370 , ND10-361 , and ND10-441 ) , with chained or clustered breakpoints localized to a subset of chromosomes in regions that also contained copy-number oscillations [53] ( S3 Fig ) . Approximately 90% of human cutaneous melanomas are driven in part by BRAF , RAS , NF1 , and KIT mutations that confer constitutive mitogenic signaling through the MAPK pathway [24 , 45 , 54] . However , these alterations are far less common in human mucosal and acral melanomas [20 , 22 , 23 , 55–57] . No somatic alterations in BRAF were identified within any platform in our canine melanoma cohort . However , RAS family members , whose protein products are predicted to share 100% sequence identity with their human orthologs , were the most commonly mutated genes in aggregate , occurring in 24% of cases in human-conserved hotspots ( Figs 1C and 2A ) . NRAS codon 61 ( Q61R/H/K ) and KRAS codon 12 ( G12C ) mutations occurred each in four cases while a single case bore an HRAS Q61R mutation ( nine total RAS mutations ) . All three acral and two cutaneous cases bore NRAS or KRAS mutations , while only 4/31 ( 13% ) of mucosal cases bore an NRAS , KRAS , or HRAS mutation . Although NF1 copy number losses occurred in six cases , no homozygous deletions or truncating mutations were identified ( S7 Table ) . KIT mutations were present in one cutaneous and two mucosal tumors ( S3 and S4 Tables ) . In the cutaneous case , the mutation results in a glutamine ( Q ) to arginine ( R ) change in codon 396 , notably a site of variation between canine and human orthologs , a change that is not predicted to be damaging by PROVEAN , and may constitute a germline SNP , but germline DNA was not available in this case [58] . KIT mutations in the mucosal cases included an in-frame deletion of amino acids 560–562 , a likely damaging mutation in a commonly mutated region of the human ortholog , as well as an aspartic acid ( D ) to valine ( V ) change in codon 815 corresponding to the most common hotspot D816V mutations occurring in the kinase domain of KIT in human cancers ( S5 Fig ) [59] . Copy number gains encompassing KIT were also present in 10 samples ( eight mucosal , one acral , and one cutaneous–Jones , 17CM98 , ND10-104 , ND10-158 , ND10-365 , ND10-370 , ND10-376 , ND10-361 , ND10-363 , and ND10-441 ) , although no focal amplifications were identified ( S7 Table ) . The most commonly mutated gene in this cohort was the putative tumor suppressor gene PTPRJ , not previously shown to have frequent inactivating point mutations in cancer ( Figs 1C and 2C ) . PTPRJ ( also known as density-enhanced phosphatase 1 ( DEP-1 ) or CD148 ) is a protein tyrosine phosphatase receptor originally discovered by virtue of its overexpression in dense cultures of human lung fibroblasts [60] . It has since been shown to be frequently involved in allelic loss or loss of heterozygosity ( LOH ) in human cancers and mouse models [61 , 62] and to potentially play a role in oncogenesis in diverse cancer types , but somatic homozygous deletions or truncating mutations have yet to be described in cancer from any species and its tumor suppressor status remains controversial [63–71] . Canine and human orthologs share 70% sequence identity with a highly conserved C terminus containing the protein tyrosine phosphatase catalytic domain that is nearly 100% identical between species ( S6 Fig ) . Sequencing of PTPRJ across all 37 tumors revealed nine mutations in seven cases ( all mucosal ) , comprising 19% of all tumors and 23% of mucosal cases . Six frameshifts or stop gains were discovered in addition to two splice site mutations , a C-terminal 10-amino acid deletion , and a single predicted damaging missense mutation . Two cases–ND10-190 and ND10-376 –contained two mutations each , consistent with putative bi-allelic inactivation of a tumor suppressor gene . Further , LOH was evident by analysis of adjacent SNPs in WGS data in case ND10-166 bearing the M110fs mutation ( S9 Table ) . Consistent with this finding , the PTPRJ frameshift in the ND10-166 tumor occurred at an allele ratio of 61% in DNA and 100% in RNA . Finally , PTPRJ transcript was observed in RNAseq data from the two PTPRJ-mutant tumors profiled by WGS and RNAseq ( 270 . 21 Fragments Per Kilobase of transcript per Million mapped reads ( FPKMs ) in ND10-166 and 92 . 37 FPKMS in ND10-376 ) as shown in S7 Fig . ND10-376 , containing two somatic PTPRJ mutations ( a frameshift and a splice site mutation ) and 92 . 37 FPKMs , bore the lowest transcript abundance among all seven profiled tumors . Median PTPRJ FPKMs for the five PTPRJ wild-type tumors was 171 . 76 ( range 92 . 37–512 . 25 ) . Inactivation of the p53 network is a critical step in tumorigenesis in nearly all cancers [72] . Both truncating TP53 mutations and amplifications of MDM2 , a negative regulator of p53 , are key routes to p53 inactivation [73] . Although TP53 mutations and MDM2 amplifications in human melanoma are less common [23–25 , 45 , 54 , 56] , 16/37 ( 43% ) of the cases in our cohort of canine melanoma bore focal amplifications of MDM2 or truncating TP53 mutations ( Fig 1C ) . A recurrent focal amplification on CFA10 was identified by whole genome analysis in three of seven tumors in the discovery cohort with extended SNP array analysis in the prevalence cohort revealing an additional eight tumors bearing these amplifications ( minimal region 10 . 9–11 . 8 Mb ) ( Figs 1C and 2C ) . In total , 11/38 cases ( 29% ) bore this amplification involving seven genes , with MDM2 being the likely amplification target ( Fig 2B ) . All such amplifications occurred in mucosal melanomas ( 11/31 , 35% ) . CDK4 , a cancer gene 10 Mb proximal to MDM2 in both human and canine genomes and often the target of bipartite amplification alongside MDM2 [74 , 75] , was co-amplified in three of these cases . Identification of focal MDM2 or CDK4 amplification in the WGS- and RNA-sequenced cohort coincided with high transcript abundance for these genes relative to their wild-type counterparts ( S7 Fig ) . MDM2 FPKMs were 357 . 48 and 331 . 21 for the amplified cases ( ND09-345 and ND10-370 ) relative to a median of 67 . 02 ( range 37 . 57–82 . 26 ) for wild-type cases . CDK4 FPKMs were 2 , 730 . 13 for the amplified case ( ND09-345 ) versus a median of 201 . 24 ( range 69 . 87–471 . 2 for the wild-type cases ) . Additionally , twenty tumors were additionally assessed for MDM2 expression by IHC ( S10 Table and S8 Fig ) . Three of five cases with MDM2 focal amplifications also showed prominent MDM2 staining while no cases lacking MDM2 amplifications were positive by IHC . We additionally discovered seven tumors with mutations in TP53 whose protein product shares 80% identity with its human ortholog ( S9 Fig ) . Three of these mutations were truncating–a homozygous T90X in ND10-252 , heterozygous K151fs in ND11-201 , and a heterozygous Q306X in ND10-564 ( Fig 2D and S4 Table ) . Of the three missense mutations , R145C and R270H were predicted to be damaging . R145C occurred in two tumors and R270H in a single tumor , with both mutations confirmed somatic through analysis of matched germline DNA . Codon 270 in canine TP53 is homologous to codon 282 in human TP53 , the fifth most common hotspot for mutations in human cancer[59] . The missense G290R variant is a likely SNP . It occurs in a tumor for which matched germline DNA is unavailable and it is predicted to be neutral , although it has not been previously described [76–78] . In keeping with findings in other cancers , no sequence mutations were present in MDM2 and MDM2 amplifications were mutually exclusive with TP53 mutations . Further , TP53 and MDM2 alterations were mutually exclusive with RAS mutations in all but one case ( ND10-748 , Fig 1 ) . Common genomic subtypes of human cutaneous melanoma ( BRAF , RAS ( N/H/K ) , and NF1 in 90% of cases ) that engage oncogenic signaling through the MAPK pathway are less common in human non-cutaneous melanoma and in canine malignant melanoma ( 24% of cases here , Fig 1C ) . Therefore , to undertake unbiased identification of pathways contributing to canine melanomagenesis , we performed pathway analysis using WGS data from the discovery cohort . We generated a list of all genes bearing nonsynonymous mutations , lying within chromosomal breakpoints or significant CNV regions from GISTIC ( n = 1047 ) in order to determine enrichment of these mutated genes within specific KEGG and Reactome pathways ( Materials and Methods ) [79–81] . Network analysis of the affected genes identified 97 pathways with significant Benjamini-Hochberg corrected P-values ( S11 Table ) . The most significantly enriched pathways were Insulin Receptor Substrate ( IRS ) -mediated signaling , and IRS-related events , for which 23% ( 19 genes ) of the pathway members are mutated in this cohort . Notably , these pathways converge on MAPK and PI3K mitogenic signaling and contain core pathway members such as FGFs , EIF4G1 , HRAS , KRAS , NRAS , and RPTOR . Indeed the majority of the enriched pathways contain members of MAPK , PI3K , or growth factor receptor signaling ( e . g . PI3K cascade P = 0 . 002 , mTOR signaling P = 0 . 008 , signaling by Rho GTPases P = 0 . 012 , VEGF signaling P = 0 . 017 , RAF activation P = 0 . 017 , melanoma signaling P = 0 . 021 , RAS signaling P = 0 . 031 , and MEK activation P = 0 . 036 ) and , in many cases , intersections with MDM2 signaling .
Melanoma is a clinically significant disease in dogs , the study of which holds untapped potential for developing clinical approaches to improve the lives of pet dogs while also informing human melanoma biology and treatment . Few treatment options are available for locally advanced or metastatic canine melanoma in part because the molecular etiology is still largely unknown . Similarly , limited molecular understanding of rare sun-shielded and BRAFwt subtypes of human melanoma has constrained clinical innovation . In order to identify the molecular alterations underlying canine melanoma , we undertook a comprehensive multi-platform genomic investigation . Our integrated analysis confirms that although these tumors are driven by mutational landscapes distinct from those in human cutaneous melanoma , they share important similarities with BRAFwt and rare histological subtypes of human melanoma . These data not only guide biological and therapeutic studies in canine melanoma , but they also lend further support for the use of the naturally occurring canine model in comparative studies of human cancers . This study builds on knowledge of the cytogenetic landscape of canine melanoma [41] to provide a comprehensive view of numbers and types of somatic coding mutations in this cancer . Given the dearth of genomic data for canine melanoma , we focused overall on collecting primary tumors from diverse breeds . While this study was not sufficiently powered to draw conclusions regarding breed associations with somatic mutations , it is nonetheless important to consider potential associations between breed and somatic mutational landscapes , particularly because such associations have been shown to occur in other canine cancers such as lymphoma [82] . Several breeds have been suggested to be at increased risk for malignant melanoma , particularly breeds with heavily pigmented skin or oral mucosa such as Cocker Spaniels , Schnauzers , Scottish Terriers , Poodles , Chow Chows , and Golden Retrievers [83] . Our WGS discovery cohort primarily consisted of Cocker Spaniels ( four Cocker Spaniels , one English Cocker Spaniel , and one Labrador ) , a breed reported to be at higher risk of oral melanoma , but our extended cohort then included targeted sequencing of 13 melanoma hallmark genes ( as well as PTPRJ , which was the only additional recurrently mutated gene in the WGS cohort ) and copy number assessment from SNP arrays across 20 total breeds . Given that our WGS cohort was predominantly Cocker Spaniel , it is possible that other recurrent , breed-specific somatic SNVs in non-melanoma-hallmark genes could exist that were not captured here . Thus , future expanded study of breed-specific cohorts will be critical for further understanding the role of germline variation in shaping somatic cancer landscapes across species . It will also be important to further define subtype differences in expanded cohorts of canine acral and cutaneous tumors as well as benign and precursor lesions . Overall , the genomic landscapes of human melanoma vary by anatomic site and degree of sun exposure [22 , 26 , 57] . Cutaneous sun-exposed melanoma is characterized both by high point mutation frequencies linked to UV damage [45] and also only modest burdens of structural variation . In contrast , sun-shielded and non-cutaneous melanomas harbor a low point mutation , but high structural mutation burden . Here , we establish that the canine malignant melanoma genome landscape resembles that reported in human sun-shielded melanoma . Canine melanoma of all subtypes in our discovery cohort is likely sun-shielded , including cutaneous tumors which occur in densely hair-bearing skin , although cropping or shaving during summer months may in some cases increase UV exposure . In keeping with this status , WGS in these two canine cutaneous malignant melanoma cases provides a deep view of their genome-wide mutation burden revealing low point mutation frequencies ( median 2 . 03 somatic mutations per Mb ) similar to that seen in human acral and mucosal melanoma WGS data from Hayward et al . 2017 ( Fig 3A ) [26] . Although we only profiled two such cases and larger cohorts are needed , a low point mutation burden relative to human sun-exposed melanoma has potential bearing on expected responses to immunotherapy such as anti-CTLA4 and anti-PD1 checkpoint blockade . Numerous studies have shown a clear positive correlation between mutation burden , abundance of neoantigens , and clinical benefit in human melanoma and other cancers [84 , 85] . Nonetheless , other molecular determinants of response to immunotherapy exist beyond simply mutation burden and the activity of such agents in canine malignant melanoma remains to be determined . Notably , CNV and SV burden from our WGS in canine malignant melanoma was markedly lower than all subtypes as described in Hayward et al . ( Fig 3B and 3C ) [26] . WGS additionally provides a deep view of genome-wide mutation signatures . High point mutation burden in sun-exposed cutaneous melanoma is understood to result from UV-induced over-representation of C>T transitions occurring in dipyrimidines versus non-dipyrimidines . UV-induced C>T mutations occurring in dipyrimidines comprise a low proportion of total SNVs in our cohort ( 25% ) , reflective of human sun-shielded cutaneous , mucosal and acral melanoma , in contrast to 85–90% of C>Ts occurring in dipyrimidines in human sun-exposed melanoma ( Fig 3C ) [24 , 26 , 45 , 55 , 56 , 86] . This lends support for a non-UV etiology of canine melanoma . The genome-wide SNV spectrum further revealed that C>T transitions in CpGs were the most common sequence alterations ( S2 Fig ) . These mutations correlate with age in human cancers and are due to spontaneous deamination of 5-methylcytosine [46] . Enrichment for these mutations in canine melanoma is not surprising given that the largest risk factor for cancer in humans and dogs is biological ( not chronological ) age [87–92] and that the mean age of these dogs at the time of surgical resection was 13 years ( range: 10–16 ) . Relative to the average number of human somatic mutations , these data provide further evidence that not only cancer incidence , but also mutational burden increases with biological , rather than chronological , age [93] . Commonly observed mutational patterns in human melanoma such as kataegis were not observed , although four tumors exhibited clustered or chained translocations suggestive of breakage-fusion-bridge events due to telomere crisis or of chromothripsis , in which one or a few chromosomes undergo punctuated shattering and reassembly events [53] . Such events have been linked to poor outcome in human melanoma [94] and may be enriched in tumors with p53 dysfunction or those that lack means to extend telomeres [95 , 96] . Notably , we show here that MDM2 and mutually exclusive TP53 alterations are common in canine melanoma . Similarly , inactivating p53 mutations have been found in human mucosal and acral melanoma , suggesting p53 pathway dysregulation may be crucial in non-UV induced melanoma development . Further , UV-induced TERT promoter mutations are common in human cutaneous melanoma , and , although they are rare in sun-shielded subtypes , these subtypes have been shown to bear enrichment for other types of mutation that drive TERT overexpression such as SVs and CNVs [57] . The cutaneous tumors in this cohort do not bear somatic TERT promoter mutations or other known genetic lesions that would enable telomere extension . Thus , telomere crisis and the survival of structurally aberrant genomes may play a significant role in the molecular etiology of canine melanoma . Our comprehensive analysis of canine melanoma reveals that most canine melanomas bear a low coding mutation burden and are also less structurally complex than human melanoma . Two WGS approaches coupled with array-based platforms have enabled deep interrogation of these changes , complementing recent cytogenetic analyses of this tumor type [41] . Significant copy number gains on CFA10 and 30 that have been reported as a defining signature of these lesions are recapitulated in this dataset ( S6 Table ) . Our multi-platform approach was also able to further elucidate complex chromosomal rearrangements present in these regions . Both regions are involved in multiple intra- and inter-chromosomal structural events across this cohort ( S8 Table ) . Additionally , focal amplification of the CFA10 10-12MB region encompasses MDM2 , a gene which is known to drive human cancers and is currently being explored as a drug target in TP53 wild type tumors [97] . CNVs associated with canine melanoma also include gain of CFA13 and loss of CFA22 . While not statistically significant via GISTIC in this cohort , both events are present in individual samples . Overall , extensive copy number and structural variation suggest high levels of large-scale chromosome instability , i . e . gain and loss of whole chromosomes or chromosome arms . Intriguingly , mutually exclusive focal amplification of MDM2 or inactivating mutation in TP53 have been shown to be enriched in BRAF- , NRAS- , and NF1-wild-type human melanoma , although human TP53-mutant melanomas tend to also display higher mutation burden and presence of C>T transitions [98] . Taken together the high degree of structural complexity , the lack of TERT mutations ( barring one putative translocation ) or telomere-lengthening mechanisms , and the frequency of MDM2/TP53 mutations all suggest that chromosome instability plays a key role in canine melanomagenesis . In the discovery cohort , putatively pathogenic somatic mutations in orthologs of human cancer genes were present in a single tumor each including ATF6 , EPAS1 , FAT2 , FAT4 , FOXA3 , FOXO1 , GAB2 , HRAS , KIT , KRAS , MMP21 , NRAS , PBX1 , and XPO1 ( S3 Table ) . Of the 14 melanoma hallmark genes evaluated in the prevalence cohort ( including PTPRJ ) , an additional 24 putatively pathogenic somatic mutations were identified in seven genes–NRAS , TP53 , PTPRJ , KIT , KRAS , GNA11 , and BAP1 ( S4 Table ) . Overall , across discovery and prevalence analyses , RAS gene family members were the genes most commonly bearing somatic SNVs , occurring in 24% of cases ( Figs 1C and 2A ) , followed by TP53 and PTPRJ mutations each in 19% of cases , KIT in 8% and PTEN in 5% . Combined , these mutations most commonly impact proliferative and cell cycle/apoptosis pathways in patterns that display both key similarities and differences with human melanoma subtypes [26] ( Fig 3D ) . For example , despite an absence of BRAF and lower abundance of RAS and KIT mutations in canine versus human mucosal melanoma , these tumors display likely MAPK activating events in 35% of cases . Further , canine mucosal melanoma displays a higher burden of cell cycle and apoptotic events ( 51% ) than all subtypes from the Hayward comparator human melanoma cohort assessed here due largely to enrichment for mutually exclusive MDM2 and TP53 mutations in canine mucosal melanoma . However , these mutations are common in human cutaneous melanoma ( ranging from 36% of cases in the Hayward comparator cohort to 62% in the TCGA cutaneous melanoma study [98] ) . Some of these differences are likely due to the still small sizes of non-cutaneous human and canine melanoma cohorts and the need for greater resolution , particularly across different anatomic sites . For example , all but two of the canine mucosal cases described in this study originate from the oral cavity , whereas the eight human mucosal comparators are largely vulvar ( 3 ) or from nasal cavity ( 2 ) . Thus , an ongoing need exists for broader profiling of these melanoma subtypes by anatomic site . Overall , however , these findings nonetheless suggest that both MAPK pathway inhibition ( e . g . via MEK inhibitors ) and p53 pathway inhibition ( e . g . via MDM2 inhibitors ) are important therapeutic axes for development in canine melanoma just as they are in human [38] . The oncogenic MAPK pathway is critically important in many cancers given its central role in conveying extracellular signals to the nucleus in order to regulate cancer hallmarks including proliferation , invasion , metastasis , and angiogenesis . The majority of human cutaneous melanomas are driven in part by constitutive activation of the MAPK pathway through mutation of genes such as BRAF , NRAS , NF1 , KIT , GNAQ , and GNA11 , often in a mutually exclusive pattern [99] . The high frequency of these mutations has motivated the TCGA classification of these tumors according to MAPK mutation status: BRAF ( ~50% of cases ) , RAS ( ~30% ) , NF1 ( ~15% ) , and TWT ( ~10% ) [98] . These genomic categories are correlated with clinical , pathological , molecular , and biological features of melanoma and thus may comprise distinct subtypes . However , less common histological subtypes of melanoma such as mucosal , acral , and uveal melanoma bear unique mutation spectra that are not uniformly centered on canonical activating mutations in the MAPK pathway . Correspondingly , it has been shown that BRAF mutations are exceedingly rare in predominantly oral canine malignant melanoma and , to date , few alterations in other MAPK members have been discovered . These findings were recapitulated in our cohort , which showed no canonical BRAF or NF1 mutations . Nonetheless , MAPK and/or PI3K signaling have been shown to be activated in nearly all cases [100] . Additional mutations impacting the MAPK and PI3K pathways include three KIT mutations , two PTEN mutations , and one GNA11 mutation . In total , 35% of mucosal and 43% of all canine melanomas bear an alteration impacting the MAPK pathway ( Figs 1C and 3D ) . Prior to our studies described here , the mutations underlying such activation have remained largely unknown . Here we show a complete absence of somatic BRAF mutations ( SNVs , CNVs , or translocations encompassing the BRAF locus ) in canine malignant melanoma in keeping with prior studies [32 , 37 , 41 , 101] . We also did not uncover truncating SNVs in or homozygous deletions of NF1 . A higher proportion of our cohort bear RAS mutations than the 6–13% previously described [32 , 100] , although prior studies have focused almost exclusively on NRAS exons one and two . All three major RAS family members are highly conserved ( 100% protein identity ) between canine and human . In humans , of these family members , malignant melanomas predominantly bear NRAS mutations with only very rare KRAS and HRAS mutations . In our cohort , we found four NRAS codon 61 alterations ( 11% ) , four KRAS G12C mutations and one HRAS Q61R mutation . Further , four of these RAS alterations ( two NRAS , one KRAS , and one HRAS mutation ) occur in mucosal tumors , a frequency of 13% in this subtype . However , in our cohort all three acral tumors and both cutaneous tumors had detectable RAS alterations ( three KRAS and two NRAS mutations ) . This unusual pattern of RAS mutation in canine melanoma may reflect important differences in biological , tissue , and species specificities of RAS family members . These data point to the genomic lesions underlying MAPK and PI3K activation in a substantial proportion of canine melanomas , and to subtle genetic differences in disease subtypes within and across species . Most striking is the discovery of a putative novel tumor suppressor gene , PTPRJ , a receptor-type protein tyrosine phosphatase , which has been genetically and functionally implicated in cancer [61 , 62] , but for which clear genetic mechanisms of inactivation have yet to establish its definitive role as a canonical tumor suppressor gene . PTPRJ consists of an extracellular domain with eight fibronectin III motifs , a transmembrane domain , and an intracellular catalytic domain . It was originally cloned from HeLa cells and characterized by overexpression and hyper-activation in dense cultures of fibroblasts , by regulation of contact inhibition , and by its role in regulation of cancer cell proliferation and invasion [60 , 102–107] . Early genetic studies of quantitative trait loci for mouse cancer susceptibility with homologous regions in human cancers pointed to recurrent PTPRJ deletions , LOH , and missense mutations in small cohorts of colorectal ( 49% ) , lung ( 50% ) , and breast ( 78% ) carcinomas in addition to a correlation between PTPRJ LOH and colorectal cancer progression [61 , 62] . Additional sequencing studies in larger cohorts have identified nonsynonymous SNPs in the extracellular fibronectin repeats associated with risk of developing thyroid , colorectal , head and neck squamous cell , and esophageal cancers [67 , 70 , 108–110] . More recently , a subclonal K1017N missense mutation in the non-catalytic cytoplasmic domain of PTPRJ was identified in a primary breast tumor with significant enrichment in a brain metastases and patient-derived xenograft [111] . PTPRJ substrates that may mediate its tumor suppressive potential include ERK1/2 , Akt , various receptor tyrosine kinases , and Src kinases [42 , 112–116] . However , Ptprj knockout mice have normal development with no cancer predisposition and thus inactivation of this gene does not appear to be sufficient to induce tumorigenesis[65] . Across all TCGA studies published to date ( 10 , 951 cases from 33 tumor types in the TCGA PanCancer Atlas accessed via cBioPortal ) , the frequency of somatic PTPRJ point mutations and/or deep deletions is low– 211/10 , 951 ( 1 . 9% , S12 Table ) [117 , 118] . Only 21 somatic PTRPJ mutations are present in the TCGA human cutaneous melanoma data set consisting of 363 cases ( a single homozygous deletion , five truncating mutations , and fifteen missense mutations ) . However , a related receptor-type protein tyrosine phosphatase , PTPRD , is thought play a role in regulation of STAT3 signaling and has been frequently implicated as a tumor suppressor in human cancers through inactivating somatic mutation , focal deletion or methylation in glioma , melanoma , neuroblastoma , colorectal , liver , head and neck , and lung cancers [119–122] . In human cutaneous melanoma , PTPRD is deleted or truncated in 9–12% of cutaneous cases , but has not been determined to occur at high frequency in rare histological subtypes [50 , 55 , 56 , 120 , 123] . Here , we present the first report of recurrent somatic truncating mutations in PTPRJ in a naturally occurring cancer . We have discovered seven cases ( 19% ) of canine melanomas bearing somatic PTPRJ mutations . Canine and human PTPRJ orthologs share 70% sequence identity with a highly conserved C-terminus containing the protein tyrosine phosphatase catalytic domain ( S6 Fig ) . Sequencing of PTPRJ across all 37 tumors revealed nine mutations in seven cases ( seven mucosal and one uveal ) comprising 19% of all tumors and 23% of mucosal cases . Six frameshifts or stop gains were discovered in addition to one splice site mutation , a C-terminal 10-amino acid deletion , and a single predicted damaging missense mutation . Two cases–ND10-190 and ND10-376 –contained two mutations each , consistent with bi-allelic inactivation of a tumor suppressor gene . Further , LOH was evident by analysis of adjacent SNPs in WGS data in case ND10-166 bearing the M110fs mutation ( S9 Table ) . Although regional LOH on chromosome 18 was observed by SNP array in three of six cases bearing single mutations in PTPRJ , these regions were not observed to directly overlap the coding region of PTPRJ . Overall , the enrichment for PTPRJ truncating mutation in canine malignant melanoma bears intriguing implications both for a previously underappreciated role for this gene in human melanoma ( e . g . through as-yet understudied roles for hemizygous deletion [124] and/or epigenetic modifications ) and for the possibility of unique mechanisms of tumorigenesis across species . Through deep integrated genomic analysis combining WGS , LI-WGS , RNA sequencing , aCGH , SNP arrays , and targeted Sanger sequencing we have determined that canine melanoma is driven by frequent dysregulation of MAPK and cell cycle/apoptosis pathways and , in some cases as is seen in our WGS cohort of predominantly Cocker Spaniels , extensive chromosomal instability . In keeping with prior comparative melanoma studies that have incorporated histology , targeted sequencing , and aCGH [32 , 36 , 38 , 41] , this work highlights the striking resemblance of canine malignant melanoma to BRAFwt subtypes of human melanoma . Finally , we have additionally discovered a putative novel tumor suppressor that may reflect unique species-specific biology and/or may highlight a tumor suppressive axis more subtly altered and as-yet underappreciated in human melanoma . This work bears immediate relevance for development of improved diagnostic and treatment approaches in canine malignant melanoma and provides further evidence to credential the naturally occurring canine melanoma model for study of relevant genomic subsets of human melanoma .
Samples were obtained under institutional review protocols at the Van Andel Research Institute in collaboration with local speciality veterinary clinics ( protocol #08-06-14 ) . Tumors and whole blood were obtained from 36 dogs recruited from 21 veterinary specialty centers in 10 states ( AZ , CA , FL , IL , MA , MI , OH , TX , VA , WI ) under VARI IACUC and ethical review ( protocol #08-06-14 ) . Material was collected at surgery and snap frozen in optimal cutting temperature ( OCT ) compound . Patient matched control DNA was obtained from peripheral blood mononuclear cells . Each resected tumor was evaluated by an on-site board-certified veterinary pathologist and then by BD to estimate tumor content and extent of tissue heterogeneity . Diagnosis of malignant melanoma was histologically confirmed according to criteria defined by the American College of Veterinary Pathologists in addition to criteria recently established by comparative analyses of canine and human melanoma focusing on architecture , pigmentation , and the presence of differentiation markers [32 , 100 , 125] . Two tissue microarrays ( TMAs ) , designated Dog MEL A TMA and Dog MEL B TMA , consisted of 96 individual dogs and 131 tissue samples placed in duplicate and two tissue samples placed in quadruplicate ( 272 array spots ) . Multiple tumors from nine dogs were present on the array and multiple samples from varying sites within the same tumor were present for twelve dogs . Additionally , non-melanoma stromal or control normal tissues were included . TMAs were hematoxylin and eosin-stained and evaluated via routine immunohistochemical procedures for melanoma cocktail ( anti-melan A , anti-melanosome , and anti-tyrosinase ) , and antibodies to vimentin , MDM2 and p53 . Samples scoring positive for MDM2 staining were then confirmed for positive staining with melanoma cocktail and re-evaluated for p53 staining . Positive staining was counted if at least one of the two duplicate samples could be evaluated for both MDM2 and melanoma cocktail on the TMA . Antibodies were purchased from Santa Cruz Biotechnology or Cell Marque . A total of 98 dogs and 189 spots/samples ( 132 tumors ) met these criteria for evaluation for MDM2 protein expression by IHC . Of these 98 dogs , 18 dogs ( 17% ) had melanocytic tumors positive for MDM2 staining in 33 spots/samples ( 25% ) . MDM2 staining was predominantly cytoplasmic highest intensity at junction between epithelial and subepithelial ( submucosa , dermis ) . Staining was observed in both malignant pigmented and amelanotic melanoma and benign melanocytomas . Most intense staining ( 4+ cytoplasmic and nuclear ) was observed in a benign cutaneous melanocytoma from a boxer that had additionally a malignant melanoma ( negative for MDM2 staining on the array ) and multiple cutaneous mast cell tumors . Tissue was disrupted and homogenized in Buffer RLT plus ( Qiagen AllPrep DNA/RNA Mini Kit ) , using the Bullet Blender , Next Advance , and transferred to a microcentrifuge tube containing Buffer RLT plus and 1 . 6 mm stainless steel beads or 0 . 9 mm–2 . 0 mm RNase free stainless steel beads . Blood leukocytes ( buffy coat ) were isolated from whole blood by centrifugation at room temperature and resuspended in Buffer RLT plus . All samples were homogenized , centrifuged at full speed , and lysates were transferred to Qiagen AllPrep spin columns . Genomic DNA and RNA were then purified following the manufacturer's protocol . DNA was quantified using the Nanodrop spectrophotometer and quality was accessed from 260/280 and 260/230 absorbance ratios . RNA was analyzed on the Agilent Bioanalyzer RNA 6000 Nano Chip to validate RNA integrity ( RIN≥7 . 0 ) . Three μg of genomic DNA from each sample was fragmented to a target size of 300–350 base pairs ( bp ) . Overhangs in the fragmented samples were repaired and adenine bases were ligated on . Diluted paired end Illumina adapters were then ligated onto the A-tailed products . Following ligation , samples were run on a 3% TAE gel to separate products . Ligation products at 300 bp and 350 bp were selected for each sample , isolated from gel punches , and purified . 2× Phusion High-Fidelity PCR Master Mix ( Finnzymes; catalog#F-531L ) was used to perform PCR to enrich for these products . Enriched PCR products were run on a 2% TAE gel and extracted . Products were quantified using Agilent's High Sensitivity DNA chip ( catalog#5067–4626 ) on the Agilent 2100 Bioanalyzer ( catalog#G2939AA ) . Long insert whole genome libraries were constructed as previously described [126] with the following modifications: 1100ng inputs were used; following DNA fragmentation , a bead purification was performed at a 1:1 . 8 ( sample volume: bead volume ) ratio; a 1% size selection gel was used; and during library enrichment , 10 PCR cycles was used . Libraries were clustered onto Illumina V3 flowcells ( San Diego , CA ) using Illumina’s TruSeq PE Cluster Kit V3 ( cat#PE-401-3001 ) and sequenced for paired 100bp reads using Illumina’s TruSeq SBS Kit V3 ( cat#FC-401-3002 , n = 3 ) on the Illumina HiSeq . 10 ng of total RNA was used to generate whole transcriptome libraries for RNA sequencing . Using the Nugen Ovation RNA-Seq System ( cat#7100–08 ) , total RNA was used to generate double stranded cDNA , which was amplified using Nugen's SPIA linear amplification process . Amplified cDNA was input into Illumina's TruSeq DNA Sample Preparation Kit–Set A ( cat#FC-121-1001 ) for library preparation . In summary , 1 μg of amplified cDNA was fragmented to a target insert size of 300 bp and end repaired . Samples were then adenylated and indexed paired end adapters were ligated . Ligation products were run on a 2% TAE gel and size selected at 400 bp . Ligation products were isolated from gel punches and purified . Cleaned ligation products were input into PCR to enrich for libraries . PCR products were cleaned and quantified using the Agilent Bioanalyzer . Tumor and normal libraries were prepared for paired end sequencing as described above . Clusters were generated using Illumina's cBot and HiSeq Paired End Cluster Generation Kits ( catalog#PE-401-1001 ) and sequenced on Illumina's HiSeq 2000 using Illumina's HiSeq Sequencing Kit ( catalog#FC-401-1001 ) . All informatic tools , versions , and flags are shown in S13 Table . BCL to FASTQ file conversion was performed using Illumina's BCL converter tool . Read alignment was performed with BWA ( Burrows-Wheeler Aligner ) v . 0 . 7 . 8 [127] using the canine reference genome CanFam 3 . 1 . Aligned BAMs were indel ( insertion/deletion ) realigned and recalibrated using GATK v3 . 3 . 0 [128 , 129] and duplicate reads marked using Picard v1 . 128 ( http://broadinstitute . github . io/picard/ ) . Variants were called using Strelka v . 1 . 0 . 13 [130] , Seurat v2 . 6 [131] and MuTect v . 1 . 1 . 4 [132] and calls were annotated according to dbSNP 151 , SNPs on the Illumina CanineHD BeadChip , and SnpEff-3 . 5 [133] . Final somatic SNVs were called by at least 2/3 callers . LI-WGS data were utilized for CNV and SV detection . For CNV detection , read depths at every 100 bases across sequenced regions were determined . Next , normalized log2 fold-changes between tumor and normal were calculated and a smoothing window applied . Tumor allele frequencies of known heterozygous germline SNPs were utilized to evaluate potential false positives and correct biases . Finally , the Circular Binary Segmentation algorithm [134] was used to correct log2 fold-changes . For mutation burden metrics , a focal CNV is included if the log2 change is > = |2| . SV detection was performed utilizing Delly v0 . 76 [52] . A minimum tumor allele ratio of 0 . 10 and a minimum quality score of 20 is required for an SV to be called . RNA sequencing data in FASTQ format was checked for quality using cycle-by-cycle quality plots and biases such as GC content . Reads were aligned to the canine reference genome CanFam 3 . 1 using STAR-2 . 4 to generate alignment files in BAM format [49] . Somatic SNVs were called with HaplotypeCaller ( GATK v3 . 3 . 0 ) and verified in IGV . Transcript abundance in FPKMs ( Fragments Per Kilobase of transcript per Million mapped reads ) was obtained by CuffDiff v2 . 2 . 1 [134] and annotated using ENSEMBL ( CanFam 3 . 1 . 68 ) . Next generation sequencing data from this study have been submitted to the NCBI Biosample Database ( http://www . ncbi . nlm . nih . gov/bioproject/389294 ) under project number PRJNA389294 with sample accession IDs SAMN07376261 , SAMN07376262 , SAMN07376263 , SAMN07376264 , SAMN07376265 , SAMN07376266 , SAMN07376267 , SAMN07376268 , SAMN07376269 , SAMN07376270 , SAMN07376271 , SAMN07376272 , and SAMN07376273 . A list of 1 , 405 genes with single nucleotide variation or structural variation or copy number variation from the discovery cohort were analyzed using ClueGo4 [79] , a Cytoscape plug-in , to create a functionally organized pathway network . Kappa scores were then used to measure association between the networks . Functional networks were created with a minimum Kappa score threshold of 0 . 5 and a minimum of 3 affected genes in every network forming at least 10% of the total associated genes in that particular network . The genes were assigned to the networks based on the predefined pathways from KEGG , REACTOME and Wiki Pathways . 97 pathways were obtained , all with Benjamini-Hochberg corrected p-value <0 . 05 . These pathways were grouped together based on inter-term kappa score and named by the most significant pathway in the respective groups . PCR amplification of 15 genes ( NRAS , KRAS , BRAF , GNA11 , GNAQ , PTPRJ , TP53 , MDM2 , BAP1 , CDK4 , PTEN , c-KIT , MITF and NF1 ) was performed using primers targeting all coding exons ( S4 Table ) . All amplification reactions were performed using Platinum Taq DNA Polymerase #10966–034 ( Life Technologies; Carlsbad , CA ) . Briefly , each primer pair was mixed with 10 ng of genomic DNA and subjected to the following cycling parameters: 94°C for 2 min . , 3 cycles at each temperature: 30 sec . at 94°C , 30 sec . at 60–57°C , 45 sec . at 72°C; 25 cycles: 30 sec . at 94°C , 30 sec . at 62°C , 45 sec . at 72°C; final extension of 5 min . at 72°C . PCR amplicons were sequenced using M13 forward and reverse primers at the Arizona State University’s DNA Lab ( Tempe , AZ ) . Oligo array CGH ( aCGH ) was performed by co-hybridization of tumor ( test ) DNA and a common reference DNA sample , where the latter comprised an equimolar pool of genomic DNA samples from multiple healthy individuals of various breeds . DNA was labeled using an Agilent SureTag Labeling Kit ( Agilent Technologies , Santa Clara , CA ) with all test samples labeled with Cyanine-3-dCTP and the common reference sample labeled with Cyanine-5-dCTP . Fluorochrome incorporation and final probe concentrations were determined using routine spectrophotometric parameters with readings taken from a Nanodrop1000 . Fluorescently labeled test and reference samples were co-hybridized to Canine G3 180 , 000 feature CGH arrays ( Agilent , AMADID 025522 ) for 40 h at 65°C and 20 rpm , as described previously [135 , 136] . Arrays were scanned at 3 μm using a high-resolution microarray scanner ( Agilent , Model G2505C ) and data extracted using Feature Extraction ( v10 . 9 ) software . Scan data were assessed for quality by the ‘Quality Metrics’ report in Agilent’s Feature extraction software ( v10 . 5 ) ( Agilent Technologies ) . SNP genotyping was performed using the Illumina CanineHD array ( cat#WG-440-1001 ) . Per manufacturer’s protocol , 200ng of DNA was first denatured then neutralized with 0 . 1N NaOH before amplification at 37°C for 24 hours . The amplified DNA was then enzymatically fragmented and precipitated using 100% 2-propanol before drying for one hour at room temperature . After resuspension the fragmented DNA was then denatured and loaded onto the CanineHD BeadChip and hybridized for 16 hours at 48°C . BeadChips were washed , a single base extension of hybridized primers added followed by multi-layer staining of the primers . Arrays were then washed , coated with the XC4 reagent ( Illumina ) and dried under vacuum for one hour . Coated arrays were read on the HiScan system and data visualized using the Illumina GenomeStudio Genotyping 2 . 0 software with an average sample call rate of 97% . For both aCGH and SNP arrays , copy number data were analyzed with NEXUS Copy Number v8 . 0 software ( Biodiscovery Inc . , CA , USA ) . For cross-platform comparisons , LI-WGS BAMs were also analyzed utilizing Nexus software . CNVs were identified using a FASST2 segmentation algorithm with a significance threshold of 5 . 5×10−6 . Aberrations were defined as a minimum of three consecutive probes with log2 tumor: reference value of >1 . 14 ( high gain ) , 1 . 13 to 0 . 2 ( gain ) , −0 . 23 to −1 . 1 ( loss ) , <−1 . 1 ( big loss ) . Recurrent CNVs within each subtype were determined within NEXUS using an involvement threshold of 50% . Significance of these regions was then determined in NEXUS using the GISTIC algorithm ( to identify regions with a statistically high frequency of CNVs over background ) with a G-score cut off of G>1 . 0 and a significance of Q<0 . 05 . CNV frequency comparisons amongst sample groups were performed in NEXUS using Fisher’s exact test with differential threshold of >50% and significance p<0 . 05 . Significance of each probe between the two groups was calculated in NEXUS using a Mann–Whitney test for median comparison .
|
Melanoma , a cancer arising from transformed melanocytes , commonly occurs in pet dogs . Unlike human melanoma , most often occurring in sun-exposed cutaneous skin , canine melanoma typically arises in sun-shielded oral mucosa . Canine melanoma clinically resembles human melanoma , particularly sun-shielded subtypes . However , canine melanoma genomics remain poorly defined . Similarly , although new treatments for human melanoma based on genomic understanding have shown improvements in outcomes for these patients , treatments for canine melanoma are limited and outcomes remain poor . Detailing the genomic basis of canine melanoma thus provides untapped potential for improving lives of pet dogs and helping to define canine melanoma’s role in comparative studies that also inform human melanoma understanding . To better define the genomic landscape of canine melanoma , we performed multi-platform characterization of 37 tumors . Our integrated analysis confirms that these tumors commonly contain mutations in canine orthologs of human cancer genes such as RAS , MDM2 , and TP53 alongside mutational patterns sharing important similarities with human melanoma subtypes . We have also found a new putative cancer gene , PTPRJ , frequently mutated in canine melanoma . These data will guide biologic and therapeutic studies in canine melanoma while broadly framing the utility of comparative studies of canine and human cancers .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"cancer",
"genomics",
"dermatology",
"medicine",
"and",
"health",
"sciences",
"cancers",
"and",
"neoplasms",
"basic",
"cancer",
"research",
"oncology",
"skin",
"neoplasms",
"mutation",
"cutaneous",
"melanoma",
"mammalian",
"genomics",
"molecular",
"genetics",
"mapk",
"signaling",
"cascades",
"malignant",
"skin",
"neoplasms",
"melanomas",
"molecular",
"biology",
"animal",
"genomics",
"somatic",
"mutation",
"signal",
"transduction",
"point",
"mutation",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"cell",
"signaling",
"genomic",
"medicine",
"signaling",
"cascades"
] |
2018
|
Somatic inactivating PTPRJ mutations and dysregulated pathways identified in canine malignant melanoma by integrated comparative genomic analysis
|
Type I interferons ( including IFNαβ ) are innate cytokines that may contribute to pathogenesis during Mycobacterium tuberculosis ( Mtb ) infection . To induce IFNβ , Mtb must gain access to the host cytosol and trigger stimulator of interferon genes ( STING ) signaling . A recently proposed model suggests that Mtb triggers STING signaling through bacterial DNA binding cyclic GMP-AMP synthase ( cGAS ) in the cytosol . The aim of this study was to test the generalizability of this model using phylogenetically distinct strains of the Mtb complex ( MTBC ) . We infected bone marrow derived macrophages with strains from MTBC Lineages 2 , 4 and 6 . We found that the Lineage 6 strain induced less IFNβ , and that the Lineage 2 strain induced more IFNβ , than the Lineage 4 strain . The strains did not differ in their access to the host cytosol and IFNβ induction by each strain required both STING and cGAS . We also found that the three strains shed similar amounts of bacterial DNA . Interestingly , we found that the Lineage 6 strain was associated with less mitochondrial stress and less mitochondrial DNA ( mtDNA ) in the cytosol compared with the Lineage 4 strain . Treating macrophages with a mitochondria-specific antioxidant reduced cytosolic mtDNA and inhibited IFNβ induction by the Lineage 2 and 4 strains . We also found that the Lineage 2 strain did not induce more mitochondrial stress than the Lineage 4 strain , suggesting that additional pathways contribute to higher IFNβ induction . These results indicate that the mechanism for IFNβ by Mtb is more complex than the established model suggests . We show that mitochondrial dynamics and mtDNA contribute to IFNβ induction by Mtb . Moreover , we show that the contribution of mtDNA to the IFNβ response varies by MTBC strain and that additional mechanisms exist for Mtb to induce IFNβ .
Type I interferons ( including IFNαβ ) are innate cytokines that are protective during most viral infections , but may be pathogenic during infections with bacteria such as Mycobacterium tuberculosis ( Mtb ) [1] . Studies have shown that active tuberculosis ( TB ) is associated with expression of interferon-inducible genes [2 , 3] , lepromatous Mycobacterium leprae lesions are enriched in IFNαβ-inducible mRNAs [4] , and that interleukin-1 confers resistance to Mtb by limiting IFNαβ induction [5] . There is also evidence that IFNαβ is protective in certain contexts [6] , and thus it is likely that a balance of this cytokine is required for optimal protection . Given the complex role of IFNαβ signaling during Mtb infection , discovering a model for how Mtb induces IFNβ gene transcription—the first step required for IFNαβ signaling—has been an active and challenging area of research . Several groups have recently proposed a mechanism for IFNβ induction by Mtb . In this model , the first step in the pathway occurs when Mtb gains access to the host cytosol [7] , such as through phagosome permeabilization [8] . The second step is initiation of the STING ( stimulator of interferon genes ) signaling pathway . STING can be triggered by bacterial cyclic dinucleotides [9] or through DNA binding to cGAS ( cyclic GMP-AMP synthase ) in the cytosol [10–12] . The established model suggests that bacterial DNA shed from cytosolic mycobacteria binds to and activates cGAS in order to induce IFNβ [12 , 13] . However , this model does not take into account the contribution of DNA from other sources , such as mitochondria . Furthermore , this model has only been tested with strains from MTBC Lineage 4 and it is unknown whether the mechanism is generalizable to strains from other phylogenetic lineages . The two important outstanding questions that we address in this study are: 1 ) Do phylogenetically distinct mycobacterial strains induce distinct levels of IFNβ in vitro ? and 2 ) What are the mechanisms underlying IFNβ induction by these strains ? Identifying a mechanism by which distinct MTBC strains promote IFNαβ induction would provide crucial insight into a mechanism of Mtb pathogenesis and into the evolution and diversity of mycobacterial strains . Therefore , in order to examine the model for IFNβ induction by Mtb , we infected bone marrow derived macrophages ( BMDM ) with bacterial strains from three phylogenetically distinct MTBC lineages and assayed cytosolic signaling and IFNβ induction by each . The MTBC strains we chose were 1182 ( Lineage 6; also known as M . africanum ) [14 , 15] , H37Rv ( Lineage 4 ) [16] , and 4334 ( Lineage 2 ) [17 , 18] ( S1 Table ) . We find that 1182/Lineage 6 induces significantly less , and 4334/Lineage 2 induces significantly more , IFNβ than H37Rv/Lineage 4 . Additionally , we find that strain differences in IFNβ induction are not due to differences in cytosolic access or bacterial DNA shedding . Instead , we provide evidence that IFNβ differences are due , at least partially , to differences in mitochondrial stress and mitochondrial DNA ( mtDNA ) in the cytosol . Moreover , we show that additional mechanisms exist for Mtb to induce IFNβ and thus that the mechanism for IFNβ induction by Mtb is much more complex than the established model has implied .
To determine whether distinct mycobacterial strains induce distinct levels of IFNβ in vitro , we infected bone marrow derived macrophages ( BMDM ) with strains from three distinct phylogenetic lineages ( S1 Table ) . We found that 1182/Lineage 6 induced less IFNβ , and 4334/Lineage 2 induced more IFNβ , than H37Rv/Lineage 4 ( Fig 1 ) . Differences in mRNA expression were detected at 3 hr post infection and continued at 6 and 24 hr post infection ( Fig 1A ) . Differences in secreted protein were detected at 48 hr post infection and were consistent across several multiplicities of infection ( MOI ) ( Fig 1B ) . The H37Rv ΔEsx-1 strain that is unable to access the cytosol [8 , 19] did not induce IFNβ ( Fig 1A and 1B ) . Importantly , the differences in IFNβ protein secretion were not due to differences in bacterial numbers or intracellular survival ( Fig 1C ) , and were not mirrored by differences in tumor necrosis factor ( TNF ) secretion ( Fig 1D ) . To verify this , we included colony forming units ( CFU ) and TNF as covariates in an ANCOVA model ( S2 Table ) . We found that CFU and TNF did not explain variation in IFNβ , while bacterial strain did , at each MOI . This suggested that IFNβ induction by Mtb was bacterial strain-dependent and that strain differences were not explained by differences in bacterial growth or accompanied by global differences in inflammatory cytokine production . We also assayed interleukin-1 ( IL-1 ) secretion , which negatively regulates IFNβ [5] . We found that IL-1 levels were below the limit of detection during infection with each strain , and thus variation in IL-1 likely did not explain IFNβ variation in this system . Therefore we hypothesized that differences in IFNβ induction were due to differences in cytosolic signaling . To induce IFNβ gene transcription , Mtb must gain access to the host cytosol [7] . To determine whether cytosolic access varied by mycobacterial strain , we infected BMDM and assayed colocalization of each strain with FK2 and Galectin-3 . FK2 labels ubiquitinated proteins and thus labels bacteria that gain access to the cytosol and are ubiquitinated [20] . Galectin-3 labels damaged vacuole membranes [21] . There were no detectable differences in the percent of bacteria that colocalized with FK2 ( Fig 2A and 2C ) or Galectin-3 ( Fig 2B and 2D ) at any time point examined during infection with 1182/Lineage 6 , H37Rv/Lineage 4 , or 4334/Lineage 2 . 4334/Lineage 2 interacted differently with FK2 and Galectin-3 than 1182/Lineage 6 and H37Rv/Lineage 4 , as indicated by staining patterns ( Fig 2A and 2B ) , however this was not associated with IFNβ induction . In addition , we measured the mean fluorescence intensity ( MFI ) of FK2 and Galectin-3 surrounding each colocalized bacterium and found no significant differences between the strains ( Fig 2E and 2F ) . The ΔEsx-1 negative control strain did not colocalize with FK2 or Galectin-3 . These data suggested that the differences in IFNβ induction between MTBC strains were not attributable to differences in access of the bacteria to the cytosol . We confirmed these results using a LiveBLAzer FRET assay [8] . In this assay BMDM are incubated with Cephalosporin Coumarin Fluorescein 4 ( CCF4 ) , a cephalosporin substrate labeled with two fluorophores that form a fluorescent resonance energy transfer ( FRET ) pair . In BMDM where Mtb is sequestered in a phagosome , CCF4 is uncleaved and when excited at 409 nm emits fluorescence with a peak at 520 nm . In BMDM where Mtb gains access to the cytosol , Mtb’s endogenous β-lactamase cleaves CCF4 and disrupts the FRET signal and phagosome permeabilization is detected by BMDM that emit a signal at 450 nm . We found no difference in 450:520 nm fluorescence ratios between the three MTBC strains ( S1 Fig ) , which confirmed the FK2 and Galectin-3 results . Together , these data indicated that differences in IFNβ induction between the MTBC strains were not due to differences in access to the host cytosol . Once Mtb gains access to the cytosol it triggers STING signaling , either by bacterial cyclic dinucleotides [9] or through dsDNA binding to and activating cGAS in the cytosol [10–12] . To determine whether IFNβ induction by each strain was dependent on STING signaling we infected STING-/- and wild type BMDM and assayed IFNβ protein secretion . We found that IFNβ induction by each strain was completely abrogated in the absence of STING ( Fig 3A ) , indicating that IFNβ induction was STING-dependent , and suggested that under these conditions differential production of cGAMP or release of DNA into the cytosol may contribute to differential IFNβ induction . Next we determined whether IFNβ induction by each strain was due to direct STING signaling or due to STING signaling via cGAS . Therefore we infected cGAS-/- and wild type BMDM and assayed IFNβ protein secretion . We found that IFNβ induction by each strain was reduced in the absence of cGAS to the same degree that IFNβ induction was reduced in the absence of STING ( Fig 3B ) . These data suggested that differences in IFNβ induction between the MTBC strains might depend on differences in the availability of cytosolic DNA for binding to and activating cGAS . cGAS is a cytosolic DNA sensor that can recognize dsDNA of microbial , nuclear or mitochondrial origin [22 , 23] . To determine whether accumulation of dsDNA in the cytosol was mycobacterial strain-dependent we infected BMDM with each MTBC strain and quantified cytosolic DNA following cell fractionation . We found that infection with H37Rv/Lineage 4 was associated with increased mitochondrial and nuclear DNA in the cytosol compared with 1182/Lineage 6 ( Fig 4A and 4B ) , and we found no differences in release of bacterial DNA ( Fig 4C and 4D ) . The procedure for preparing subcellular fractions to quantitate DNA permeabilized phagosome membranes as well as the plasma membrane , as is indicated by bacterial DNA detected in the cytosolic fraction of cells infected with the ΔEsx-1 mutant ( Fig 4C and 4D ) . To verify that the fractionation buffer did not permeabilize other organelle membranes , we used immunoblotting to determine that the mitochondrial proteins Complex Vα ( CVα ) and pyruvate dehydrogenase E1α ( PDH ) remained in the organelle fractions and were not found in the cytosol fractions ( Fig 4E ) . Given these data we hypothesized that reduced IFNβ induction by 1182/Lineage 6 was due to reduced mtDNA in the cytosol . Although we also found reduced nuclear DNA in the cytosol of 1182/Lineage 6-infected cells , a recent study has shown that mtDNA–and not nuclear DNA–stimulates IFNβ induction [24] . We further hypothesized that the reduced mtDNA in the cytosol was due to reduced mitochondrial stress in cells infected with 1182/Lineage 6 . MtDNA may accumulate in the cytosol under conditions of stress [25–27] , and mitochondrial stress and mtDNA have been implicated in IFNβ induction [24 , 28–31] . To determine whether the release of mtDNA into the cytosol was associated with mitochondrial stress , we infected BMDM with each Mtb strain and quantified ATP as a measure of the bioenergetic state of the cell . We infected BMDM cultured in media lacking glucose and supplemented with galactose to prevent BMDM from using glycolysis for ATP production . We found that 1182/Lineage 6 was associated with higher cellular ATP concentrations compared with H37Rv/Lineage 4 , with ATP levels similar to uninfected cells and the negative control ΔEsx-1 ( Fig 4F ) . This suggested that 1182/Lineage 6 infection might induce especially low levels of mitochondrial stress . These data were consistent with the finding of lesser quantities of mtDNA in the cytosol of BMDM infected with 1182/Lineage 6 ( Fig 4A ) . To further examine mitochondrial stress , we quantified superoxide production in BMDM following infection with each bacterial strain . Superoxide is a reactive oxygen species ( ROS ) that is a byproduct of mitochondrial oxidative phosphorylation as electrons that leak from the electron transport chain are transferred to molecular oxygen . Conditions of stress result in an increase in the number of leaking electrons and result in accumulation of superoxide [32 , 33] . We found reduced mitochondrial superoxide production during 1182/Lineage 6 infection compared to H37Rv/Lineage 4 ( Fig 4G and 4H ) . This supported our conclusion that 1182/Lineage 6 was associated with reduced mitochondrial stress . Importantly , we also found that 4334/Lineage 2 was associated with less host DNA in the cytosol ( Fig 4A and 4B ) and less mitochondrial stress ( Fig 4F and 4G ) than H37Rv/Lineage 4 . This indicated that 4334/Lineage 2 induced high IFNβ levels by a different , additional mechanism . Several studies have shown that mitochondrial ROS-induced oxidation of DNA contributes to the inflammatory response to DNA [25 , 30 , 34 , 35] . Our data suggest that this may also be true for the IFNβ response to mycobacterial infection ( Figs 1 and 4 ) . To determine whether mycobacterial strain-dependent differences in IFNβ were due to differential accumulation of ROS , we treated BMDM with the antioxidant MitoQ and assayed accumulation of cytosolic DNA and IFNβ protein secretion induced by each strain . MitoQ is coenzyme Q attached to a lipophilic triphenylphosphonium cation . Coenzyme Q is a strong reducing agent and thus acts as an antioxidant by transferring electrons to superoxide . The lipophilic cation causes accumulation of the molecule specifically in the mitochondria [36] . We used decyltriphenylphosphonium bromide ( dTPP ) as the negative control . We found that MitoQ treatment partially reduced superoxide accumulation ( Fig 5A and 5B ) . Higher doses of MitoQ were toxic to the cells and therefore we were not able to completely eliminate superoxide accumulation . Correspondingly , we found that MitoQ treatment reduced mtDNA accumulation in the cytosol during H37Rv/Lineage 4 and 4334/Lineage 2 infections ( Fig 5C ) , but did not impact accumulation of bacterial DNA ( Fig 5E ) . MitoQ treatment also reduced nuclear DNA accumulation in the cytosol , but the levels of nuclear DNA in the cytosol were lower than the levels of mtDNA ( Fig 5D ) . MitoQ treatment had no significant effect on total cellular levels of mitochondrial , nuclear , or bacterial DNA ( Fig 5F–5H ) . Furthermore , we found that MitoQ treatment partially reduced IFNβ induction by H37Rv/Lineage 4 and 4334/Lineage 2 ( Fig 6A ) , and that the percent of IFNβ inhibited by MitoQ treatment was positively correlated with the amount of IFNβ a strain induced ( Pearson r = 0 . 88 , p = 0 . 002 and Fig 6C and 6E ) . IFNβ reduction during MitoQ treatment was not mirrored by a reduction in TNF ( Fig 6B and 6D ) or by a reduction in CFU ( Fig 6F ) . In addition , the amount of mtDNA in the cytosol of infected cells positively correlated with IFNβ induction ( Pearson r = 0 . 73 , p = 0 . 0006 and S2 Fig ) , while bacterial DNA did not ( Pearson r = -0 . 23 , p = 0 . 36 and S2 Fig ) . Nuclear DNA in the cytosol also correlated with IFNβ induction , but the correlation was not a strong as for mtDNA ( Pearson r = 0 . 60 , p = 0 . 01 and S2 Fig ) . These data indicated that IFNβ induction by the MTBC strains was due , at least in part , to mitochondrial stress and cytosolic mtDNA accumulation .
In this study we show that the M . tuberculosis complex strain 1182 from Lineage 6 induces less mitochondrial ROS , less mtDNA in the cytosol , and lower IFNβ induction than H37Rv/Lineage 4 . Further , we show that reducing mitochondrial ROS during Mtb infection reduces IFNβ induction . Therefore we propose that mitochondrial stress contributes to IFNβ induction by Mtb ( Fig 7 ) . We also show that 4334/Lineage 2 induces similar to lower levels of mitochondrial ROS and cytosolic mtDNA than H37Rv/Lineage 4 , yet induces higher IFNβ induction . Thus we propose that 4334/Lineage 2 induces additional , unidentified pathways to promote IFNβ induction ( Fig 7 ) . Together these results show that the mechanism for IFNβ induction by Mtb is much more complex than the established model suggests . Mitochondria are integral centers in the cell that recognize and emit danger signals–such as mtDNA and ROS–to direct cellular responses [37] , including various innate immune responses [32] . Here we have shown that mitochondrial dynamics contribute to the IFNβ response to Mtb infection . This has already been shown in other pathological contexts . MtDNA instability and accumulation of mtDNA in the cytosol contribute to the cGAS-mediated IFNβ response to HSV-1 [28] . Additionally , systemic lupus erythematosus is associated with mitochondrial dysregulation and type I interferon signatures [38–40] , oxidized mtDNA released from lupus neutrophils drives type I interferon [24] , and mitochondrial antioxidants attenuate IFNβ responses and lupus-like disease in mice [30] . Our finding that mitochondrial stress contributes to IFNβ induction by distinct MTBC strains suggests that it will be worthwhile to determine how widespread this pathway may be in other infections or diseases . An important remaining question is how 1182/Lineage 6 induces less mitochondrial stress than H37Rv/Lineage 4 . It is possible that 1182/Lineage 6 secretes reduced levels of a virulence factor that targets and disrupts mitochondria . Examples of such virulence factors in other pathogens include Staphylococcus aureus α-toxin and Streptococcus pneumoniae pneumolysin [41] . 1182/Lineage 6 contains a premature stop codon in Rv3879c of the Esx-1 gene locus [42] , which does not impact ESAT-6 secretion [14] but may impact the secretion of other virulence factors . It is also possible that 1182/Lineage 6 secretes higher levels of an antioxidant that prevents mitochondrial toxicity . A third possibility is that 1182/Lineage 6 secretes reduced levels of a metabolite that feeds into the citric acid cycle and drives mitochondrial ROS . Succinate secretion is upregulated in mycobacteria under conditions of hypoxia [43] and perhaps a similar metabolic change occurs upon macrophage entry that drives mitochondrial ROS . This function been proposed for citrate produced by Salmonella during infection [44] . These possibilities are not mutually exclusive and studies are ongoing to address these questions . Our results provide an alternative view of cytosolic signaling and IFNβ induction by Mtb than previous studies . First , we show that IFNβ induction is not necessarily a measure of cytosolic access , as 1182/Lineage 6 induced lower levels of IFNβ than the other two bacterial strains but , as indicated by three distinct assays , gained similar access to the host cytosol . Second , we show that bacterial DNA is probably not the predominant danger signal in the cytosol during Mtb infection; mtDNA released from stressed mitochondria likely binds cGAS to induce IFNβ . We did not observe complete reduction in IFNβ with MitoQ treatment . This could be due to residual superoxide and cytosolic mtDNA present in MitoQ-treated cells or it could be due to an additional mechanism that acts in concert with mitochondrial stress to promote IFNβ signaling . Perhaps bacterial DNA contributes to differential IFNβ induction but cannot be detected by the gene-specific primers that we used in our assays . We propose that we were able to discover this mechanism , while previous studies did not , for several reasons . First , we used diverse clinical isolates to test the model . Second , we used low bacterial doses ( MOI of 1 , 5 , or 10 ) throughout the study and carefully controlled for any effect that CFU might have had on our results ( Figs 1 and 6 and S2 Table ) . The study that showed mycobacterial DNA bound to cGAS used a high dose of bacteria ( MOI of 20 ) and did not show CFU or mtDNA results [12] . Our results are consistent with previous studies that showed that Lineage 2 strains induce more IFNβ than Lineage 4 strains [45 , 46] . This suggests that the ability to induce mitochondrial stress and/or IFNβ may contribute to the global dissemination and virulence of this lineage [47] . However , further studies would be required to determine how widespread this pathway is in Lineage 2 . Our data also suggest that additional mechanisms exist for Lineage 2 strains to induce high levels of IFNβ , as 4334/Lineage 2 induced less mitochondrial stress than H37Rv/Lineage 4 . We propose that 4334/Lineage 2 induces cGAS signaling by an as of yet undetermined mechanism . We noted that 4334/Lineage 2 interacted differently with FK2 and Galectin-3 than H37Rv/Lineage 4 and 1182/Lineage 6 , indicating that this strain may behave differently in the host cytosol and may induce entirely distinct signaling pathways . It also remains unknown if the reduced ability to induce mitochondrial stress and IFNβ is widespread in Lineage 6 strains . If so , these factors likely played a role in the evolution of Lineage 6 . Lineage 6 strains exhibit low virulence and are geographically restricted to West Africa [48]; reduced ability to induce mitochondrial stress and IFNβ may contribute to this attenuation . Interestingly , Lineage 6 is currently not being outcompeted by Lineage 4 and Lineage 2 in areas where it is prevalent [49 , 50] . Thus reduced mitochondrial toxicity and the ability to cause slower disease progression may have been selected for in Lineage 6 , and future studies are warranted to investigate this possibility . Ultimately , understanding IFNβ induction by Mtb may facilitate the development of host-directed TB treatments . Treating mice with prostaglandin E2 and zileuton , which limit IFNβ induction , confers tolerance to Mtb [5] and zileuton has been developed as a drug that can be administered to TB patients to enhance the efficacy of antibiotic treatment [51] . Our data suggest that mitochondria-specific antioxidants could be another means to limit the pathogenic type I interferon response . Studies that identify additional mechanisms by which Mtb induces type I interferon would then be indispensible as they could provide additional targets for TB prevention and therapy .
All animal experiments were done in accordance with procedures approved by the NYU School of Medicine Institutional Animal Care and Use Committee ( IACUC , Laboratory Animal Care Protocol: 150502–01 ) . These IACUC regulations conformed to the national guidelines provided by the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . BMDM were generated from 8–12 week old wild-type C57BL/6 mice from The Jackson Laboratory . BMDM were generated using 20% L929-cell-conditioned media for 7 days , unless otherwise indicated . For antioxidant treatment , BMDM were treated with 0 . 25 μM MitoQ ( gift from Michael Murphy , Ph . D . , MRC Mitochondrial Biology Unit , Cambridge , United Kingdom ) or 0 . 25 μM decyltriphenylphosphonium bromide ( Santa Cruz Biotechnology ) for 4 hours prior to infection and throughout infection . H37Rv/Lineage 4 was obtained from American Type Culture Collection , 1182/Lineage 6 was obtained from an HIV-uninfected male with pulmonary tuberculosis in The Gambia ( courtesy of Bouke de Jong , M . D . , Ph . D . , Institute of Tropical Medicine , Antwerp , Belgium ) , and 4334/Lineage 2 was obtained from a patient of unspecified HIV status with pulmonary tuberculosis in San Francisco ( courtesy of Diane Ordway , Ph . D . , Colorado State University , Fort Collins , USA and Midori Kato-Maeda , M . D . , San Francisco General Hospital ) . Strains were grown at 37°C in Middlebrook 7H9 liquid or 7H11 solid media supplemented with 10% albumin , dextrose and catalase . Bacteria were grown to exponential phase , resuspended in 0 . 5% PBS-Tween 80 , and centrifuged at 150 g for 8 min to remove clumped and dead bacteria . OD580:CFU ratios were calculated individually for each strain . Bacteria were added to BMDM at the MOI reported in the figure legend . Plates were incubated at 37°C under 5% CO2 . Cells were washed 3–4 hr post infection and given fresh BMDM media . BMDM were infected in 6-well plates . Digitonin extracts were generated as previously described [52] , with the digitonin concentration optimized for our assays . 24 hr post infection , cells were collected and resuspended in ice cold PBS . One eighth ( by volume ) of the suspension was subjected to bead-beating to lyse BMDM and bacteria and saved as the undisturbed cell normalization control . The remainder of the cells were resuspended in ice cold lysis buffer containing 150 mM NaCl , 50 mM HEPES pH 7 . 4 , and 0 . 01% digitonin ( Sigma ) . The homogenates were incubated end over end for 10 min at room temperature to allow selective membrane permeabilization and then centrifuged at 650 g for 5 min at 4°C to pellet intact cells . Supernatants were transferred to fresh tubes and centrifuged at maximum speed ( 20 , 800 g ) for 10 min at 4°C to pellet organelles and obtain cytosolic supernatants . Organelle fractions were washed in PBS and resuspended in TN1 lysis buffer ( 50 mM Tris pH 8 , 150 mM NaCl , 10% glycerol , 1 mM EDTA , 0 . 1% Triton-X , protease inhibitor cocktail ) . Organelle and cytosolic fractions were filtered with 0 . 22 μm SpinX filters ( Corning ) . DNA and protein were extracted from undisturbed cell and cytosolic fractions by phenol-chloroform separation . DNA was precipitated from the aqueous phase using isopropanol . Protein was precipitated from the cytosolic fraction phenol phase using 0 . 1M ammonium acetate in methanol [53] . BMDM were infected in 6-well plates . RNA was extracted at indicated time points using RNeasy Mini Kits ( Qiagen ) , DNA was removed using RQ1 RNase-Free DNase ( Promega ) , and cDNA was generated using Reverse Transcription System ( Promega ) . IFNβ was normalized to β-actin ( ng IFNβ/ng β-actin ) . DNA was quantified directly following cell fractionation and DNA precipitation . Cytosolic DNA was normalized to undisturbed cell DNA ( ( ng cytosolic DNA ) / ( ng total undisturbed cell DNA ) *100 ) . Amount in ng of each gene was determined using standards that were generated independently of experimental samples and that contained abundant levels of each gene . The sequences were as follows: IFNβ fwd CAGCTCCAAGAAAGGACGAAC , rvs GGCAGTGTAACTCTTCTGCAT; β-actin fwd AGTGTGACGTTGACATCCGTA , rvs GCCAGAGCAGTAATCTCCTTCT; D-loop fwd AATCTACCATCCTCCGTGAAACC , rvs TCAGTTTAGCTACCCCCCCAAGTTTAA; SigF fwd GCGGGTCGGGCTGGTCAAC , rvs CCTCGCCCATGATGGTAGGAAC; dnaA fwd CGACAACGACGAGATTGATGA , rvs CGGTAGCGGAATCGGTATTG . BMDM were infected in 24-well plates . 48 hr post infection cell culture supernatants were collected from infected cells , filtered with 0 . 22 μm SpinX filters ( Corning ) , and assayed for mouse IFNβ ( PBL Interferon Source ) and TNF ( eBioscience ) . BMDM were infected in Lab-Tek chamber slides . Cells were washed at indicated time points and fixed in 1% PFA overnight at 4°C . Primary antibodies FK2 ( Millipore ) and Galectin-3 ( eBioscience ) and secondary antibodies Alexa Fluor 488 goat anti-mouse and Alexa Fluor 488 chicken anti-rat ( Invitrogen ) were used at 1:1000 . Slides were visualized on a Leica Leitz DMRB upright microscope . For superoxide analysis BMDM were incubated with 5 μM MitoSOX in DMEM for 25 min at 37°C , washed in warm PBS , and fixed in 1% PFA overnight at 4°C . Slides were visualized on a Zeiss LSM710 Multiphoton microscope . Mean cell fluorescence intensity was determined using ImageJ . BMDM were infected in black-wall , clear-bottom 96-well plates . 24 hr post infection CCF4-AM with 2 . 5 μM probenecid was added to cells at room temperature ( LiveBLAzer FRET-B/G Loading Kit; Invitrogen ) . After 2 hr cells were washed 3x in PBS with 2 . 5 μM probenecid . Cells were fixed in 4% PFA for 30 min and then overnight in 1% PFA at 4°C . Slides were visualized on a Nikon Eclipse Ti inverted microscope and 450nm:520nm ratios of the area directly surrounding each bacterium were determined using NIS-Elements Imaging Software . BMDM were grown for 6 days using 20 ng/mL M-CSF ( PeproTech ) and then were plated in 96-well plates in M-CSF media without glucose and with 5 mM galactose . Cells were grown for 24 hours in the presence of galactose and absence of glucose prior to infection and throughout infection . On day 7 BMDM were infected at the MOI reported in the text . 24 hr post infection intracellular ATP was measured using a Luminescent ATP Detection Assay Kit ( Abcam ) . Protein from TN1 lysates or phenol extracts was denatured at 95°C for 5 min and quantified using Pierce BCA Protein Assay Kit . Samples were normalized to total amount of protein , separated by SDS-PAGE on 12% Tris-HCl gels ( BioRad ) , and transferred onto 0 . 2 μm nitrocellulose membranes . Blots were incubated with anti-β-actin ( Cell Signaling ) , anti-Complex Vα ( Abcam ) , and anti-pyruvate dehydrogenase E1α ( Abcam ) .
|
Bacterial strains from the Mycobacterium tuberculosis complex ( MTBC ) infect one in three humans , however not all infected individuals progress to active tuberculosis disease . It is unknown why immunocompetent individuals develop tuberculosis , and this presents a significant challenge in preventing the disease . One proposed explanation is that individuals that progress produce higher levels of the innate cytokine type I interferon , due to bacterial and/or host determinants . Therefore we set out to determine whether MTBC strains from distinct phylogenetic lineages induce distinct levels of type I interferon , and to determine a mechanism of type I interferon induction by each strain . We found that a Lineage 6 strain induced lower levels of type I interferon in macrophages than Lineage 2 and 4 strains . Additionally , this strain induced low levels of mitochondrial stress . We showed that improving mitochondrial function by treating macrophages with a mitochondria-specific antioxidant reduced type I interferon induction by Lineage 2 and 4 strains . Our results indicate that differential mitochondrial stress contributes to differential type I interferon induction by distinct MTBC strains . This suggests that mitochondria-specific antioxidants may be a means of reducing the pathogenic type I interferon response .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"chemical",
"compounds",
"oxides",
"mitochondrial",
"dna",
"physiological",
"processes",
"forms",
"of",
"dna",
"mitochondria",
"dna",
"bioenergetics",
"cellular",
"structures",
"and",
"organelles",
"bacteria",
"proteins",
"chemistry",
"actinobacteria",
"biochemistry",
"cell",
"biology",
"nucleic",
"acids",
"mycobacterium",
"tuberculosis",
"physiology",
"interferons",
"genetics",
"biology",
"and",
"life",
"sciences",
"secretion",
"physical",
"sciences",
"energy-producing",
"organelles",
"superoxides",
"cytosol",
"antioxidants",
"organisms"
] |
2016
|
The Mechanism for Type I Interferon Induction by Mycobacterium tuberculosis is Bacterial Strain-Dependent
|
Among the several new antimalarials discovered over the past decade are at least three clinical candidate drugs , each with a distinct chemical structure , that disrupt Na+ homeostasis resulting in a rapid increase in intracellular Na+ concentration ( [Na+]i ) within the erythrocytic stages of Plasmodium falciparum . At present , events triggered by Na+ influx that result in parasite demise are not well-understood . Here we report effects of two such drugs , a pyrazoleamide and a spiroindolone , on intraerythrocytic P . falciparum . Within minutes following the exposure to these drugs , the trophozoite stage parasite , which normally contains little cholesterol , was made permeant by cholesterol-dependent detergents , suggesting it acquired a substantial amount of the lipid . Consistently , the merozoite surface protein 1 and 2 ( MSP1 and MSP2 ) , glycosylphosphotidylinositol ( GPI ) -anchored proteins normally uniformly distributed in the parasite plasma membrane , coalesced into clusters . These alterations were not observed following drug treatment of P . falciparum parasites adapted to grow in a low [Na+] growth medium . Both cholesterol acquisition and MSP1 coalescence were reversible upon the removal of the drugs , implicating an active process of cholesterol exclusion from trophozoites that we hypothesize is inhibited by high [Na+]i . Electron microscopy of drug-treated trophozoites revealed substantial morphological changes normally seen at the later schizont stage including the appearance of partial inner membrane complexes , dense organelles that resemble “rhoptries” and apparent nuclear division . Together these results suggest that [Na+]i disruptor drugs by altering levels of cholesterol in the parasite , dysregulate trophozoite to schizont development and cause parasite demise .
Billions of people living in regions endemic for malaria are confronted with the looming threat of Plasmodium falciparum parasites resistant to currently effective artemisinin combination therapies [1] . From an evolutionary point of view , emergence of resistant parasites has to be anticipated , especially in light of the fact that the drug pressure is being applied on a vast population of parasites , whose transmission requires obligatory sexual reproduction favoring recombinatorial selection of beneficial drug resistance alleles . Thus for a foreseeable future , efforts to control and eliminate malaria will require a robust pipeline of antimalarial drugs under development . Over the past decade , efforts by academic and industrial investigators have begun to prime this pipeline with new chemical entities with potent antimalarial activity [2] . Understanding the mechanism by which these new compounds cause the demise of malaria parasites would reveal vulnerable aspects of parasite physiology , which in turn could identify other new potential drug targets for further investigations . Three new antimalarials designated as clinical candidates , each with a distinct chemical structure , appear to share a common mode of action . These antimalarial drugs—belonging to the spiroindolone [3 , 4] , pyrazoleamide [5] and dihydroisoquinolone ( DHIQ ) [6] chemical classes ( see Fig 1A for structures ) were all shown to induce a rapid influx of Na+ into isolated trophozoite stages of P . falciparum . An investigation of about 400 antimalarial compounds included in the “Malaria Box” supplied by the Medicines for Malaria Venture ( MMV ) revealed 28 compounds belonging to 16 distinct chemical classes that also caused rapid Na+ influx into isolated P . falciparum trophozoites [7] . Resistance to several of the compounds causing Na+ influx was found to be associated with mutations within a P-type cation ATPase , PfATP4 [4 , 5 , 6] , a plasma membrane transporter initially annotated as a non-SERCA Ca2+ ATPase pump [8] , but now revealed to have characteristics of a Na+ pump [3] . The existence of resistance-associated mutations in PfATP4 has been interpreted as indicating that the large number of chemical classes causing Na+ influx into isolated trophozoites are all direct inhibitors of PfATP4 , and that it is the inhibition of the Na+ efflux function of PfATP4 that leads to the rapid increase in [Na+]i . While this appears to be a reasonable interpretation , there is no direct evidence at this time to show that PfATP4 is indeed a Na+ pump , or that various antimalarials causing [Na+]i disruption directly inhibit its Na+ pumping activity . Genetic investigations of pyrazoleamide-resistant P . falciparum have revealed potential epistatic interactions between PfATP4 and regulatory molecules such as a Ca2+-dependent protein kinase [5] . Thus , it appears possible that Na+ homeostasis within the intraerythrocytic stages of P . falciparum might be regulated by a complex network , and that perturbation of this network by a variety of small molecules could also lead to Na+ homeostasis disruption . Cellular life depends upon maintenance of appropriate intracellular ionic concentrations . All cells expend substantial amount of energy for this purpose . Numerous transporters and their associated regulators manage ionic homeostasis , especially for intracellular [H+] , [Na+] , [K+] and [Ca2+] . This results in concentration gradients of ions across the plasma membranes , which in turn permit rapid changes in intracellular ionic concentrations , the basis for a wide range of signaling cascades affecting cellular physiology . Maintenance of ionic homeostasis is especially challenging to an intracellular eukaryotic parasite such as Plasmodium . Intraerythrocytic stages of the parasite are confronted with a changing ionic environment of the erythrocyte cytosol . As the parasite matures , new permeability pathways induced by the parasite result in increased [Na+] within the host cell [9] . Hence , Na+ pumping by PfATP4 is proposed as a means to maintain physiological levels of [Na+]i [3] . Although drug-induced influx of Na+ within the parasite is likely to be disruptive to parasite physiology , details of the events triggered by the resultant ionic imbalance that eventually lead to parasite death have not been explored . We report here dramatic alterations in parasite plasma membrane permeability and morphology induced by two chemically distinct [Na+]i disruptor drugs . These changes resemble the last stages of schizogony , which leads us to suggest Na+ influx as a signaling event that is prematurely induced by these antimalarial drugs .
The initial publication on the spiroindolone KAE609 reported rapid protein synthesis inhibition in intraerythrocytic P . falciparum as a possible reason for the parasite demise ( Fig 2 in Rottman et al . [4] ) . The authors used 35S-amino acid incorporation into trichloroacetic acid ( TCA ) insoluble radioactive material over a 1 h exposure to the drug to assess protein synthesis by control and drug-treated parasites . The TCA precipitation was carried out after the parasites were released by treatment with saponin to eliminate the quenching effect of hemoglobin released from the erythrocytes . When we assessed protein synthesis inhibition by autoradiography of newly synthesized proteins displayed by SDS-polyacrylamide gel electrophoresis ( PAGE ) , we also found a much reduced amount of labeled proteins in saponin-released KAE609 and PA21A050-treated parasites ( Fig 1B ) . However , when parasites were released by a milder treatment using anthrolysin-O ( ALO ) , a cholesterol-dependent cytolysin [10] , we observed a minimal reduction of newly synthesized proteins in KAE609- and PA21A050-treated samples , as determined by autoradiography ( Fig 1C ) . We used cycloheximide and artemisinin , as positive and negative controls respectively , and observed the expected protein synthesis inhibition by cycloheximide and minimal inhibition by artemisinin , whether the parasites were freed by saponin or ALO . These results suggested that a short treatment with KAE609 or PA21A050 causes the parasite plasma membrane to become sensitive to saponin detergent , resulting in the leakage of cytosolic proteins , and that these drugs do not induce rapid protein synthesis inhibition . Saponin treatment is widely used to free intact parasites from erythrocytes without apparent permeation of the parasite plasma membrane ( PPM ) , and has permitted extensive investigations of transport across the PPM [11–15] . This is likely due to the highly reduced cholesterol content of the PPM compared to the erythrocyte membrane and the parasitophorous vacuolar membrane ( PVM ) [16–18] . The increased susceptibility of the parasites to permeation by saponin in drug-treated parasites led us to examine this phenomenon in some detail . Both PA21A050 and KAE609 induced saponin sensitivity to the parasites in a dose dependent manner , as judged by the loss of aldolase , a cytosolic protein . Treatment with artemisinin did not cause saponin-induced aldolase leakage ( S1B Fig ) . Exposure to PA21A050 and KAE609 did not affect the level of an integral PVM protein Exp2 ( Fig 2A and 2B ) , or the level of an integral PPM protein PfATP4 ( S1A Fig ) . This indicated that the loss of cytosolic aldolase likely occurred through saponin-induced pores within the PPM of treated parasites rather than by dissolution of the membranes . The effective concentrations for 50% loss of aldolase ( EC50 ) were about 10 fold higher in this 2 h assay compared to the EC50 observed for parasite growth inhibition carried out for 48 h ( Fig 2C ) . Induction of saponin-sensitivity was also observed when freed parasites were exposed to the drugs . This was observable within 30–60 min of exposure ( Fig 2D and 2E ) . Furthermore , a pyrazoleamide-resistant P . falciparum line , Dd2-R21 [5] , did not develop saponin sensitivity when exposed to PA21A050 , thereby showing a clear link between the antimalarial drug action and the acquisition of saponin sensitivity ( Fig 2F ) . The Dd2-R21 line is not cross-resistant to KAE609 [5] , and therefore acquired saponin sensitivity when exposed to this drug ( Fig 2G ) . Parasites released from the erythrocytes by exposure to hypotonic buffer did not acquire saponin sensitivity ( S2 Fig ) . Taken together , these data establish that the rapid induction of the saponin sensitivity of the trophozoites is a result of exposure to these drugs . Saponin is a mixture of cholesterol-dependent glycoside detergents similar to digitonin . We found that digitonin also caused aldolase leakage in treated parasites ( S1C Fig ) . These results strongly suggested that exposure to these new antimalarials rapidly leads to cholesterol acquisition by the parasites . We assessed this by using methyl-β-cyclodextrin ( MβCD ) to extract cholesterol from membranes . MβCD is composed of a cyclic polymer of 7 glucose monomers bearing varying degrees of methylation . Thus , MβCD , while being water soluble , possesses a hydrophobic cavity within the cyclic polymer with the capacity to extract highly hydrophobic cholesterol from lipid bilayers . Measurement of cholesterol content in freed parasites using a biochemical assay ( Amplex Red Cholesterol Assay kit; ThermoFisher Scientific ) showed >50% reduction in total cholesterol content after the freed parasites were extracted with 5 mM MβCD ( S1 Table ) . Re-incubation of MβCD-extracted ( i . e . cholesterol depleted ) freed parasites with cholesterol-saturated MβCD resulted in repletion of the total cholesterol content of the sample ( S1 Table ) . Thus , the use of MβCD with or without added cholesterol permits assessment of the involvement of cholesterol in inducing saponin sensitivity following drug treatments . For many years , MβCD extraction has been widely employed for investigating cholesterol dynamics in cell biology [17 , 19–21] . Freed parasites were subjected to MβCD extraction , either prior to , or after the treatment with PA21A050 , which was then followed by assessment of the saponin-sensitivity of the PPM as revealed by aldolase leakage . Fig 3A shows that cholesterol extraction by MβCD either prior or subsequent to the treatment with PA21A050 prevented saponin-dependent aldolase leakage from the parasite . Similar results were also seen when KAE609 was used to treat parasites ( Fig 3B ) . Thus , ~50% depletion of cholesterol content of freed parasites ( S1 Table ) appears to eliminate drug-induced saponin sensitivity of the PPM . Pre-loading MβCD with cholesterol ( which would prevent its ability to extract cholesterol from the parasites ) prior to its use in these experiments resulted in aldolase leakage when exposed to saponin ( Fig 3A and 3B , lanes 5 and 8 ) . We also observed that freed parasites extracted with MβCD , while being resistant to saponin-mediated aldolase leakage when exposed to PA21A050 and KAE609 , were made sensitive to saponin if co-incubated with MβCD preloaded with cholesterol ( Fig 3C and 3D ) . Exposure of MβCD extracted freed parasites to cholesterol-loaded MβCD did not result in acquisition of saponin sensitivity when not treated with the compounds ( Fig 3C and 3D ) , suggesting the possibility of an active process that excludes cholesterol from the parasites under normal conditions ( discussed further below ) . Overall these observations provide strong support to the interpretation that treatment with these compounds results in rapid cholesterol incorporation into the parasite including PPM , and that the source of cholesterol is external to the parasite , likely being derived from the PVM , the RBC plasma membrane , or both . Cholesterol incorporation into the usually cholesterol-deficient PPM could have significant impact on the functioning of PPM-resident proteins such as transporters . Since the rapid increase in [Na+]i in treated parasites ( S3 Fig ) is proposed to be due to inhibition of a P-type Na+ pump , PfATP4 , we wondered whether the disruption of Na+ homeostasis by these drugs preceded cholesterol acquisition by the parasites , or whether the elevated cholesterol caused inhibition of Na+ pumping by PfATP4 , leading to an increased [Na+]i . To answer this question , we took advantage of P . falciparum parasites adapted to grow in a low [Na+] growth medium [22] . Parasites growing in low [Na+] medium would be exposed to low [Na+] within the host cell cytosol , since Na+ concentration within the host cell would be expected to equilibrate with the growth medium . Treatment of low [Na+] adapted parasites with the drugs did not result in induction of saponin sensitivity of the PPM ( Fig 4A and 4B ) . These results suggest that cholesterol acquisition in treated parasites is a consequence of Na+ homeostasis disruption . We observed that drug treatment of the ring-stage parasites did not result in the acquisition of saponin-sensitivity ( Fig 4C ) , which is consistent with the fact that ring-stage parasites do not display new permeability pathways and thus do not have high [Na+] within the host cell cytosol [9] . These results strongly support the notion that the influx of Na+ into the parasite cytosol is a requisite step preceding cholesterol accumulation in parasites treated with these drugs . As reported previously , Na+ influx induced by treatment with KAE609 and PA21A050 is also accompanied by an increase in cytosolic pH within the parasites , presumably due to an inhibition of PfATP4 wherein Na+ pumping is believed to be countered by H+ import . Thus , the effects we describe here could potentially be due to an increase in cytosolic pH . We assessed this possibility by treating trophozoite stage parasites with maduramicin , a Na+ ionophore anti-coccidial drug ( structure in the inset of Fig 5A ) , that has recently been shown to have potent antimalarial activity and synergy with PA21A050 [23] . As an ionophore , maduramicin is not likely to alter cytosolic pH , and thus effects of Na+ influx could be assessed independent of the pH change . As shown in Fig 5A , maduramicin inhibited P . falciparum growth with EC50 of 0 . 44 nM . Within the range of concentrations used for parasite growth inhibition , no toxic effects or morphological changes were apparent in uninfected erythrocytes . Maduramicin induced rapid influx of Na+ into freed parasites ( Fig 5B ) at a very low concentration with half-maximal effect observed at 4 nM of the compound ( Fig 5C ) . The initial rate of Na+ influx was much more rapid compared to that observed for KAE609 and PA21A050 ( compare Fig 5B with S3 Fig ) . This is likely due to the ionophoric activity of maduramicin , which is expected to rapidly equilibrate [Na+] across a membrane , compared to the putative Na+ pump inhibitors . A 2 h treatment with maduramicin also induced saponin sensitivity to the PPM as judged by aldolase leakage with EC50 for this effect ( ~ 5 nM ) being about 10-fold higher than the EC50 for growth inhibition ( Fig 5D ) , as was observed for KAE609 and PA21A050 . The time course of cholesterol accumulation within the parasites following the treatment with maduramicin was similar to that seen for compounds KAE609 and PA21A050 ( Fig 5E; compare with Fig 2D and 2E ) . This suggests that the processes that are triggered by Na+ influx have similar lag period irrespective of the stimulus used for the influx . Malaria parasites do not synthesize cholesterol de novo [24] . In intraerythrocytic stages , the invading merozoite forms the PVM derived mostly from the RBC plasma membrane [16] , and thus , the PVM contains a significant amount of host-derived cholesterol . As the parasite grows within the RBC , it ingests the host cytosol through structures called cytostomes , which incorporate portions of both the PVM and PPM in the process . Cholesterol and cholesteryl ester are detectable within the parasite [17 , 25] but the PPM remains poor in cholesterol despite its close interaction with the cholesterol-rich PVM , as indicated by the resistance to cholesterol-dependent detergents . Therefore , we wondered whether the parasites might have an active process to exclude cholesterol from the trophozoite stage . We examined parasites that were treated with the drugs for 2 h followed by removal of the drugs and assessed the saponin sensitivity of their PPM at various time points . Remarkably , within 60 min following the removal of either PA21A050 ( Fig 6A ) or KAE609 ( Fig 6B ) there was a complete reversal of saponin sensitivity . We interpret these results as suggesting that an active process of the parasite excludes cholesterol from the trophozoite stage , that this process is inhibited by the increased [Na+]i caused by the drugs , and that the removal of the drugs following a 2 h exposure allows re-activation of the cholesterol exclusion process . PA21A050-treated trophozoite stage parasites showed a 20% reduction in 3H-hypoxanthine incorporation over a 24 h period following the 2 h exposure , whereas KAE609 treatment led to a 60% reduction ( Fig 6C ) . Reduced parasite growth was also observed over a 96 h period following the drug exposure ( Fig 6D ) . Cholesterol plays a critical role in membrane organization and formation of microdomains that display a higher concentration of signaling molecules including proteins embedded in the plasma membrane through GPI anchors [26–28] . Therefore , we next examined the effect of drug treatment on the localization of the merozoite surface protein-1 ( MSP1 ) , a GPI-anchored PPM protein [29 , 30] . MSP1 synthesis begins at the mid trophozoite stage and the protein is inserted into the PPM via a GPI anchor at its C-terminus . A 2 h exposure to 10 nM of either PA21A050 or KAE609 resulted in a dramatic clustering of MSP1 in >90% of the parasites ( Fig 7A and 7D ) . Washing the drugs off from the parasites restored even distribution of MSP1 along the PPM in 2 h ( Fig 7B ) . The MSP1 clustering was not observed when parasites adapted to grow in low [Na+] medium were exposed to PA21A050 or KAE609 ( Fig 7C ) . Immunoelectron microscopy also revealed clustered MSP1 in PA21A050 treated parasites ( Fig 7E ) . The clustering was also observed by IFA for another GPI-anchored protein , MSP-2 ( S4 Fig ) . These observations suggest that rapid cholesterol accumulation in drug treated parasites may lead to clustering of the GPI-anchored proteins , which is reversible upon removal of the drugs and concurrent with restoration of a low cholesterol content in the parasite . Electron microscopy of drug-treated parasites revealed dramatic alterations in parasite morphology ( Fig 8 ) . Control parasites exposed to the vehicle were all uni-nucleated ( Fig 8A; representative of images examined from 25 thin sections ) . In contrast , trophozoites exposed to KAE609 ( Fig 8B , panels a , c , e , f , i ) or PA21A050 ( Fig 8B , panels b , d , g , h , j ) for 2 h showed an acceleration in stage progression from trophozoites to aberrant schizont-like parasites . Images shown here are representative >75 thin sections examined for each drug treatment in two biological replicate experiment . Signs of schizogony included nuclei fragmentation ( Fig 8B , panel a ) resulting in accumulation of from two ( Fig 8B , panels b , c ) to six ( Fig 8B , panels d , e ) nuclear profiles , and initiation of cytokinesis , as evidenced by the wrapping of the inner membrane complex ( IMC ) around premature parasites that contain apparently mature rhoptries ( Fig 8B , panels f , g ) . Several parasites exhibited more advanced stages of this atypical schizogony with the appearance of individualized parasites floating in the RBC cytoplasm and surrounded by a PVM like membrane ( Fig 8B , panels h-j ) . Overall , 95% of the treated parasites showed at least one of the features associated with schizogony , with 70% of the parasites showing more than one nucleus in a given section . We wondered whether the observation of multiple nuclei in some of the treated trophozoites indicated stimulation of DNA synthesis by these compounds . To assess this , the DNA content of treated parasites was examined by flow cytometry of SYBR Green stained parasites . As shown in Fig 9 , the overall flow cytometry profiles of control and treated parasites did not change after a 2 h exposure to the drugs . Previous studies have shown that mid- to late stage trophozoites consist of sub-populations that have undergone DNA replication without nuclear division [31] . Thus , the acceleration of events normally associated with schizogony that are observed following the drug treatment does not appear to involve DNA replication per se , and the multiple nuclei seen in some of the treated parasites likely indicate an induction of nuclear division of polyploid nuclei present within various sub-populations of trophozoites [31] .
The results above provide a glimpse of a complex set of events that are triggered by an influx of Na+ into the cytoplasm of trophozoite stages of P . falciparum . By using a simple saponin sensitivity assay , we have demonstrated profound changes in plasma membrane permeability of the trophozoite stage P . falciparum induced by two chemically distinct antimalarial drugs . We recognize that this approach does not directly measure the parasite plasma membrane properties . However , the membrane organization of intraerythrocytic parasites makes it technically very difficult , if not impossible , to separate PPM from PVM to carry out direct biochemical assessment of their content . Close proximity of the two membranes also makes it very difficult to use fluorescent probes to monitor their relative cholesterol contents . The approach of fluorescence lifetime imaging microscopy to assess cholesterol content of PPM vs . PVM does not provide adequate resolution in live imaging of parasites and thus is not suitable for this purpose [18] . Thus , we have relied on the well-documented propensity of saponin ( and digitonin ) to insert pores within membranes in a cholesterol-dependent manner , leading to the leakage of cytosolic proteins , to provide evidence for dramatic but reversible changes in the cholesterol content of the PPM induced by these antimalarial drugs . Experiments described in Fig 3 , in which MβCD and MβCD loaded with cholesterol were used prior to saponin sensitivity assessment of treated and untreated freed parasites , provide strong support to our interpretations of rapid accumulation of cholesterol in the PPM by drug treated parasites . Reasons for the very low cholesterol content of the PPM during intraerythrocytic development of the parasites have remained unexplored at this point . One possibility is that the low cholesterol content of the PPM reduces its rigidity , thus permitting greater flexibility to the intraerythrocytic parasite as it traverses through the host’s circulation . Uninfected erythrocytes in circulation need to endure a high level of shear stress as they pass through the tight spaces of the peripheral vasculature , and the high degree of deformability afforded by their plasma membrane-associated cytoskeleton permits erythrocytes to withstand this stress [32 , 33] . Parasite-infected erythrocytes also face the same shear stress that too needs to be mitigated in some manner [34 , 35] . Cytoadherence of infected erythrocytes bearing later stages of P . falciparum to endothelial cells is one mechanism that minimizes exposure to the stress [36 , 37] . Here , we propose that the reduced cholesterol content of the PPM resulting in reduced rigidity is another way by which the parasite manages to withstand the shear stress of circulation . The results described here could also provide an explanation for the observation that , while the full lethal effects of the [Na+]i disruptor drugs on P . falciparum culture appear to take 24 to 48 h ( as judged by in vitro parasite clearance experiments [38]; F . J . Gamo , personal communication ) , parasites are cleared much more rapidly in vivo . Treatment with these drugs of immunodeficient mice , engrafted with human erythrocytes and infected with P . falciparum , cleared the parasites from circulation with a time course that was faster than any other antimalarials tested [5 , 6] . A clinical trial of KAE609 revealed that a single 30 mg dose resulted in clearance of both P . falciparum and P . vivax from the circulation of malaria patients to undetectable levels within 12 h [39] , an unprecedented finding that has raised enthusiasm for this drug ( and possibly others in its class ) as a new weapon against malaria [2 , 40–42] . We suggest that the accumulation of cholesterol into the PPM induced by KAE609 , PA21A050 and other [Na+]i disruptor drugs ( in addition to the morphological changes observed following drug exposure ) results in increased rigidity of the membrane , rendering the parasite vulnerable to damage and/or removal from the circulation . This proposition is supported by a recent report showing increased rigidity of KAE609-treated parasitized erythrocytes when assessed in a microfluidic device in vitro [43] . In addition to imparting saponin sensitivity to the parasite , the [Na+]i disruptor drugs also caused coalescence of the GPI-anchored plasma membrane proteins MSP1 and MSP2 . Cholesterol , in conjunction with other lipids within the membrane , is involved in appropriate formation of microdomains wherein GPI-anchored proteins are present in greater concentration [28] . Although MSP1 is synthesized during the trophozoite stage and is displayed within the PPM , its function is critical at the surface of merozoites released at the end of the intraerythrocytic development cycle , where it has been shown to form a complex with other merozoite surface proteins , playing an essential role in host cell recognition and invasion by merozoites [44–46] . Recently , MSP1 processing was shown to result in activation of its spectrin binding function and aid in parasite egress from the RBC [47]; however , drug treatment of trophozoites for 2 h , while causing clustering of MSP1 , did not result in its processing ( S5 Fig ) . The clustering of MSP1 observed here following the drug treatments likely results from the increased cholesterol within the PPM and the absence of other merozoite surface proteins required for appropriate display of the complex . It is interesting to note that , unlike the PPM of the trophozoite stage parasites , merozoites have been reported to contain a substantial amount of cholesterol [48] , which would be consistent with the suggestion that cholesterol is required for appropriate formation of the MSP1 containing complex involved in receptor interaction and invasion . A remarkable observation was the reversibility of cholesterol accumulation and MSP1 clustering within minutes after the [Na+]i disruptor drugs were washed away from the trophozoites . We interpret this to indicate the presence of an active process that excludes cholesterol from the parasite plasma membrane . We propose this putative “cholesterol pump” ( in a manner similar to the Niemann-Pick disease associated protein NPC-1 , which transports cholesterol from the endosomal compartment in mammalian cells [49] ) to be present in the PPM and being responsible for maintaining low cholesterol levels , thereby preventing increased rigidity of the PPM . The net advantage to the parasite would be protection against damage and removal when passing through the narrow labyrinths of the peripheral vasculature . At the time of merozoite formation , however , this process of cholesterol exclusion would need to be reversed so that the merozoite plasma membrane could acquire cholesterol necessary for the appropriate display of merozoite surface proteins . Furthermore , cholesterol in the merozoite membrane is likely to be required for its rigidity and for imparting proper shape to this invasive extracellular stage . We hypothesize that the [Na+]i disruptor drugs are able to reverse or inhibit this putative cholesterol pump . Although most of the parasites appear to undergo rapid reversal of cholesterol accumulation and MSP1 clustering upon the removal of the drugs following a 2 h treatment , viability of the parasites appeared to be reduced to varying degrees . Ultrastructural examination of treated parasites provides a possible explanation for this variation . The drug treatment appears to trigger events resembling schizogony in the nominally trophozoite stage parasites . There appeared to be significant variations among parasites with regard to the morphological changes that were apparent by electron microscopy ( Fig 8 ) . As indicated by flow cytometry of DNA content ( Fig 9 ) , the trophozoite stage parasites used in these studies contained a significant sub-population that had higher DNA content than the majority , indicating that these parasites had undergone DNA replication and possessed polyploid nuclei . A potential explanation for the variation in viability could be that the parasites with polyploid nuclei [31] might not be able to recover from the premature schizogony-like events ( nuclear division , formation of inner-membrane complex and rhoptries etc . ) induced by the drugs . The observation that cholesterol accumulation and MSP1 coalescence in the PPM were not seen in treated parasites adapted to grow in low [Na+] medium constitutes the main evidence to support the conclusion that the massive changes induced by a short term exposure to the drugs are consequences of increased [Na+]i within the parasite . Experiments with the Na+ ionophore maduramicin ( Fig 5 ) further support this conclusion . Furthermore , these changes are not observed in ring-stage parasites even when grown in regular medium ( Fig 4C ) . This is likely due to the fact that , in the absence of parasite-induced new permeability pathways , the host cell cytosol of the ring-stage parasites has yet to acquire high level of [Na+] , thus precluding the influx of excess Na+ into the parasite cytoplasm [9] . Influx of Na+ beyond a threshold level , then , could be a normal physiological signal for initiation of a complex set of events resulting in schizogony and merozoite formation . We hypothesize that an inappropriate Na+ influx induced by the new antimalarial drugs investigated here leads to parasite demise . Fig 10 provides a schematic to depicts this hypothesis . Green arrows and hammers indicate a normal process of parasite development . Under normal physiological conditions , developmental signaling event at the initiation of schizogony is hypothesized as being able to inhibit Na+ pumping by the PPM-localized PfATP4 , resulting in Na+ influx within the parasite cytoplasm . The increased [Na+]i provides further signaling that leads to the inhibition of a putative cholesterol pump , resulting in increased accumulation of cholesterol in the PPM required for appropriate display of merozoite surface proteins as well as for imparting rigidity to the membrane and its shape . We also hypothesize that the increased [Na+]i also constitutes a signal for further developmental progression such as nuclear division and formation of the inner membrane complex . Red arrows and hammers in Fig 10 indicate the processes hypothesized as being induced by the drugs . In this model , spiroindolones and pyrazoleamides prematurely usurp this finely tuned process by either directly inhibiting PfATP4 or by mimicking a developmental signal , either of which would result in influx of Na+ into the parasite cytoplasm . Premature inhibition of the putative cholesterol pump and induction of developmental progression , we hypothesize , would result in parasite death . This model predicts the existence of a complex cascade of events that are unleashed by Na+ influx . It would be of great interest to identify various players that may participate in this cascade .
P . falciparum lines 3D7 , Dd2 , and Dd2-R21 were cultured using type O+ human RBC ( purchased from Interstate Blood Bank , TN ) , in RPMI1640 supplemented with 15 mM HEPES , 2 g/L sodium bicarbonate , 10 mg/L hypoxanthine , 50 mg/L Gentamysin sulfate , and 0 . 5% Albumax ( cRPMI ) . Asexual stages of P . falciparum were maintained at 5% haematocrit in cRPMI at 37°C in a humidified incubator containing 90% N2 , 5% O2 and 5% CO2 . Parasite cultures were periodically tested for Mycoplasma contamination to ensure that they were free of Mycoplasma . Plasmodium aldolase antibody was purchased from Abcam ( cat no . ab38905 ) . Anti-hemagglutinin ( HA ) antibody was purchased from Santa Cruz Biotechnology ( cat . no . sc-393579 ) . Antibodies to MSP142 and Exp2 were raised in rabbits as described previously [50 , 51] . Rabbit anti-serum to MSP2 was obtained from MR4 ( Malaria Research and Reference Reagent Resource Center ) . P . falciparum 3D7 and/or Dd2 ATP4-HA , Dd2-R21 parasites were grown up to 5% parasitemia in cRPMI . Cultures were synchronized three times using 500 mM alanine in 10 mM HEPES , pH 7 . 4 [52] . Synchronized trophozoite stage parasites were treated with the indicated concentration of the compounds or DMSO . For most experiments the treatment was carried out for 2 h . For some of the experiments , the time of treatment was varied as described in figure legends . To assess reversibility of compound effect , the treated parasites were washed with cRPMI and incubated with compound-free medium for the indicated length of time . The treated and rescued parasites were processed according to the experimental design as described below . Synchronized trophozoite stage parasites were exposed to varying compound concentrations in cRPMI for 2 h , collected by centrifugation and resuspended in 5 volumes of cRPMI . Equal volumes of 0 . 04% saponin ( in cRPMI ) was added to achieve final concentration 0 . 02% . After quick inversions ( 4–5 times ) parasites were collected by centrifugation and washed with PBS twice at 1500x g for 5 min . Anthrolysin O [10] ( a kind gift from Dr . Richard Rest , Drexel University ) was activated with 5 mM dithiothreitol on ice for 30 min . Parasites were collected by centrifugation at 500x g for 5min and resuspended to 10% hematocrit in RPMI without Albumax to remove any remaining cholesterol from the culture medium . Anthrolysin O was added at final concentration of 3 . 2 μg/ml and incubated for 5 min at 37°C , followed by centrifugation at 1500x g for 5 min . The parasite pellet was resuspended in 20 x volumes of RPMI without Albumax . A second exposure to anthrolysin O at 8 μg/ml for 5 min at 37°C was carried out to ensure complete lysis of uninfected erythrocytes . Freed parasites were collected by centrifugation and washed with cRPMI to quench any remaining anthrolysin O with cholesterol present in Albumax . For hypotonic lysis , compound treated trophozoite stage parasites were collected by centrifugation at 500x g for 5min and resuspended in hypotonic solution ( 2 mM NaCl , 10 mM Tric-Cl , pH 7 . 4 ) and incubated for 5 min at 37°C . After the incubation , parasites were collected by centrifugation and washed twice with PBS and finally collected at 1500x g for 5 min . This was followed by SDS-PAGE and imuunobloting ( S2 Fig ) . In another set of experiments shown in S2A Fig , hypotonically released untreated parasites were exposed to 5 mM MβCD in IRPMI at 37°C for 30 min . MβCD treated parasites were washed twice with cRPMI . Hypotonically released and MβCD treated parasites both were subjected to saponin treatment and harvested by centrifugation . Parasites from different treatment regimens were subjected for SDS-PAGE and immunobloting . Freed parasite pellets were lysed in RIPA buffer ( 20 mM Tris HCl ( pH 7 . 5 ) , 150 mM NaCl , 1 mM Na2 EDTA , 1 mM EGTA , 1% NP-40 , 1% Na-deoxycholate , 2 . 5 mM Na-pyrophosphate , 1 mM β-glycerophosphate ) containing protease and phosphatase inhibitors ( Sigma-Aldrich , Inc , St . Louis , MO , USA ) by a brief sonication at 4°C . In some experiments , parasite pellets were lysed in SDS loading buffer directly . The lysates were centrifuged at 15 , 000x g at 4°C for 30 min and the supernatant was used for the immunoblots after protein estimation using the Bradford reagent ( Sigma ) . Before loading , the protein was mixed with gel loading buffer ( 50 mM Tris HCl pH 6 . 8 , 10 mM DTT/100 mM β-Mercaptoethanol , 2% SDS , 0 . 1% bromophenol blue , 10% glycerol ) and heated at 90°C for 5 min . Samples were resolved on 12% SDS–PAGE and proteins were transferred to methanol-activated polyvinylidene fluoride ( PVDF ) membrane ( Millipore ) using anode buffer ( 25 mM Tris HCl pH 10 . 4 , glycine , 10% Methanol ) and wet transferred for 2 h at 4°C . Membranes were blocked with 5% non-fat skim milk powder in 1x PBS overnight and probed with specific antibodies . Primary antibody dilution was made in 1x PBS containing Tween-20 ( 0 . 2% ) and incubated with the membrane for 3 h at RT on a rocker . Primary antibody binding was detected by appropriate secondary antibodies conjugated to horseradish peroxidase ( Santa Cruz Biotechnology ) . Dilution of secondary antibody was made in 1x PBS Tween-20 ( 0 . 2% ) . After each incubation , the membrane was washed with 1x PBS-Tween-20 ( 0 . 2% ) for 5 min at least 5–6 times . The immunoblots were developed using the SuperSignal West Pico chemiluminescent substrate ( Thermo Scientific , USA ) . IFA of P . falciparum 3D7 infected RBCs were performed in solution . Infected RBCs were centrifuged at 500g for 5 min , washed twice , and re-suspended in PBS . Cells were fixed using 4% formaldehyde/0 . 0075% glutarldehyde in PBS for 20 min at 4°C . All subsequent steps were carried out at room temperature ( 24–26°C ) . Cells were permeabilzed using 0 . 1% Triton-X 100 in PBS for 30 min and washed twice with PBS . 3% BSA was used for blocking . Antibodies to MSP1 were added at 1:1000 dilution in PBS containing 0 . 01% Triton X-100 and incubated for 3–4 hrs . Specificities of MSP1 and MSP2 antibodies have been shown by immunoblots ( S4 and S5 Figs ) . RBCs were pelleted at 500x g , washed and treated with appropriate Alexa 488- or 594-conjugated secondary antibodies ( Molecular Probes , invitrogen , USA ) at 1:500 dilution for 2 hrs . After washing 3–4 times , cells were incubated for 5 min with DAPI ( 0 . 1 μg/ml ) . Cells were imaged using a Nikon ECLIPSE TiE inverted microscope . Acquired IFA images were processed by Image J software . For electron microscopy , preparations of P . falciparum 3D7 infected erythrocytes ( DMSO or compound treated ) were fixed in 2 . 5% glutaraldehyde ( Electron Microscopy Sciences; EMS , PA ) in 0 . 1 M sodium cacodylate buffer ( pH 7 . 4 ) for 1 h at room temperature , and then processed for thin-section transmission electron microscopy as previously described [53] . Thin sections were examined with a Philips CM120 Electron Microscope ( Eindhoven , the Netherlands ) under 80 kV . For immunogold staining , infected erythrocytes were first fixed in 4% paraformaldehyde ( PFA from EMS ) in 0 . 25 M HEPES ( pH 7 . 4 ) for 1 h at room temperature , then in 8% PFA in the same buffer overnight at 4°C . Samples were infiltrated , frozen and sectioned as described [54] . The sections were immunolabeled with rabbit anti-MSP1 antibody at 1/75 dilution in PBS/1% fish skin gelatin , followed by incubation with anti-IgG antibodies and 10 nm protein A-gold particles before electron microscopy . Parasite growth inhibition was assessed by a modified version of the method originally described by Desjardin et al . [55] . The method assessed parasite growth as reflected by incorporation of 3H-hypoxanthine by parasites . P . falciparum 3D7 parasites in culture were exposed to graded dilutions of test compounds for 48 h and incorporation of 3H-hypoxanthine into parasite nucleic acids during the last 24 h was determined by liquid scintillation spectroscopy . To assess rescue of treated parasites , synchronized P . falciparum parasites at 1% parasitemia in 3% hematocrit culture were treated with 10 nM PA21A050 or 10 nM KAE609 for 2h . Parasites were harvested and washed 3 times with low hypoxanthine-containing medium ( LHM ) and resuspended to the original volume with LHM . 100 μl of this culture was mixed with 100 μl of LHM containing 4 μC/ml 3H –hypoxanthine ( Perkin Elmer ) and plated in 8 wells of a 96 well plate . 3H-hypoxanthine incorporation was assessed as above . The remaining culture was resuspended in cRPMI and the parasitemia was monitored after every 24 h by Giemsa staining of thin blood smears . Synchronized P . falciparum 3D7 infected erythrocytes at trophozoite stage were gently treated with 0 . 02% saponin in cRPMI as described above . The freed parasites were immediately harvested by centrifugation and washed once with RPMI without Albumax . An aliquot of saponin-freed parasites was first treated with MβCD ( 5 mM in RPMI without Albumax ( IRPMI ) at 37°C for 30 min ) or MβCD loaded with cholesterol . MβCD was loaded with cholesterol at 1:5 MβCD:cholesterol ratio as described by Caliceti et al . [56] and used at final concentration of MβCD at 2 . 5 mM . Following the exposure to MβCD and MβCD-cholesterol , parasites were harvested by centrifugation and washed with cRPMI and resuspended in cRPMI containing the indicated amount of the compound . After 2 h of compound treatment , an aliquot of parasites was harvested and treated with 0 . 02% saponin in cRPMI followed by immediate harvesting by centrifugation . Another aliquot of freed parasite was first treated with the compounds for 2 h at 37°C , harvested by centrifugation and then extracted with MβCD and MβCD loaded with cholesterol as above . Parasites were collected by centrifugation , washed with cRPMI and then briefly treated with 0 . 02% saponin in cRPMI prior to SDS-PAGE and immunobloting . In another set of experiment , untreated saponin-freed parasites were treated with 2 . 5 mM MβCD in IRPMI at 37°C for 30 min . Parasites were then harvested and washed twice with IRPMI . Parasites were exposed to MβCD-cholesterol at final concentrations of 0 . 625 . 1 . 25 and 2 . 5 mM and incubated in the presence or absence of 10 nM PA21A050 or 10 nM KAE609 at 37°C for 2h . After the incubation , harvested parasites were subjected to saponin treatment followed by SDS-PAGE and immunobloting . Highly synchronized trophozoite stage parasites at 15% parasitemia ( 0 . 9 ml packed cell volume ) were gently freed by treatment with 0 . 2% saponin . Freed parasites were collected and resuspended in incomplete RPMI medium . Freed parasites were divided into 3 aliquots . One aliquot was left as freed parasites . A second aliquot was extracted with 5 mM MβCD in incomplete RPMI for 30 min at 37°C . The third aliquot was extracted with MβCD as above and then resuspended in cholesterol-saturated MβCD followed by a 30 min incubation at 37°C . Parasites were collected by centrifugation and their cholesterol content was determined by using the Amplex Red Cholesterol Assay kit ( ThermoFisher; catalog no . A12216 ) using the protocol recommended by the manufacturer . P . falciparum parasites were adapted to grow in low [Na+] medium according to the method described by Pillai et al . [22] . Briefly , the medium contained all ingredients specified for RPMI1640 except that NaCl , NaHCO3 , and Na2HPO4 were replaced by 64 . 8 mM KCl , 28 . 6 mM KHCO3 , and 5 . 64 mM K2HPO4 . In addition , 84 . 3 mM sucrose was included in the medium . Addition of 10% human serum to this medium was estimated to result in about a 7 mM Na+ concentration . Using this low [Na+] medium , P . falciparum parasites were adapted by growing them for 10–12 generations . These adapted parasites were then used for various experiments . Intracellular [Na+] was determined using methods described by Spillman et al . [3] . Briefly , saponin freed parasites were loaded with the ratiometric sodium sensitive probe SBFI-AM ( 5 . 5 μM ) ( Invitrogen ) and 0 . 01% w/v Pluronic F-127 ( Invitrogen ) in suspension ( at 2 . 5–3 . 5 x 108 parasites/mL ) for 30min at 37°C in bicarbonate-free RPMI supplemented with 20 mM glucose , 10 mg/L hypoxanthine , 25 mM HEPES and 50 mg/L gentamycin sulphate ( pH 7 . 1 ) . The probe-loaded parasites were washed twice ( 2 , 000x g , 30 s ) and resuspended to a final parasite concentration of 1 . 0–1 . 5 x 108/mL in a saline buffer ( 125 mM NaCl , 5 mM KCl , 1 mM MgCl2 , 20 mM glucose , 25 mM HEPES , pH 7 . 1 ) . SBFI loaded parasites were excited at 340 nm and 380 nm with emissions recorded at 500 nm in a Horiba Scientific FluoroMax 3 spectrofluorometer . Auto-fluorescence corrected 340/380 nm emission fluorescence ratio were related to [Na+]i using an average from 3 independent calibration curves for SBFI . Calibration plots were generated for SBFI loaded parasites in solutions containing 0 , 10 , 25 , 50 , 75 , 100 , 130 mM Na+ , made by mixing solutions of 80 mM Na+/K+ gluconate and 50 mM NaCl/KCl ( 1 mM MgCl2 , 20 mM glucose , 25 mM HEPES , pH 7 . 1 ) and [Na+]i was equilibrated using a combination of the ionophores nigericin ( 5 μM ) , gramicidin ( 2 . 5 μM ) and monensin ( 5 μM ) . Protein synthesis was assessed by labeling with 35S-methinonine/35S-cysteine over a 2 h period in the absence or the presence of 10x EC50 of PA21A050 ( 10 nM ) , KAE609 ( 10 nM ) , cycloheximide ( 2000 nM ) , or artemisinin ( 100 nM ) . Briefly , parasite culture at 10% parasitemia was washed with methionine/cysteine free RPMI1640 supplemented with 0 . 5% Albumax . Labeling was done in 1 ml cultures at 5% hematocrit in methionine/cysteine free RPMI1640 containing 125 μCi/ml of Easy Tag 35S Protein Labeling Mix ( Perkin Elmer ) . After 1 h of incubation , 0 . 5 ml of regular cRPMI supplemented with 1 mg/ml of unlabeled methionine and cysteine plus the indicated amount of the compounds was added to the cultures , and incubation was continued for an additional 1 h . The cultures were divided into two 1 . 5 ml tubes , centrifuged , and washed twice with Albumax-free RPMI . Parasites were resuspended and treated with either 0 . 02% saponin or anthrolysin O as described above . Parasite pellets were resuspended in SDS-PAGE loading buffer , followed by SDS-PAGE and autoradiography . Trophozoite stage P . falciparum 3D7 parasites at 6–7% parasitemia were treated with the vehicle ( DMSO ) , 10 nM PA21A050 or 10 nM KAE609 for 2 h . Parasites were fixed with formaldehyde/glutaraldehyde and stained with the DNA-specific fluorescent dye SYBR Green . Parasites were washed twice and resuspended in 1X PBS as a single cell suspension . For each treatment , 3 million events were counted by flow cytometry using a Becton Dickinson LSR Fortessa Cell Analyzer , and the data were analyzed by FlowJo single cell analysis software . The pseudocolor density dot plot was used to represent the data .
|
Malaria remains a major public health challenge in the world , especially with the realization that parasites causing the disease are becoming resistant to currently used antimalarial drugs . There are new antimalarial drugs under development , and among these are 3 clinical candidate drugs that have the propensity to cause sodium leakage into parasites growing inside human red blood cells . We have investigated events that lead to parasite death when sodium concentration inside the parasite increases . Our findings suggest that the drug-treated parasite rapidly acquires cholesterol and clusters containing lipid-embedded proteins MSP1 and MSP2 form within the plasma membrane . Because these changes are reversible when the drugs are removed , we suggest that there is an active process that keeps cholesterol out of the parasite . We also observed massive morphological changes resembling premature steps of parasite division following drug treatment . The changes we describe appear to be a direct consequence of increased sodium level in the parasites . We hypothesize that sodium influx constitutes a normal signaling process in malaria parasites and that the new antimalarial drugs initiate this process prematurely , which results in the death of the parasites .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"trophozoites",
"parasite",
"groups",
"medicine",
"and",
"health",
"sciences",
"plasmodium",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"parasitology",
"membrane",
"proteins",
"apicomplexa",
"protozoans",
"pharmaceutics",
"cellular",
"structures",
"and",
"organelles",
"lipids",
"malarial",
"parasites",
"cell",
"membranes",
"cholesterol",
"biochemistry",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"drug",
"therapy",
"organisms"
] |
2016
|
Na+ Influx Induced by New Antimalarials Causes Rapid Alterations in the Cholesterol Content and Morphology of Plasmodium falciparum
|
An enduring theme in pathogenic microbiology is poor understanding of the mechanisms of host specificity . Metarhizium is a cosmopolitan genus of invertebrate pathogens that contains generalist species with broad host ranges such as M . robertsii ( formerly known as M . anisopliae var . anisopliae ) as well as specialists such as the acridid-specific grasshopper pathogen M . acridum . During growth on caterpillar ( Manduca sexta ) cuticle , M . robertsii up-regulates a gene ( Mest1 ) that is absent in M . acridum and most other fungi . Disrupting M . robertsii Mest1 reduced virulence and overexpression increased virulence to caterpillars ( Galleria mellonella and M . sexta ) , while virulence to grasshoppers ( Melanoplus femurrubrum ) was unaffected . When Mest1 was transferred to M . acridum under control of its native M . robertsii promoter , the transformants killed and colonized caterpillars in a similar fashion to M . robertsii . MEST1 localized exclusively to lipid droplets in M . robertsii conidia and infection structures was up-regulated during nutrient deprivation and had esterase activity against lipids with short chain fatty acids . The mobilization of stored lipids was delayed in the Mest1 disruptant mutant . Overall , our results suggest that expression of Mest1 allows rapid hydrolysis of stored lipids , and promotes germination and infection structure formation by M . robertsii during nutrient deprivation and invasion , while Mest1 expression in M . acridum broadens its host range by bypassing the regulatory signals found on natural hosts that trigger the mobilization of endogenous nutrient reserves . This study suggests that speciation in an insect pathogen could potentially be driven by host shifts resulting from changes in a single gene .
An enduring theme in pathogenic microbiology is poor understanding of the mechanisms of host specificity . That is , what factors limit a pathogens host range , how is host specificity linked with virulence and what changes in pathogens or hosts can open new host ranges ? These are fundamental questions that relate both to the co-evolution of host susceptibility and pathogen virulence , as well as to factors underlying host switching and the emergence of new pathogens that originate in different host species . The molecular mechanisms controlling host selectivity in fungi are particularly poorly understood . A number of plant pathogenic fungi produce secondary metabolites with biological specificities that correspond with the host range of the producing fungi [1] , [2] , and small secreted proteins produced by the pathogen can trigger resistance in some plants limiting host range [3] . However , to date these have not been found in animal pathogenic fungi , many of which seem to be broadly opportunistic . This has allowed researchers on emergent human pathogenic fungi to employ insects as model systems [4] . Most human pathogens are not normally transmitted between insect hosts . However , specialization to entomopathogenicity is a major fungal lifestyle with ∼1000 known species that can be highly infectious . In contrast to the opportunistic pathogens , many of these have evolved narrow host ranges by as yet unknown mechanisms . Metarhizium is a cosmopolitan genus of Ascomycetes ( class Sordariomycetes ) comprising species that exhibit varied lifestyles . Metarhizium robertsii ( formerly known as M . anisopliae var . anisopliae [5] ) is a generalist able to infect hundreds of insect species . It has been at the forefront of efforts to develop biocontrol alternatives to chemical insecticides in agricultural and human disease-vector control programs [6]–[8] . M . robertsii has also been used to study the interactions between invertebrate model hosts and pathogenic fungi as host innate immune responses are broadly conserved across many phyla [9] . In contrast M . acridum is a specialist with a narrow host range for certain locusts and grasshoppers [10] . This specificity is one of the reasons it is being mass produced as an environmentally safe alternative to pesticides [10]–[12] . The infection strategies of Metarhizium species resemble those of most plant pathogens . Infection proceeds via spores that adhere to the host surface and germinate to form a germ tube that continues undifferentiated hyphal growth if nutrient quality and quantity are not conducive for differentiation of infection structures . On a host , however , apical elongation terminates and germ tubes produce infection structures , called appressoria , which promote the localized production of cuticle degrading enzymes and also build up turgor providing a mechanical component to penetration [13]–[15] . The formation of appressoria by broad host range strains of M . robertsii such as ARSEF2575 ( Mr2575 ) can also be induced efficiently by low levels of complex nitrogenous nutrients [13] . However , pathogens with a narrow host range such as M . acridum ARSEF324 ( Ma324 ) ( = CSIRO FI 485: the active ingredient of ‘Green Guard’ ) germinate poorly under these conditions and only produce abundant appressoria in the presence of a lipid extract from host insects [16] . M . robertsii Mr2575 sharply up-regulates the Mest1 gene when germinating on insect cuticles [17] . This gene is absent in Ma324 [18] , which suggests that it is unlikely to have an essential function in the related M . robertsii but we speculated that it might have a niche role in pathogenicity , perhaps facilitating an opportunistic life-style . In this study , we functionally characterized MEST1 and demonstrated that inserting Mest1 into M . acridum is sufficient to expand its host range to include lepidopterans . Mest1 is thus the first gene identified in an entomopathogenic fungus that encodes a determinant of specificity and is to our knowledge the first example where a single metabolic protein assumes such a crucial role for host selectivity in a animal pathogenic fungus . We speculate that speciation in insect pathogens can be driven by host shifts that become fixed in populations due to the gain or loss of a pathogen gene that confers wide host specificity .
The M . robertsii Mest1 is 1188 bp long and lacks introns . It encodes a predicted protein of 395 amino acids , with a deduced molecular weight of 42244 Da and a pI of 5 . 34 . The SignalP 3 . 0 program ( http://www . cbs . dtu . dk/services/SignalP/ ) revealed no signal sequence suggesting that MEST1 is a cell-bound protein . MEST1 contains the sequence GGS341VG which conforms to the motif G-X1-S-X2-G , commonly observed in serine esterases and many lipases [19]–[21] . According to M . robertsii genome sequence data [22] , Mest1 ( MAA_03283 ) is downstream of three secreted cuticle degrading subtilisins ( Pr1F ( MAA_03280 ) , Pr1E ( MAA_03281 ) and a subtilisin-like protease ( MAA_03282 ) . The M . robertsii genome also contains a paralog ( MAA_08059 ) with 42% identity to MEST1 that is downstream of three hypothetical proteins and upstream of a glutathione S-transferase . The genome of M . acridum strain CQMa 102 lacks an ortholog of MEST1 but contains a sequence ( MAC_02852 ) with 90 . 2% identity to MAA_08059 ( the median sequence identity of orthologs is 89 . 8% ) . MAC_02852 is downstream of four hypothetical proteins and upstream of glutathione S-transferase indicating that it is syntenic with its ortholog MAA_08059 . Homologs to MEST1 and MAA_08059 were identified in four other ascomycete fungi , but only the related Nectria haematococca had a sequence ( EEU_38198 ) that was highly similar ( 82% identity ) to MEST1 . The related Gibberella zeae ( XP_380200 ) as well as Aspergillus niger ( XP_001397035 ) and Penicillium chrysogenum ( XP_002567997 ) contain single copy sequences resembling MAA_08059 ( Figure 1 ) . The functions of these genes have not been reported and contain a putative penicillin-binding domain characteristic of β-lactamase class C proteins . Homologs of MEST1 were absent in Neurospora crassa , Magnaporthe grisea , Schizosaccharomyces japonicus , Trichoderma harzianum and T . reesei , as well as Basidiomycetes , Zygomycetes and Chytridiomycetes . However , MEST1 shows up to 42% identity with bacterial sequences , including known esterases [19] . A phylogenetic tree ( Figure 1 ) confirmed that MEST1 and the fungal homologs formed a separate well supported clade distinct from bacterial clades containing actinomycete and pseudomonad sequences . To study the function of Mest1 , Mest1 null mutants ( ΔMest1 ) , were generated in Mr2575 by homologous replacement ( Figure 2 and Table 1 ) . Pathogenicity assays against Galleria mellonella and Manduca sexta caterpillars revealed a significant reduction in mortality and speed of kill by ΔMest1 relative to wild type M . robertsii Mr2575 ( Figure 3A and 3B ) , while virulence against acridid grasshoppers ( Melanoplus femurrubrum ) was unaltered ( Figure 3C ) . To demonstrate that the altered phenotype of ΔMest1 was specifically due to gene inactivation , the Mest1 gene was reintroduced into ΔMest1 in single copy . Six isolates of the resulting complemented strain ( Mest1-Com ) infected Galleria in an identical fashion as the wild type indicating successful complementation . Overexpression of Mest1 under control of the constitutive glyceraldehyde-3-phosphate dehydrogenase ( gpd ) promoter ( Mr2575-gpd::Mest1 ) resulted in a significant increase in virulence against both caterpillars ( Manduca and Galleria ) and grasshoppers ( Figure 3C ) . Mr2575 transformed with an additional copy of Mest1 under control of its native promoter increased pathogenicity to lepidopterans but not grasshoppers , suggesting that Mest1 is not activated by the wild type fungus on grasshoppers . To elucidate whether Mest1 is needed for developmental processes , wild type Mr2575 , ΔMest1 and Mest1-Com strains were grown on SDA ( Sabouraud dextrose agar ) or SDB ( Sabouraud dextrose broth ) ( nutrient rich conditions ) , in 0 . 01% YE ( Yeast extract ) plus Manduca cuticular lipids ( 0 . 25 mg/ml ) and on ground grasshopper or Manduca cuticles . There was no significant difference in sporulation , germination and growth rates between M . robertsii Mr2575 , ΔMest1 and Mest1-Com on SDA or SDB , indicating that Mest1 does not facilitate these processes in nutrient rich conditions . Germination , germ tube formation and appressorial formation by Mr2575 on intact grasshopper and Manduca cuticles were similar ( Figure 4 ) . The germination rate of ΔMest1 was significantly ( P<0 . 01 ) lower than that of wild type M . robertsii in 0 . 01% YE with or without Manduca cuticular lipids , and 24 h post-inoculation only 2 . 7±1 . 5 and 26 . 5±2 . 4% of ΔMest1 germlings had appressoria in YE and YE+lipids , respectively , as compared to 25 . 5±1 . 5 ( YE ) and 81 . 6±3 . 2% ( YE+lipids ) of wild type Mr2575 and 24 . 7±1 . 2% ( YE ) and 80 . 7±3 . 5% ( YE+lipids ) of complemented strain Mest1-Com ( Table 2 ) . Furthermore , ΔMest1 appressoria ( 3 . 6±0 . 1 µm×3 . 9±0 . 8 µm ) were significantly smaller ( P<0 . 01 ) than those of the wild type ( 4 . 1±0 . 2 µm×12 . 7±0 . 4 µm ) on insect cuticles ( Figure 4A and 4B ) . Conversely , overexpression of Mest1 in Mr2575-gpd::Mest1 resulted in multiple lobed appressoria on branched hyphae ( Figure 4C ) . Wild type M . acridum Ma324 forms appressoria in locust cuticle lipid extracts and on locust wings [16] . We expressed a 2 . 7 kb clone encoding M . robertsii Mest1 under native control in Ma324 and studied its impact on pathogenicity and growth on caterpillar cuticle lipid extracts . Compared to the wild type Ma324 , Ma324-Mest1 conidia exhibited significantly faster germination in Manduca cuticle lipids with or without YE ( P<0 . 01 ) . Approximately 7% of Ma324-Mest1 germlings produced appressoria in Manduca lipids compared to 35% of germlings in lipids+YE ( Table 3 ) . In contrast , wild type Ma324 did not form appressoria in either YE or Manduca lipid extracts , and <10% of germlings formed appressoria in Manduca lipid extracts+YE ( Table 3 ) . These appressoria were atypically small ( 2 . 8±0 . 2 µm×3 . 6±0 . 7 µm ) , whereas Ma324-Mest1 formed compound appressoria ( 3 . 9±0 . 3 µm×10 . 2±0 . 6 µm ) at the end of germ tubes similar to those produced by Mr2575 ( Figure 4D ) . Neither wild type Ma324 nor Ma324-Mest1 formed appressoria in water or YE , even though there was hyphal growth in the latter ( Table 3 ) . Thus heterologous expression of Mest1 by M . acridum is not sufficient by itself to trigger differentiation of appressoria . Unlike wild type Ma324 , Ma324-Mest1 produced appressoria within 24 hours of inoculation onto G . mellonella and M . sexta cuticles ( Figure 4F and 4G ) . Pathogenicity assays showed that Ma324-Mest1 kills M . sexta and G . mellonella larvae , even though these are very poor hosts for the wild type at the spore dose tested ( Figure 5A and 5B ) . The LT50 ( time required to kill 50% ) values of M . robertsii Mr2575 and Ma324-Mest1 against M . sexta were similar , being 5 . 1 days and 6 . 2 days , respectively . Inoculation of caterpillars with Ma324-Mest1 caused localized melanization ( indicating cuticle penetration ) and sluggishness similar to Mr2575 , whereas caterpillars developed no symptoms with the wild type Ma324 8 days post-inoculation . Ma324-Mest1 was able to complete the full pathogenic life cycle on caterpillars as cadavers quickly became covered in spores . In contrast , wild type Ma324 only produced spores on the cadavers of a preferred acridid host M . rubrum ( Figure S1 ) . The LT50 values of Ma324 ( 4 . 9±0 . 2 d ) and Ma324-Mest1 ( 4 . 7±0 . 6 d ) against M . rubrum were not significantly different ( Figure 5C ) , suggesting that the unknown esterases/lipases used by Ma324 for mobilizing internal nutrients are as efficient as MEST1 , but only expressed on its specific hosts . Transcript levels of Mest1 gene and two reference genes gpd ( glyceraldehyde 3-phosphate dehydrogenase ) and tef ( translation elongation factor 1-α ) were measured using SYBR dye technology ( Applied Biosystems , CA ) and quantitative real-time PCR ( qRT-PCR ) analysis . Real-time PCR analysis demonstrated stronger expression of Mest1 by wild type Mr2575 in nutrient-poor media including water , basal medium and 1% Manduca cuticle relative to nutrient-rich media such as insect hemolymph or SDB ( Figure 6A ) . We also analyzed Mest1 expression in time course studies . Mest1 was activated within 2 h of conidia being incubated in H2O but activation took up to 4 h when cultured in 1% ground Manduca cuticle medium ( Figure 6B and 6C ) . Catabolite repression is a common mechanism by which easily available carbon sources decrease the expression of enzymes required for the use of other more complex nutrients such as lipids [23]–[25] . Real-time PCR analysis demonstrated that Mest1 expression was repressed in Mr2575 when grown on glucose , galactose , sorbose , fructose , trehalose or sucrose as sole carbon sources ( Figure 6D ) . These results suggest that Mest1 expression occurs when Mr2575 needs to access nutrient reserves . However Mest1 expression was increased by 1% alanine , a common component of insect cuticles and not by locust cuticle , suggesting that the availability of easily accessible nutrients is not the only controlling factor for Mest1 expression . To visualize the intracellular targeting of MEST1 in vivo , MEST1 tagged at its C-terminus with the green fluorescent protein ( GFP ) was analyzed by fluorescence microscopy of living Mr2575 cells . We also determined whether lipid droplet localization is a general quality of MEST1 by expressing it in the yeast S . cerevisiae , as these lack an endogenous MEST1-like protein . The GFP signal co-localized with lipid droplets stained with the neutral lipid stain Nile red in Mr2575 and the transformed yeast cells , confirming that MEST1 is binding to lipid droplets ( Figure 7 ) . No additional diffuse cytoplasmic signal was seen with either GFP or Nile Red . The expression patterns and the intracellular localization of MEST1 are therefore consistent with the protein playing a part in mobilizing global triacylglycerol storage by acting at the level of lipid droplets . In spite of possessing endogenous nutrient reserves , germination of M . robertsii requires external nutrients , albeit these can be at very low levels . When conidia are incubated in water they swell but do not germinate [13] . To test the involvement of MEST1 in lipid metabolism , we compared the lipid content of M robertsii Mr2575 , ΔMest1 , Mest1-Com , Mr2575-gpd::Mest1 , M . acridum Ma324 and Ma324-Mest1 . Except for the reduced lipid content of Mr2575-gpd::Mest1 , which constitutively expresses Mest1 , conidia from the tested strains showed no significant differences in their lipid content , indicating that MEST1 is not required for lipid storage ( Figure 8A ) . As expected , total lipid content in all strains fell significantly ( P<0 . 01 ) as nutrient reserves were mobilized during nutrient stress ( conidia incubated in water ) and during germination on 1% alanine . However , germlings of the wild type M . robertsii Mr2575 contained only 44 . 4% of the original lipid present in conidia as compared to 65 . 2% in the mutant ΔMest1 . The complemented strain Mest1-Com had similar lipid content as wild type strain Mr2575 . Ma324-Mest1 contained 50 . 3% of the original lipid as compared to 70% in wild type M . acridum Ma324 , demonstrating that the transgenic Mest1 was hydrolyzing lipids in Ma324-Mest1 ( Figure 8A ) . Triglycerides are degraded into fatty acids and glycerol . Consistent with more rapid hydrolysis of triglycerides , the glycerol content of M . robertsii Mr2575 was 1 . 98-fold higher ( P<0 . 01 ) than in the disruptant ΔMest1-2575 , while the over-expression strain Mr2575-gpd::Mest1 had 1 . 35-fold higher glycerol than the wild type strain Mr2575 ( Figure 8B ) , suggesting that MEST1 contributes to the generation of turgor pressure in M . robertsii . The complemented strain had similar glycerol content as WT Mr2575 . Similarly , heterologous expression of M . robertsii MEST1 in M . acridum Ma324 resulted in a 1 . 95-fold increase in glycerol content compared to wild type Ma324 ( Figure 8B ) . To determine if addition of exogenous nutrients overcomes the inability of wild type M . acridium to infect lepidopterans , we inoculated G . mellonella caterpillars with Ma324 spores suspended in 1% nutrient solutions ( SDB , glucose , glycerol , or N-acetylglucosamine ) , or we topically applied pre-germinated conidia ± exogenous nutrients . Neither exogenous nutrients nor pregermination triggered differentiation of infection structures on insect cuticle , and consequently M . acridum was unable to infect caterpillars ( Table S2 ) . To investigate the substrate specificity of MEST1 , we expressed Mest1 in E . coli Rosetta ( DE3 ) cells . SDS-PAGE and western blot analysis confirmed that a novel 47 kDa band in the transformed E . coli ( ED3 ) cell lysates was MEST1- ( His ) 6 fusion ( Figure 2D and Figure S2 ) . Attempts to purify six-His-tagged MEST1 expressed in E . coli Rosetta ( DE3 ) cells failed , which could be because the six-His tag is inaccessible in this protein . Therefore , esterase activity was measured in crude extracts , with similar extracts from E . coli Rosetta ( DE3 ) transformed with the corresponding empty vector used as control . The substrate specificity of the expressed MEST1 was determined against p-nitrophenyl esters with different carbon chain-lengths ( Figure 9 ) . MEST1 exhibited a marked preference for short-chain fatty acids , with highest activity against p-NP propionate ( C3 ) , p-NP butyrate ( C4 ) and p-NP caproate ( C6 ) . As is typical for esterases [26] , the enzyme was much less active against p-NP esters with longer-chain lengths .
The genus Metarhizium provides a novel model system to study evolutionary processes as it includes species such as M . robertsii with broad host ranges , as well as M . majus , M . flavoviride and M . acridum that are specific for scarabs , hemipterans and acridids , respectively [5] . M . acridum Ma324 in particular has been widely used for locust control . The genetic distinctness of M . acridum from generalist strains implies evolutionarily conserved host use patterns . However , being a generalist does not preclude M . robertsii strains from showing adaptations to nutrients on frequently encountered hosts . For example , nutrients on Hemiptera ( e . g . , aphids ) are supplemented by insect secretions rich in sugars while beetles carry low levels of nitrogenous nutrients . Consistent with this , many hemipteran-derived lines produce appressoria in glucose medium whereas coleopteran-derived lines do not [27] . Closely related strains show these differences indicating that there are genetic mechanisms allowing rapid adaptation [28] . Lipids are the main nutrient reserve in fungal spores [16] , [25] . In addition , lipid bodies are transported to the developing appressoria and degraded to release glycerol , which contributes to the hydrostatic pressure that provides a driving force for mechanical penetration [15] . This process is controlled by a perilipin that surrounds the lipid droplets and is predominantly expressed when M . robertsii is engaged in accumulating lipids in nutrient rich conditions . The perilipin layer is broken down during nutrient deprivation which presumably allows esterases/lipases to hydrolyze the lipid [15] . In this paper we show that unlike perilipin , Mest1 is predominantly expressed when lipids are being broken down . Disrupting Mest1 ( ΔMest1 ) did not interfere with saprophytic growth of Mr2575 in nutrient rich conditions showing that its function is only important when it is adaptive for the fungus to mobilize endogenous nutrient reserves . Compared to ΔMest1 , wild type Mr2575 grown in nutrient poor media or with cuticular lipids germinated faster and the germlings contained less lipid and more glycerol , while overexpression of Mest1 further boosted germination as well as appressorial differentiation . Intracellular lipid content was also significantly reduced and glycerol content increased in Ma324-Mest1 compared to wild type Ma324 . These findings indicate that MEST1 is required by M . robertsii for rapid hydrolysis of endogenous energy reserves during germination and for infection structure formation . Lipid droplet enzymes have not been studied in filamentous fungi , but in yeasts all lipid droplet associated proteins are involved in the mobilization of triacylglycerols under conditions of fatty acid starvation [29] . The nutrient poor conditions that induce expression of Mest1 are a necessary trigger for virulence as M . robertsii strain Mr2575 only produces infection structures on cuticle surfaces with low levels of nutrients [13] . However , in ΔMest1 the mobilization of stored lipids was delayed , but not abolished , suggesting the existence of additional enzymes involved in breakdown of lipid droplets . It seems clear from culturing the Ma324-Mest1 in yeast extract medium , that MEST1 expression is not sufficient by itself to induce appressoria production . Consequently , production of appressoria by Ma324-Mest1 on M . sexta and G . mellonella cuticles suggests that these have at least some of the inducers required by Ma324 for appressorial differentiation ( Figure 2F and 2H ) . The incidence of infection and the severity of the disease symptoms shown by Ma324-Mest1 were similar to those observed with M . robertsii Mr2575 , an authentic pathogen of caterpillars . We have therefore shown that inserting a single gene able to mobilize nutrient reserves from a broad spectrum pathogen into a specialist is enough to broaden the latter's host range . This implies that mobilization of nutrient reserves or lack thereof can account for the broad host range of M . robertsii and the restricted host range of M . acridum , respectively . Restricted host range could therefore be due essentially to biochemical limitations . Addition of exogenous nutrients ( SDB , glucose , N-acetylglucosamine , glycerol ) or pregermination on rich medium ( SDB ) did not overcome the inability of M . acridum to infect caterpillars . The implication is that intracellular lipid reserves are not only a nutrient source but also a source of chemical signals triggering infection processes . Consistent with this , disrupting the Mest1 gene in M . robertsii ( ΔMest1 ) reduces virulence against caterpillars but not grasshoppers , indicating that the physiological conditions on grasshoppers and caterpillars are different , and there are different mechanisms for linking signals on the host surfaces with differentiation of infection structures . Hydrocarbons comprise over 90% of the wax layer on the surface of grasshoppers , with the balance being composed of wax esters , free fatty acids and triacylglycerides , whereas in larval Lepidoptera , aliphatic alcohols are the most abundant compounds , and triacylglycerides are absent [30] . M . acridum extensively hydrolyzes surface lipids and waxes during germination and pre-penetration growth on locust cuticles [31] . A possible explanation for the different responses to grasshopper and caterpillars may be that breakdown products of triglycerides on grasshoppers provide signaling molecules to trigger infection processes by Ma324 , but on caterpillars MEST1 provides these signals from stored triglycerides . As the transgenic M . acridum Ma324-Mest1 is able to kill caterpillars , wild type M . acridum must have retained most of the genetic machinery required for parasitism of insects outside its natural host range . This is consistent with the developmental processes within M . anisopliae and M . acridum being very similar , e . g . formation of germ tubes , appressoria , penetration pegs unicellular blastospores , and multi-cellular hyphal bodies that facilitate the infection of target insects , proliferation within haemolymph , and eventual eruption through the host cadaver . Besides novel lineage specific proteins such as MEST1 , host recognition may therefore be determined by regulatory controls that allow expression of pathogenicity genes that are not expressed on non-hosts . Expression of Mest1 under its native M . robertsii promoter in M . acridum may have broadened host range by bypassing its need for esterases that are regulated by specific locust-related stimuli . The impact of a single gene on host range suggests that host shifts may have occurred during Metarhizium speciation by the acquisition or loss of novel pathogenicity factors . The presence of a MEST1 ortholog in the related N . haematococca is consistent with M . acridum and M . robertsii having inherited MEST1 from a common ancestor . Their patchy distribution could be explained by rapid mutation or multiple lineage specific gene loss events in M . acridum and other fungi . An important question is whether utilizing MEST1-like esterases during infection processes is the ancestral state . There are 16 esterases in M . robertsii ( 6 of which are secreted ) and 18 in M . acridum ( 6 secreted ) [22] , and Metarhizium spp . express multiple esterases on insect cuticles [32] . It seems likely that Ma324 uses the same esterases for mobilizing lipids as other fungi , and the Mest1 gene in M . robertsii has acquired unique functions in this species as it is clearly highly dispensable . The new functions could potentially have turned the recipient into a novel pathogen or allowed it to infect new hosts . Expression of Mest1 in Ma324 does not change virulence against its preferred grasshopper host suggesting that acridids do not represent a specialized ecological niche in which a M . robertsii-like MEST1 activity is detrimental , and there is no evidence in this study for any conditional benefits in losing MEST1's function . This is consistent with the infection-related functions of MEST1 arising de novo in M . robertsii . This study has important safety implications for field applications of M . acridum as it shows that Ma324 lacks a gene important for opportunism and this should severely constrain the possibility of host switching to non-target beneficials . In addition , an understanding of how Mest1 affects fungal responses to hosts , and identification of the signaling cascades involved in regulating the mobilization of nutrient reserves will provide fundamental new insights into the initial steps that are required for the establishment of a compatible interaction between fungi and their hosts .
M . robertsii Mr2575 and M . acridum Ma324 are wild type strains that were obtained from the U . S . Department of Agriculture Entomopathogenic Fungus Collection ( ARSEF ) in Ithaca , N . Y . Strain Mr2575 can infect caterpillars ( Manduca sexta ) , beetles ( Cucurlio caryae ) [33] , grasshoppers ( Melanoplus femurrubrum ) and locusts ( Schistocerca gregaria ) [34] . M . acridum Ma324 ( = CSIRO FI 485 ) is the active ingredient of “Green Guard” used for locust control in Australia . In the field it is found exclusively in acridids , and is only infectious to caterpillars in laboratory conditions at very high spore concentrations . Fungal strains were maintained on Potato Dextrose Agar ( PDA ) at 27°C . Conidia were obtained from 10 day old PDA cultures . For preparation of genomic DNA and RNA , fungal spores were cultured in Sabouraud dextrose broth ( SDB ) ( 2×106 conidia/ml ) at 27°C . Mest1 was originally identified as an EST expressed when strain Mr2575 was grown on insect cuticle [17] . The full-length sequence of the Mest1 cDNA was obtained from the EST using RACE , and a genomic clone was obtained using the DNA Walking Speed Up Kit II ( Seegene Inc . , Rockville , Maryland , USA ) . The primers are listed in Supplementary Information , Table S1 . For targeted deletion of Mest1 , the 5′ and 3′ flanking regions of the Mest1 ORF were amplified by PCR from Mr2575 genomic DNA , and then subcloned into the XbaI and SpeI sites of the binary vector pBarGFP [35] . The gene disruption construct ( pBarGFP-Mest1 ) was then transformed into Agrobacterium tumefaciens AGL-1 for targeted gene disruption by homologous recombination as described previously [36] . Replacement-specific PCR amplifications of the Mest1 locus were performed with specific primer pairs ( primers are listed in Table S1 ) that amplify either the wild type or the mutant gene locus . To revert disruptant ΔMest1 , the full-length Mest1 gene with its native promoter and terminator sequences was amplified by PCR and cloned into the XbaI site of pBenGFP [36] to generate complementation vector pBenGFP-Mest1 . The complementary strain Mest1-Com was generated by reintroduction of pBenGFP-Mest1 into the disruptant ΔMest1 using Agrobacterium tumefaciens-mediated transformation . M . robertsii MEST1 was heterologously expressed in specialist M . acridum Ma324 using vector pBenGFP-Mest1 to generate Ma324-Mest1 . Four putative transformants were chosen and verified for Mest1 gene expression by RT-PCR ( data now shown ) . A transformation vector was constructed by amplifying the coding region of Mest1 from the cDNA clone with primer pairs Mest1F and Mest1R ( Table S1 ) containing a 5′ BamHI site and a 3′ EcoRI site plus a 6×His-tag . The resulting PCR fragment was cloned into the pGEM-T/A cloning vector ( Promega ) , and the sequence of the Mest1 amplicon was confirmed by sequencing . The Mest1 gene fragment was released with BamH I and SmaI , and subcloned into pBarGPE1 [37] downstream of a constitutive Aspergillus nidulans gpdA promoter to obtain pBarGPE1-Mest1 . The Mest1 cassette was released by cleavage with BglII , and then inserted into the BglII site of pBarGFP to generate pBarGFP-gpdA::Mest1 for A . tumefaciens-mediated transformation into Mr2575 . The resulting strain was designated as Mr2575-gpd::Mest1 . Genomic DNA was extracted from putative transformants for Southern blot analysis as previously described [18] . Mest1 gene expression in transformants was verified by RT-PCR [38] . To determine the subcellular localization of MEST1 , the promoter region ( ∼1 . 5 kb ) together with the Mest1 ORF minus the 3′ TAA stop codon was amplified with primer pairs Mest1F2 and Mest1R2 ( Table S1 ) and inserted into the BamHI and EcoRI sites of pBarGPE1 [37] to generate pBarGPE1-Pmest1::Mest1 . An enhanced Green Fluorescent gene ( eGFP ) was amplified from pEGFP ( Clontech ) with primers gfpF1 and gfpR1 containing a 5′ EcoRI site and a 3′ XhoI site , and integrated into the EcoRI and XhoI sites of the plasmid pBarGPE1-Pmest1::Mest1 to generate pBarGPE1-Pmest1::Mest1:GFP . The construct was restricted with Pml1 and BamHI , and the released cassette Pmest1::Mest1:GFP was subcloned into the BamHI and EcoRV sites of the binary vector pPK2 [39] to generate pBar-Pmest1::Mest1:GFP . The final construct was transformed into wild type Mr2575 using A . tumefaciens AGL-1 to generate Mr2575-Mest1:GFP . We also determined whether MEST1 localizes to lipid droplets in Saccharomyces cerevisiae , a fungus unrelated to Metarhizium and lacking an endogenous MEST1 protein . The Mest1 ORF was amplified with primers Mest1yesF and Mest1yesR ( Table S1 ) and the product integrated into the EcoRI and NotI sites of pYES2 ( Invitrogen ) . The resulting pYES2-Mest1 or the parent plasmid pYES2 ( used as a control ) were transformed into S . cerevisiae strain INVSc1 according to the manufacturer's instructions ( Invitrogen ) . Nile red , ( 9-diethylamino-5H-benzo [alpha] phenoxazine-5-one ) was used to stain intracellular lipid droplets , which were viewed by fluorescence microscopy as previously described [15] . To monitor Mest1 expression in different growth conditions , fungal spores were incubated ( 6 hrs ) in 10 ml of fresh SDB , Manduca sexta hemolymph [17] , basal medium [BM , 0 . 02% KH2PO4 , 0 . 01% MgSO4 ( pH 6 ) ] , water or water supplemented with either 0 . 1% bean root exudate [40] ) or 1% insect cuticle as described [38] . Fungal cells were also incubated in BM supplemented with 1% glucose , 1% galactose , 1% sorbose , 1% trehalose , 1% sucrose or 1% alanine . Total RNA was extracted using RNeasy Plant Mini Kit ( Qiagen ) . First strand cDNA was synthesized using Verso cDNA Kit ( ABgene ) according to manufacturer's instructions . Real-time quantitative reverse transcription PCR ( qRT-PCR ) reactions were performed using a Quantitative real-time SYBR Green MasterMix Kit ( Applied Biosystems ) on an Applied Biosystems 7300 real-time instrument and ABI Prism SDS 1 . 2 . 2 . software . The qPCRs were performed using the following conditions: 50°C for 2 min , then denaturation at 95°C for 10 min followed by 40 cycles of denaturation at 95°C for 20 s , annealing and extension at 60°C for 1 min . The primers used for gene Mest1 and the reference genes gpd and tef are listed in Table S1 . To test the involvement of MEST1 in lipid metabolism , the cell lipid content was quantified by the sulfo-phospho-vanillin method as previously described [15] . A reference standard curve was generated using triolein ( Sigma ) . To determine whether starvation stress induces the hydrolysis of residual stored lipids , conidia from wild type M . robertsii Mr2575 , mutant ΔMest1 , complementary strain Mest1-Com , over-expression strain Mr2575-gpd::Mest1 , wild type M . acridum Ma324 and transgenic Ma324-Mest1were incubated in H2O for 36 h , which causes spores to swell but not germinate , or induced to germinate in basal medium plus 1% alanine ( wt/vol ) for up to 12 h . The total lipid content of the conidia was assayed as described above . Fungal conidia were inoculated into BM supplemented with 0 . 25 mg/ml of Manduca cuticle lipids [15] for up to 48 h . Cultures were harvested by filtration through Whatman No . 1 filter paper and 0 . 2 µm Millipore filter units , washed quickly with ice-cold BM , and resuspended in 0 . 5 M Tris-HCl , pH 7 . 5 . The samples were heated to 95°C for 10 min and cell debris pelleted by centrifugation as described [41] . The glycerol concentration was assayed enzymatically using a glycerol determination kit according to the manufacturer's instructions ( Sigma ) . The data represent the average of three independent experiments . The germination rate of conidia was measured by inoculating 20 µl of spore suspension ( 2×107 spores ml−1 ) in 5 . 5 cm polystyrene Petri dishes containing either 2 ml of water , 0 . 01% yeast exact ( YE ) and/or insect cuticle lipids ( 0 . 25 mg/ml ) . Three hundred spores from each of three replicates were recorded microscopically to assess germination and appressorial differentiation against the hydrophobic surface of the Petri dish . Appressoria were also induced against locust ( Schistocerca gregaria ) hind wings , Galleria mellonella and Manduca sexta cuticle as described previously [16] . For prokaryotic expression of MEST1 , the full length cDNA of Mest1 was cloned by RT-PCR using specific primers Mest1EexF and Mest1EexR ( Table S1 ) , restricted with EcoRI and NotI , and subcloned into the prokaryotic expression vector pET28a at the EcoRI and NotI sites to form pET28a-Mest1 . E . coli Rosetta ( DE3 ) cells ( Novagen ) were transformed with the recombinant expression vector . His–tagged MEST1 production was induced by 0 . 2 mM IPTG in LB medium for 16 h at 28°C , and cells were lysed with the B-PER Bacterial Protein Extraction Reagent ( Thermo Scientific ) according to the manufactor's instructions . The His-tagged MEST1 was detected by Western blot analysis using rabbit Anti 6-Histidine Epitope Tag monoclonal antibody and anti-rabbit IgG ( Fc ) , conjugated to alkaline phosphatase ( Invitrogen ) . Ester hydrolase activities were determined in E . coli Rosetta ( DE3 ) transformed with pET28a-Mest1 or the empty vector pET28a . Cell extract ( 0 . 1 ml ) containing ∼50 µg of protein and 0 . 1 ml of a p-NP-derivative substrate in 50 mM potassium phosphate buffer ( pH 8 ) was incubated at 37°C . Enzyme activity was spectrophotometrically determined by measuring the liberation of p-nitrophenol as previously described [42] . One unit of activity was defined as the amount of enzyme that released 1 µmol of p-NP per min per mg of protein under the assay conditions . Substrate specificity towards various p-nitrophenyl esters ( Sigma-Aldrich ) was determined using p-NP propionate ( C3 ) , p-NP butyrate ( C4 ) , p-NP caproate ( C6 ) , p-NP laurate ( C12 ) , p-NP myristate ( C14 ) , p-NP palmitate ( C16 ) , and p-NP stearate ( C18 ) as substrates . The virulence of the wild type M . robertsii Mr2575 , mutant ΔMest1 , complementary strain Mest1-Com , over-expression strain Mr2575-gpd::Mest1 , wild type M . acridum Ma324 and transgenic strain Ma324-Mest1 were assayed against wild caught 5th instar Melanoplus femurrubrum grasshoppers ( College Park , Maryland ) and newly molted fifth-instar larvae of Manduca sexta and Galleria mellonella ( Carolina Biological supplies ) . An aliquot of 3 µl of fungal spore suspension was applied to the pronotum of each grasshopper as previously described [43] ) . Grasshoppers were then placed in clear plastic boxes at 28°C under an 18∶6-h photoperiod in humid conditions ( >80% ) , and supplied daily with fresh wheat seedlings . Each box contained 10 grasshoppers; three containers were used for each dosage of fungus tested ( 3×105 or 5×105 spores/insect ) . Manduca sexta and Galleria mellonella were inoculated by topical immersion in conidial suspensions ( 1×107 conidia/ml ) as previously described [38] . Mortality was recorded every 12 h . After death , cadavers were surface sterilized [35] , and incubated in Petri dishes with a sterile wet cotton ball to promote fungal emergence , and thus confirm cause of death . Each treatment was replicated three times with 30 insects per replicate , and the bioassays were repeated twice . LT50 values were calculated with the SPSS program [44] . Sequence data reported here have been deposited in the GenBank database under the following accession numbers: Metarhizium robertsii Mest1 mRNA ( HM747114 ) , Mest1 genomic DNA ( HM747115 ) .
|
Host selectivity and host switching have been widely documented in diverse pathogens , but in most cases the underlying mechanisms are poorly understood . Entomopathogenic fungi in the genus Metarhizium are being used as environmentally friendly alternatives to chemical insecticides in agricultural and human disease-vector control programs , and as model systems for studying the interactions between invertebrate hosts and pathogenic fungi . In this paper we describe molecular mechanisms controlling the host selectivity of M . robertsii strain Mr2575 , a generalist able to infect hundreds of insect species , and M . acridum strain Ma324 , a specialist pathogen of grasshoppers and locusts . The esterase gene ( Mest1 ) from Mr2575 is required for virulence against caterpillars but not grasshoppers . Ma324 lacks Mest1 , but insertion of Mest1 into Ma324 broadens its host range to include caterpillars . Our results suggest that expression of Mest1 allows rapid mobilization of endogenous lipid reserves and promotes germination and infection structure formation . This study suggests that speciation in insect pathogens can be driven by host shifts due to the gain or loss of a pathogen gene that confers wide host specificity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"infectious",
"diseases",
"mycology",
"pest",
"control",
"fungal",
"evolution",
"microbial",
"control",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"infectious",
"disease",
"control",
"agriculture",
"pathogenesis"
] |
2011
|
Insertion of an Esterase Gene into a Specific Locust Pathogen (Metarhizium acridum) Enables It to Infect Caterpillars
|
Aedes albopictus , the Asian tiger mosquito , is a vector of several arboviruses including dengue and chikungunya , and is also a significant nuisance mosquito . It is one of the most invasive of mosquitoes with a relentlessly increasing geographic distribution . Conventional control methods have so far failed to control Ae . albopictus adequately . Novel genetics-based strategies offer a promising alternative or aid towards efficient control of this mosquito . We describe here the isolation , characterisation and use of the Ae . albopictus Actin-4 gene to drive a dominant lethal gene in the indirect flight muscles of Ae . albopictus , thus inducing a conditional female-specific late-acting flightless phenotype . We also show that in this context , the Actin-4 regulatory regions from both Ae . albopictus and Ae . aegypti can be used to provide conditional female-specific flightlessness in either species . With the disease-transmitting females incapacitated , the female flightless phenotype encompasses a genetic sexing mechanism and would be suitable for controlling Ae . albopictus using a male-only release approach as part of an integrated pest management strategy .
The Asian tiger mosquito , Aedes albopictus ( Skuse ) , is a vector of several arboviruses including dengue and chikungunya . This highly invasive species originating from Southeast Asia has travelled the world in the last forty years and is now established in Europe , North and South America , Africa , the Middle East and the Caribbean . In the absence of a vaccine or antiviral drugs , efficient mosquito control strategies are crucial . Novel control methods are being developed that involve the use of genetically modified mosquitoes to either suppress the target population or replace it with a pathogen-resistant strain [1] , [2] , [3] , [4] , [5] , [6] . The sterile insect technique ( SIT ) is a pest population control method developed in the 1950s which relies on releasing large numbers of males sterilised by irradiation to compete for mates with the wild-type , consequently reducing the proportion of viable offspring [7] , [8] . Despite large-scale success against some agricultural pest insects , and some promising successes of this technique against mosquitoes in the 1970s , application of the SIT to mosquito control consistently suffered from the lack of an efficient sexing system in order to eliminate disease-transmitting females before releasing the sterile males [9] . In addition , the irradiation process can impose a significant fitness cost on mosquito species [10] . Furthermore , mathematical modelling shows that the early embryonic lethality caused by the irradiation of paternal sperm is sub-optimal as it reduces the number of immature mosquitoes competing for resources during the density-dependent larval stages [11] , [12] , [13] . The RIDL system [14] is a variant of SIT which replaces irradiation by genetically engineered inducible sterilisation , an approach offering more flexibility with regards to the time of death , the sex and even the tissues targeted by the sterilising mechanism . Initial estimates suggest that this approach may provide an attractive alternative or complement to additional control methods [3] , [13] , [15] . In the RIDL systems so far developed , sterility is induced by conditional zygotic expression of a dominant lethal gene: the tetracycline-repressed transactivator tTA [16] is placed under the control of a suitable promoter , while a lethal gene is placed under the control of the tTA response element tetO . In the absence of tetracycline , the tTA transactivator binds to tetO and , via activation of a suitable minimal promoter , induces expression of the dominant lethal gene . Tetracycline prevents tTA from binding to the tetO sites , thereby repressing the system and allowing RIDL insects to develop normally on a diet supplemented with tetracycline . However , in the wild , progeny of released RIDL insects will express the lethal gene and consequently die . The development of a late-acting RIDL strain of Ae . aegypti was reported in 2007 [11] . For Ae . aegypti , efficient physical sex-separation systems based on pupal size are available; this has allowed male-only release of this strain in successful field trials [17] , [18] . A genetics-based alternative would eliminate this labour-intensive step , and is the only option for the wide range of insects , including most mosquitoes , for which reliable physical sex-separation methods are not available . Female-specific ( fsRIDL ) has additional potential advantages in terms of resistance management that could be highly advantageous in the context of an integrated pest management programme [19] . Genetic sexing RIDL strains of the Mediterranean fruit fly , Ceratitis capitata , have been produced based on the sex-specific splicing properties of the tra gene [20] , but no tra homologue has yet been found in mosquitoes . The Ae . aegypti Actin-4 gene ( AeAct-4 ) is specific to the indirect flight muscles of females with expression starting in L4 larvae [21]: an ideal combination of both female-specificity and late-acting expression allowing production of genetic sexing RIDL strains with a post-larval lethality . Fu et al . recently reported the development of a RIDL strain exploiting the promoter and sex-specific alternative splicing of the AeAct-4 gene and exhibiting a repressible female flightless phenotype [22] . Inability to fly incapacitates females at almost the latest possible stage in development prior to biting . Wise de Valdez et al . showed that periodic release of this strain could eliminate cage populations of Ae . aegypti [23] . This phenotype is indirectly lethal to females; ability to fly is essential in the field to access sugar resources and escape predation . Flight ability is also needed in both lab and field for mating , as well as – in the field – to acquire a blood meal , so flightless mosquitoes are functionally sterile . A female-flightless phenotype would therefore also permit the release of eggs directly into artificial or pre-existing breeding sites , from which RIDL males would emerge to seek wild females . With a view to applying the same type of genetic control to populations of Ae . albopictus as that proposed by Fu et al . [22] , we have isolated and characterised a segment of the Ae . albopictus Actin-4 ( AealbAct-4 ) gene . AealbAct-4 showed functional and sequence similarities to its Ae . aegypti homologue: RIDL strains of both Ae . albopictus and Ae . aegypti carrying a construct based on the AealbAct-4 promoter and 5′UTR displayed a repressible female flightless phenotype , as did an Ae . albopictus strain carrying an AeAct-4-based construct . These results indicate that the Actin-4 promoters from Ae . aegypti and Ae . albopictus are substantially interchangeable for transgenic-based RIDL strategies in these species .
The Ae . albopictus and Ae . aegypti wild-type strains originated from Malaysia and were colonised by the Institute of Medical Research ( Kuala Lumpur ) in 2006 and 1977 , respectively . The insectary was kept at 27°C ( ±1°C ) and 70% ( ±10% ) relative humidity . Larvae were fed on crushed dry fish food ( TetraMin flake food from Tetra GmbH , Germany ) and adults on 10% glucose supplemented with 14 µg/ml penicillin and 14 µg/ml streptomycin . Females were fed on horse blood using a Hemotek Insect Feeding System ( Discovery Workshops , Accrington , UK ) set at 37°C . Pre-blastoderm embryos were prepared for injection and micro-injected as described [24] . Injection mixtures consisted of 300 or 350 ng/µl of donor plasmid ( OX3688 and OX4358 , respectively ) , 300 ng/µl of piggyBac mRNA [24] and 30 µg/ml of chlortetracycline in injection buffer ( 5 mM KCl and 0 . 1 mM NaH2PO4 , pH 6 . 8 ) . phsp-pBac helper plasmid [25] was also included in the OX3688 mixture to a final concentration of 200 ng/µl as previously described [24] . Injected G0 adults were crossed in pools ( males in pools of 2 for 24 hours then merged in pools of 24; females in pools of 100 ) to wild-type counterparts . G1 larvae were screened for fluorescence using a Leica MZ95 microscope with the appropriate filter sets from Chroma Technology ( Rockingham , VT ) ( filters: AmCyan: exciter D436/20×; emitter D480/40 m; DsRed2: exciter HQ545/30×; emitter HQ620/60 m ) . Transgenic lines were established from single G1 positive adults . Lines named with different letters have founders from different G0 pools and are therefore independent genomic integrations . Lines derived from the same G0 pool were characterised and flanking sequences used to confirm their independence from each other ( data not shown ) . Only lines showing a 1∶1 fluorescent to wild-type ratio in the progeny of heterozygous to wild-type crosses , consistent with single transgene insertion , were kept for phenotype analysis ( data not shown ) . Pictures of fluorescent larvae were taken with a Canon PowerShot S5IS camera with an MM99 adaptor ( Martin microscopes ) to fit into the eyepiece . Ae . aegypti Actin-4 ( AeAct-4 , AY531222 ) , Ae . aegypti Actin-3 ( AeAct-3 , AY289765 ) and Anopheles gambiae Actin-1 ( AnAct-1 , XM315270 , which we considered from sequence analysis likely to be the An . gambiae homologue of Ae . aegypti Actin-4 ) sequences were aligned using ClustalW ( EBI ) . Primers AeA4F1 and AeA4R2 were designed in regions which were conserved between AeAct-4 and AnAct-1 but differed from AeAct-3 , and used to amplify Ae . albopictus wild-type genomic DNA . The resulting PCR product was cloned and sequenced . BLAST alignment confirmed strong sequence similarity to Ae . aegypti Actin-4 . This sequence was extended by a combination of 5′RACE and PCR techniques . 5′RACE was performed using the Ambion FirstChoice RLM-RACE kit according to the manufacturer's instructions , using primers AlbA4Race and AlbA4RaceN on 7 µg total RNA extracted from 2 pooled female pupae; adaptor-mediated PCR on genomic DNA was used to extend the sequence from the beginning of the 5′UTR back into the promoter region and from the exon 1 and 2 sequences to obtain the intron sequence . Comparison of cDNA and gDNA sequences revealed a large intron in the 5′UTR . 745 bp upstream from the start of the 5′UTR , a coding sequence with BLAST homology to Ae . aegypti sensory neuron membrane protein 2 was found , delimiting the maximum promoter fragment unless the genes overlap . The OX3688 construct is identical to the OX3604 plasmid ( [22] , JN936856 ) , apart from a correction: the 3×P3-DsRed marker cassette at one end of OX3604 was subsequently found to be 3×P3-AmCyan instead . This was corrected by exchanging a PacI-SpeI cassette to make OX3688 , which therefore represents the structure originally intended for OX3604 . Note that OX3604 encodes a tTA-like protein , tTAV [11]; relative to plasmid OX513 ( formerly LA513 , [11] ) the tTAV coding region in OX3604 has altered nucleotide sequence which we now refer to as tTAV2 . The complete sequence of the OX3604 transposon has been deposited in GenBank with accession number JN936856 . OX4358 construction: A start codon was engineered in the AealbAct-4 gene's 5′ UTR 43 bp before the 5′ donor site of the intron by PCR . Two PCR products , promoter-intron and intron-truncated exon 2 , were amplified from wild-type Ae . albopictus genomic DNA using primer pairs AlbA4proAscF-AlbA4intSpeR and AlbA4intSpeF-AlbA4ex2BglR . The two PCR products were ligated at the Spel site; the ligated product was cloned in front of the fusion gene ubiquitin–tTAV2–K10 3′UTR previously constructed [22] . The engineered start codon was in frame with the tTAV2 coding sequence . This gene cassette was inserted into an existing piggyBac construct , containing the Hr5-IE1 enhancer-promoter from the baculovirus Autographa californica MNPV [26] driving AmCyan ( Clontech ) . In order to study the endogenous Actin-4 gene from Ae . albopictus , RNA was extracted from pooled samples of three wild-type male pupae and two wild-type female pupae , using Tri Reagent ( Ambion ) , according to the manufacturer's instructions . RNA samples were treated with DNAse I ( Roche ) and quantified on a Pharmacia Biotech GeneQuant II RNA/DNA calculator . One-step RT-PCR was carried out on 200 ng RNA using SuperScript III One-step RT-PCR System with Platinum Taq DNA Polymerase ( Invitrogen ) and primers in the 5′UTR ( AlbA4UTRF ) and in exon 2 ( AlbA4FlR ) ( see Table S1 ) , according to the kit protocol . PCR conditions were 50°C for 30 min , 94°C for 2 min followed by 40 cycles of 94°C for 15 s , 55°C for 30 s and 68°C for 1 . 5 min , with a final elongation at 68°C for 5 min . RT-PCR was carried out on male and female pupae of OX4358 and OX3688 individuals as above , using primers AlbA4BsmF and UbiR2 for OX4358 , and Aeact4-ex1 , Aeact4-ex1′ and Diag2-ubi for OX3688 ( Table S1 ) , and the same PCR conditions , to confirm that sex-specific splicing was occurring as predicted in this context . Amplified fragments were verified further by sequencing ( GATC Biotech , Konstanz , Germany ) following gel extraction and purification using the Qiaquick gel extraction kit ( Qiagen ) , according to manufacturer's instructions . For phenotypic analysis of the transgenic lines , eggs were hatched on day 1 . On day 2 , “on tet” and “off tet” trays ( 11×19 cm bottom surface ) were set up with 300 heterozygous larvae in 300 ml of pure water ( 1 larva/ml ) , respectively with or without a supplement of chlortetracycline hydrochloride ( Sigma-Aldrich , Gillingham , UK ) to a final concentration of 30 µg/ml . Water was changed on days 6 and 11 . Larvae were fed crushed dry fish food ( TetraMin flake food from Tetra GmbH , Melle , Germany ) : 12 mg/tray on days 2 , 3 and 17; 24 mg/tray on days 4 , 9 , 11 and 12; 48 mg/tray on day 5; 96 mg/tray on days 6 , 7 and 8 . Sexes were separated as pupae and placed in cages into 5×5×5 cm weighing boats . Emerged adults were separated each day and their flying ability evaluated the following day by aspirating out flying individuals while tapping the cage to stimulate immobile adults . Flying adults were recorded as a proportion of pupae placed in the cage as some death - due to incomplete eclosion or drowning soon after eclosion - occurred before flying ability could be established . The proportion of pupae producing flying adults was therefore used as the measure of fitness in the present studies . For reference , the wild-type strain reared off tetracycline has an eclosion rate ( ± SE ) of 93 . 83% ( ±0 . 98% ) for males and 91 . 17% ( ±1 . 82% ) for females .
The Ae . albopictus Actin-4 gene ( AealbAct-4 ) was isolated as described in Materials and Methods . The gDNA and cDNA sequences have been deposited in GenBank ( Accession numbers: JN709493 and JN709492 , respectively ) . The sequence showed high conservation with AeAct-4 ( and also to AgAct-1 , not shown ) , particularly in the coding sequence ( Figure S1 ) . The positions of the introns are conserved , as is the gene structure with respect to sex-specific splicing ( Figure 1A ) . The sex-specific splicing was confirmed by RT-PCR ( Figure 1B ) . The AealbAct-4 promoter and 5′UTR ( containing the alternatively spliced region ) were used to make construct OX4358 ( Figure 2A ) . OX4358 includes a Hr5IE1-AmCyan-SV40 marker gene leading to strong expression of the AmCyan fluorescent protein all over the body at every developmental stage and allowing simple and reliable screening of the transgenics ( Figure 2B ) . The Ae . albopictus Actin-4 ( AealbAct-4 ) promoter was placed in front of the AealbAct-4 exon 1 , in which a start codon has been engineered . The AealbAct-4 sex-specific intron was shortened by internal deletion but preserving the male-specific transcript which provides multiple stop codons ( Figure 2A , bars below the intron line ) . The AealbAct-4 exon 2 was cloned in frame with tTAV2 , a variant of tTA optimised for expression in insects ( JN936856 ) . Expression of VP16 is activated by the binding of tTAV2; this occurs only where tTAV2 is expressed and in the absence of tetracycline . Transgenic lines carrying OX4358 were obtained for both Ae . albopictus and Ae . aegypti . RT-PCR analysis of OX4358 transgenic individuals confirmed sex-specific splicing in both Ae . albopictus and Ae . aegypti ( Figure 2C and 2D , respectively ) . Sequencing of the PCR products revealed that splicing occurs as in the native gene , except for a second male-specific transcript in which exon 1 has an extra 75 bp . This extra transcript may be a result of truncation of the intron and/or exon 2 , disrupting splicing . However , it still leads to a frame-shift between the start codon and the tTAV2 coding sequence , as with the canonical splice variant , and should not interfere with the intended function of the construct . Ae . aegypti and Ae . albopictus transgenic lines carrying the OX4358 construct were reared on and off tetracycline and their flying/non-flying phenotype assessed . piggyBac–based transgenes insert at any of a very large number of sites , therefore each transgene is embedded in a different chromatin context , which may influence its expression and associated phenotype [27] . A spectrum of phenotypes was therefore anticipated . Of the 20 independent lines obtained in Ae . albopictus , four exhibited a non-repressible flightless females phenotype , giving essentially no flying females when reared on tetracycline; five lines were found to be male-linked; these nine lines were not analysed further . The remaining eleven lines were tested on and off tetracycline ( Table 1 ) ; eight had a repressible female flightless phenotype , with no females flying off tetracycline and between 22 and 55% of females flying on tetracycline . Three lines showed no obvious sex-specific flightless phenotype . Eleven independent lines were obtained in Ae . aegypti , including two male-linked insertions . The other nine lines were tested on and off tetracycline ( Table 1 ) ; four had a repressible female flightless phenotype , with no flying females off tetracycline and 57 to 96% females flying on tetracycline . Four lines showed incomplete penetrance , with 1 to 43% females able to fly off tetracycline . The last line showed no clear flightless phenotype , with 93% females flying off tetracycline while 84% females flew on tetracycline . No impairment was observed in the ability of males to fly when reared off tetracycline . In fact there was generally a slightly higher percentage of flying males when reared off tetracycline compared to rearing on tetracycline ( Table 1 ) . The development of a tetracycline-repressible female flightless phenotype in Ae . aegypti has recently been reported using the AeAct-4 promoter [22] . We transformed Ae . albopictus with a similar construct , OX3688 ( Figure 3A ) . Three transgenic lines were produced . In one of them the flightless phenotype was not repressed by tetracycline at the concentrations used , producing flightless females even when reared on tetracycline . Line OX3688A-Aal showed a repressible female-specific flightless phenotype , with females flying on tetracycline but flightless off tetracycline , and males flying irrespective of tetracycline ( Figure 3B , Video S1 , Video S2 ) . Line OX3688D-Aal did not show female-specific flightlessness off tetracycline ( Figure 3B ) . RT-PCR was performed on OX3688A-Aal male and female pupae , finding sex-specific splicing consistent with the pattern seen in the native gene ( Figure 3C ) . Sequencing of the RT-PCR fragments indicated that the splicing occurred just as in Ae . aegypti ( data not shown ) .
Fu et al . recently reported the engineering of a conditional female flightless phenotype in Ae . aegypti using a DNA segment from the Ae . aegypti Actin-4 gene ( AeAct-4 ) [22] . The results presented in this paper show that this segment retains its key properties in Ae . albopictus and can be used to generate a similar phenotype . Moreover , replacing the AeAct-4 sequence with one from its Ae . albopictus homologue also induced a conditional female flightless phenotype in both Ae . albopictus and Ae . aegypti . This is the first report of an engineered Ae . albopictus phenotype which could be used successfully for population control of the species . The promoter , sex-specific intron and tTAV2 act as independent control elements; logic gates which combine specific inputs ( tissue , sex , tetracycline ) to give predetermined logical outputs ( Figure 4 ) . Such applied synthetic biology of pest insects is in its infancy , but already real-world applications can be seen . Although some sequence variation between the two homologous genes were observed , the two promoters and the sex-specific splicing appear to function similarly in both species . This suggests that they may also behave similarly in other closely related species . The availability of two elements of similar function but different sequence might also be advantageous if female-specific expression of two genes is required simultaneously as it would limit homologous recombinations within the construct . In spite of the general similarity of the different promoters in the two species , Ae . albopictus appeared more affected by the OX4358 RIDL constructs than Ae . aegypti , with relatively low percentages of flying females on tetracycline and males both on and off tetracycline compared to Ae . aegypti lines carrying the same construct . Several potential explanations can be proposed to explain this apparent difference between Ae . aegypti and Ae . albopictus . The Actin-4 promoters may express at a higher level in Ae . albopictus ( or , equivalently , the mRNA or protein might be more stable , or the mRNA translated more efficiently ) , leading to a higher production of tTAV2; Ae . albopictus indirect flight muscles ( IFMs ) may be more sensitive to the over-expression of VP16 than Ae . aegypti IFMs , although the apparent effect on males may also indicate a somewhat less tight regulation of the Actin-4 promoter in Ae . albopictus; or tetracycline may be metabolised slightly differently in the two species leading to sub-optimal repression of the tetO-VP16 in Ae . albopictus . Wide phenotypic variations were also noted among the OX4358-Aal lines , while the Ae . aegypti lines generally had percentages of flying females on tetracycline ( except line A3 ) and of flying males both on and off tetracycline that were more similar to wild-type . One might have expected the OX4358 construct to be more tightly controlled in Ae . albopictus , as it is based on the Ae . albopictus Actin-4 gene . On the other hand , the Ae . aegypti wild-type strain also displayed greater fitness in relation to eclosion rates/ability to fly than the Ae . albopictus wild-type strain , so the observed differences may relate in part to the much longer colonisation time of the Ae . aegypti strain which may have led to a more lab-adapted and homogeneous genetic background . Such heterogeneity in the Ae . albopictus background , despite reducing the rearing efficiency of the strains in a laboratory environment , may represent an advantage in the field as males may be more apt to survive and find mates than some more lab-adapted counterparts . We have shown that it is possible to engineer late-acting , repressible female-specific transgene expression to provide a conditional female-specific flightless phenotype in Ae . albopictus . This study represents a significant step towards genetic control of Ae . albopictus . The flightless phenotype appeared to be somewhat less tightly regulated in Ae . albopictus than in Ae . aegypti; this may lead to lower mass production efficiency and male competitiveness and thereby affect the economics of a control programme based on this technology , or at least these prototype strains . Further work will be required to develop and characterise Ae . albopictus strains homozygous for these transgenes and assess their suitability and effectiveness in suppressing wild populations . Such new methods are needed . As recent chikingunya outbreaks in the Pacific Ocean [28] , [29] highlight , the public health threat posed by the spread of Ae . albopictus , though less than that of Ae . aegypti , remains significant and cannot be considered to be adequately controlled by currently available methods .
|
The Asian Tiger mosquito , Aedes albopictus , is a highly invasive species which took advantage of human activity to spread from South-East Asia to Africa , North and South America , and Europe in the past forty years . Beyond the annoying biting nuisance , this mosquito is also a significant public health threat , capable of transmitting dengue and responsible for an important chikungunya outbreak in the Indian Ocean in 2007 . This mosquito is notoriously difficult to control using current methods , but control techniques involving the release of genetically sterile males have recently shown promising results against the closely related yellow fever mosquito , Aedes aegypti . Wild females inseminated by sterile males have non-viable progeny so if enough sterile males are released , the target population crashes . Female mosquitoes , even if sterile , would bite and potentially acquire and transmit pathogens , so it is crucial to minimise the release of such females . Here we describe the development of genetically engineered strains of the Asian Tiger mosquito to use in such control programmes: the females are unable to fly unless reared with an antidote , thus facilitating male-only releases . The daughters of released males will also be flightless , a lethal phenotype in the field , thus reducing the population and preventing disease transmission .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"biotechnology",
"entomology",
"synthetic",
"biology",
"biology",
"zoology",
"molecular",
"biology"
] |
2012
|
Female-Specific Flightless (fsRIDL) Phenotype for Control of Aedes albopictus
|
Schistosomiasis affects nearly 40 million women of reproductive age , and is known to elicit a pro-inflammatory signature in the placenta . We have previously shown that antigens from schistosome eggs can elicit pro-inflammatory cytokine production from trophoblast cells specifically; however , the influence of these antigens on other characteristics of trophoblast function , particularly as it pertains to placentation in early gestation , is unknown . We therefore sought to determine the impact of schistosome antigens on key characteristics of first trimester trophoblast cells , including migration and invasion . First trimester HTR8/SVneo trophoblast cells were co-cultured with plasma from pregnant women with and without schistosomiasis or schistosome soluble egg antigens ( SEA ) and measured cytokine , cellular migration , and invasion responses . Exposure of HTR8 cells to SEA resulted in a pro-inflammatory , anti-invasive signature , characterized by increased pro-inflammatory cytokines ( IL-6 , IL-8 , MCP-1 ) and TIMP-1 . Additionally , these cells displayed 62% decreased migration and 2 . 7-fold decreased invasion in vitro after treatment with SEA . These results are supported by increased IL-6 and IL-8 in the culture media of HTR8 cells exposed to plasma from Schistosoma japonica infected pregnant women . Soluble egg antigens found in circulation during schistosome infection increase pro-inflammatory cytokine production and inhibit the mobility and invasive characteristics of the first trimester HTR8/SVneo trophoblast cell line . This is the first study to assess the impact of schistosome soluble egg antigens on the behavior of an extravillous trophoblast model and suggests that schistosomiasis in the pre-pregnancy period may adversely impact placentation and the subsequent health of the mother and newborn .
Schistosomes are parasitic worms endemic to many parts of Africa , South America and Southeast Asia . They represent a significant disease burden in endemic regions , and have been estimated to be responsible for as many as 13–15 million disability-adjusted life years ( DALYs ) lost per year , with the true number potentially much higher [1] . Of the estimated 200 million people worldwide infected at any one time , approximately 40 million are women of reproductive age [2] , [3] . A 2002 World Health Organization policy statement recommended the use of praziquantel in pregnant and lactating women [4] , however many women still experience multiple cycles of pregnancy and lactation with schistosomiasis . This is due to the fact that in many regions of the world , pregnant and lactating women are still routinely excluded from treatment initiatives due to the Federal Drug Administration Class B designation that praziquantel still carries , as well as barriers to praziquantel acquisition and distribution . Data from our laboratory and others have demonstrated poor reproductive outcomes in rodents and humans in the context of schistosomiasis . In rodent models , schistosome infection has profound impacts on birthweight and litter size [5]–[7] . We have previously shown that schistosome infection in a population of pregnant women residing in The Philippines is positively associated with increased risk for chorioamnionitis and increased pro-inflammatory cytokines in both maternal and cord blood [8] . Although very few adult worms and/or eggs are thought to directly traffic to the tissues of the maternal-fetal interface [9] , [10] , the residency of adult worms in the mesenteric vasculature and lodging of eggs in the liver allow for continuous secretion of antigens directly into the blood stream of the host . These antigens are known to traffic to , and cross , the human placenta , and have been found in fetal circulation [11]–[17] . We hypothesized that these antigens may have a direct effect on the cells at the maternal-fetal interface , and in the studies described herein , have chosen to focus specifically on processes specific to extravillous trophoblast cells . Many events occur early in gestation that can have profound effects on the subsequent health of the fetus from gestation into adulthood . A lack of data pertaining to schistosomiasis during this critical window of development prompted us to utilize the first trimester cell line , HTR8/SVneo , to investigate the influence of schistosome infection on early events of pregnancy . Initial investigation was performed with co-culture of HTR8 cells and plasma collected from pregnant women infected with schistosomiasis or matched controls . To isolate the schistosome soluble egg antigen ( SEA ) specific effect on trophoblasts from any contribution of the host response , we next evaluated the direct impact of SEA on HTR8 cells . Events critical to placentation , including cytokine production , cellular migration and invasion were all assessed in an in vitro setting .
For the HTR8 human plasma co-culture experiment , written informed consent was obtained from each participant , and the study was approved by the institutional review boards at Rhode Island Hospital and the Philippines Research Institute of Tropical Medicine . We used the immortalized first trimester cell line HTR8/SVneo , originally obtained from a human pregnancy terminated in the first trimester , and displaying properties of invasive extravillous cytotrophoblast cells [18] . Cells were maintained at 37°C with 5% CO2 in 1∶1 DMEM/F-12 media ( Invitrogen , Grand Island , NY ) supplemented with 1% L-glutamine ( Invitrogen ) , 1% penicillin/streptomycin ( Invitrogen ) and 5% fetal bovine serum ( Atlanta Biologicals , Lawrenceville , GA ) . All experiments were performed with the addition of SEA ( 25 µg/ml ) in complete media for 24 h or media only control , unless otherwise noted . This dose was chosen based on dose response curves performed previously in our laboratory using purified primary trophoblast cells [19] . Schistosoma japonicum SEA was generously donated by Dr . Chuan-Xin Yu ( Jiangsu Provincial Institute of Schistosomiasis Control , Wuxi , Jiangsu , China ) after having been prepared according to standard procedure [20] . The SEA was prepared under endotoxin free conditions , with all reagents and equipment used to isolate the SEA from collected livers being LPS free prior to use . Preparations were evaluated for contaminating endotoxin using an FDA-cleared LAL-assay ( Lonza Group , Basel , Switzerland ) . Endotoxin levels for all SEA preparations used were <6 EU/mg protein , which , in our culture conditions , is at least 1000-fold lower than levels that have been previously shown to influence human trophoblast cells [21] . Following the treatment period , media from HTR8/SVneo cells was collected and levels of multiple cytokines , chemokines , and fibrotic markers were assessed on a bead-based platform ( BioPlex , Bio-Rad , Hercules , CA ) using a sandwich antibody-based assay as previously described [22] . Cytokines evaluated included interleukin ( IL ) -1β , IL-6 , interferon ( IFN ) -γ , tumor necrosis factor ( TNF ) -α , IL-4 , IL-5 , IL-10 , IL-13 , IL-12 , IL-8 and IL-2 . These specific cytokines were measured , as they were all components of a multiplex analysis developed and validated in our laboratory . The majority of these cytokines have been reported to be expressed by the placenta , although expression is highly variable depending on culture conditions , gestational age and disease status [23]–[26] . In addition , we measured levels of tissue inhibitor of metalloproteinases ( TIMP ) -2 , TIMP-4 , insulin-like growth factor binding protein ( IGFBP ) -5 , matrix metalloproteinase ( MMP ) -9 , tenascin C , syndecan 1 , Fas ligand , osteopontin , TIMP-1 , connective tissue growth factor ( CTGF ) , macrophage inflammatory protein ( MIP ) -1α , MMP-1 , IGF-1 , MMP-8 , monocyte chemotactic protein ( MCP ) -1 , and TIMP-3 . These analytes were developed into multiplexed assays due to their importance in schistosomiasis-associated and idiopathic pulmonary fibrosis [27]–[29] . However , they are also widely implicated in invasion and remodeling at the maternal-fetal interface , underscoring the similarities between these two processes . For the human plasma assays , we utilized plasma collected from pregnant women at 32 weeks gestation residing in Leyte , the Philippines , an area endemic for S . japonicum . The study population and sample collection has been described elsewhere [8] , with socioeconomic status ( SES ) , gravida , parity , body mass index ( BMI ) , smoking status and age determined via questionnaire [30] . Schistosomiasis and co-infections ( Ascaris lumbricoides , Trichuris trichuria , and hookworm ) were determined from stool samples using the Kato Katz method . Infection intensities for each were determined using the WHO guidelines [31] . From this larger cohort of 150 women , we selected 29 women infected with schistosomiasis ( 20 lightly infected [1–99 eggs per gram] , 9 moderate infection [100–399 epg] ) , and 29 uninfected women matched to the infected women for SES , co-infections , gravida , parity , gestational age , BMI , smoking status and maternal age ( Table 1 ) . Schistosomiasis was evaluated as a nominal ( yes/no ) variable due to the low numbers of women with moderate-high infection intensities . Serum was collected at the only pre-natal study visit ( 32 wks gestation ) . HTR8/SVneo cells were cultured to 80% confluency in complete media before being cultured for 48 hours in serum free media with the addition of 10% plasma from the aforementioned pregnancies . Following 48 h incubation , trophoblast culture supernatants were collected and analyzed for cytokine production as described above . HTR8/SVneo cells were cultured to 100% confluence in complete media . Once completely confluent , a scratch was made across the well using a sterile pipette tip . The underside of each of the wells was cross-hatched for reference . Cells were briefly washed with PBS in order to remove all detached cells after scratching , and cultured in serum-free media with the addition of SEA ( 25 µg/ml ) for 48 h . Phase contrast images of the denuded region were taken at 0 h , 24 h and 48 h after scratch formation using an Olympus IX70 inverted tissue culture microscope ( Olympus Corp . , Tokyo , Japan ) . The same region of each well was imaged at each time point , using the cross-hatching as reference . The area free of cells was quantified using ImageJ software ( NIH , Betheseda , MD ) . The denuded area at 24 h and 48 h for a specific well was expressed as a percentage of the denuded area that had been present in that well at 0 h , thus controlling for well-to-well variation in original scratch sizes . MTT assays were done on HTR8/SVneo cells in parallel to the migration assays . MTT ( Sigma Aldrich , St . Louis , MO ) was added to each well and the cells were incubated for 4 h at 37°C in a humidified environment . Media and MTT were aspirated from each well , MTT solvent ( 4 mM HCl , 0 . 1% Nonidet P-40 , in isopropanol ) added , and the plate incubated at 25°C in the dark with rotation for 15 minutes . Absorbance for each well was read at 560 nm and 630 nm . HTR8/SVneo cells at 80% confluency were treated with SEA ( 25 µg/ml ) for 24 h in complete media before being gently trypsinized , washed with complete media and resuspended in serum-free media . 25 , 000 cells/well were plated on matrigel-coated transwell inserts with an 8 µm pore size ( Corning , Tewksbury , MA ) . The bottom chamber contained complete HTR8 media . Following 48 h incubation , cells and matrigel were gently removed from the top of the transwell , and those cells that had invaded through the matrigel , traversed the pores , and reached the bottom of the transwell were stained with hematoxylin ( Sigma Chemical ) . Stained cells were visualized and counted using an Olympus BH-2 microscope ( Olympus Corp . ) . Data analysis was performed using JMP v . 10 ( SAS Institute , Cary , NC ) . All data were evaluated for normality using the Shapiro-Wilk test . Those experiments for which all data were normally distributed were further evaluated with ANOVA and t-tests , with means ± SEM reported . For data that was not normally distributed , Wilcoxon Signed Rank analyses were performed , with data reported as median ± IQR . Specifically , cytokine production by HTR8 cells was compared between cells exposed to uninfected plasma and those exposed to infected plasma ( Figure 1 ) as well as cells cultured with media alone and media with SEA ( Figures 2 and 3 ) . Similarly , HTR8 migration and invasion were compared between cells cultured with media alone and those with SEA addition to the media ( Figures 4 and 5 ) . Statistical significance was considered as P<0 . 05 .
For these experiments , we selected a sub-set of plasma samples collected from women at 32 weeks gestation as part of a previous study [8] , matching samples from 29 infected women with 29 samples from uninfected women on key potential confounding covariates ( Table 1 ) . HTR8/SVneo cells were cultured in serum-free media supplemented with 10% plasma collected at 32 weeks gestation , and allowed to remain in culture for 48 h . Media from HTR8/SVneo cells cultured with plasma from schistosome-infected women had significantly higher levels of the pro-inflammatory cytokines IL-6 and IL-8 ( 29% , P<0 . 02 and 42% , P<0 . 01 , respectively ) , compared with cells cultured in the presence of plasma from uninfected pregnant women ( Fig . 1 ) . Levels of both IL-6 and IL-8 in the plasma alone were in most cases undetectable , with the highest level of either cytokine across all plasma samples being 32 pg/ml ( data not shown ) . These data indicate that schistosomiasis results in the production of some factor ( s ) , present in maternal circulation , that stimulate a pro-inflammatory cytokine response by first trimester trophoblasts . Given that SEA are found in the circulation of infected individuals and are known to cross the placental barrier , we treated HTR8/SVneo cells with purified SEA ( 25 µg/ml ) in culture . Within 24 h of culture , HTR8 cells treated with SEA secreted higher levels of IL-6 ( 3 . 2-fold , P = 0 . 01 ) , IL-8 ( 1 . 5-fold , P = 0 . 02 ) and the pro-inflammatory chemokine , MCP-1 ( also known as CCL2; 1 . 7-fold , P 0 . 04 ) into the culture media , as compared to HTR8/SVneo treated with media alone ( Fig . 2 ) . Not only do first trimester cell models respond to SEA with a pro-inflammatory signature , they also secrete higher levels of a chemo-attractant protein that may help recruit specific immune cells to the placenta , exacerbating the inflammatory reaction initiated by SEA at the maternal-fetal interface . In addition to cytokine analysis , media from HTR8/SVneo cells treated with SEA for 24 h in culture were assessed for altered levels of a number of fibrotic markers . Of these , TIMP-1 production was increased in media from HTR8 treated with SEA , compared to media from cells with no SEA exposure ( Fig . 3 ) . In contrast , production of CTGF , CCL18 , TIMP-3 and fibronectin all showed no difference in levels secreted from HTR8/SVneo cells exposed to SEA compared to those with media alone ( data not shown ) . Given the important role of TIMP-1 in the inhibition of MMPs , SEA may influence the ability of these first trimester cells to remodel and migrate into the maternal uterine wall . We performed in vitro wound assays to assess migration of the first trimester cell line , HTR8/SVneo , following exposure to SEA . There was little difference in cell migration at 24 h ( Fig . 4a ) . By 48 h however , HTR8 cells in culture with SEA had filled only 57±8% of the denuded area compared to 92±6% for untreated cells ( P<0 . 01 , Fig . 4 ) . Cell proliferation was measured using MTT assay in all wells , however no differences were observed between those wells with SEA added compared to those with media alone indicating the wound closure was not due to a proliferative effect of SEA treatment ( data not shown ) . Together , these data suggest that migration of first trimester trophoblast cells is decreased in the presence of SEA . We next assessed the ability of the HTR8/SVneo cell line to invade through matrigel , a model of extracellular matrix , and a transwell insert after being treated with SEA , using a standard invasion assay [32] . Cells were pretreated with SEA for 24 h to minimize any direct effect of SEA on the matrigel . Enumeration of the cells that had traversed the matrigel and the pores of the transwell after an additional 48 h incubation showed 2 . 7-fold lower absolute cell numbers in those wells that had been pre-treated with SEA compared to the HTR8 cells that received media alone for the initial 24 h ( P = 0 . 04 , Fig . 5 ) . These data indicate that SEA inhibits the migratory and invasive properties of the first trimester cell line , HTR8/SVneo .
Despite a 2002 WHO recommendation that pregnant and lactating women be considered for inclusion in treatment programs [33] , pregnant women with schistosomiasis are still excluded in many regions pending further evaluation of praziquantel's safety during pregnancy . Data regarding the impact of schistosome infection on human pregnancy is rather scant , although we have reported an increase in pro-inflammatory markers in maternal , placental , and newborn compartments of pregnancies complicated by schistosomiasis , as well as increased rates of acute subchorionitis in these women [8] . Several human studies have evaluated the role of schistosomiasis during pregnancy [11] , [34] , [35] . Two observational studies reported lower birth weights among infants from infected mothers . However , methodological issues , including lack of control for important potential confounders such as socioeconomic status and maternal nutritional status [35] , and potential selection bias [11] make interpretation difficult . A recently completed RCT evaluating treatment of schistosomiasis during pregnancy in Uganda reported no change in birth weight among women treated for schistosomiasis during the second trimester and untreated controls [34] . Any effect ( s ) of schistosomiasis on early placentation however would not be detected because treatment occurred late in gestation . Thus , questions regarding the influence of schistosome infection during the first trimester of pregnancy in humans remain largely unanswered . Although direct trafficking of the adult worm or schistosome eggs to the maternal-fetal interface is thought to be a rare event [9] , schistosomiasis is known to produce a distinct antigenic signature in the circulation of infected individuals , including the presence of high levels of soluble egg antigens ( SEA ) which can cross the placental barrier [12] . Previously , we demonstrated that SEA can cause pro-inflammatory cytokine production in primary human trophoblast cells taken at term and allowed to syncytialize in vitro [19] . However , placentation is a dynamic process requiring trophoblast populations distinct in time and differentiation lineage to behave in specific and unique ways . In this manner , a syncytialized term trophoblast is responsible for very different functions ( i . e . nutrient transfer , cytokine and hormone production ) than a first trimester invasive trophoblast cell . In this report , we have focused on the effect of SEA on first trimester trophoblast cells , using the cell line HTR8/SVneo , as these cells represent one of the best model systems available for studying the behavior and characteristics of invasive , extravillous trophoblast cells [36] . As we have previously reported in term syncytialized trophoblasts , HTR8 cells exposed to SEA for 24 h ( 25 µg/ml ) exhibited a pro-inflammatory cytokine signature . These findings echoed the pro-inflammatory signature we observed in HTR8 cells exposed to plasma from pregnant women infected with schistosomiasis , compared to plasma from pregnant , uninfected controls . A potential limitation of these experiments is that HTR8 cells were cultured with maternal plasma collected during the third trimester because the original study from which these samples originated did not enroll pregnant women until 32 weeks gestation [8] . We do not expect this to influence either the validity or generalizability of our results as schistosomiasis infection status and its consequent host response should not be altered during gestation . Although little is known regarding the impact of localized pro-inflammatory cytokines during the first trimester , they have been suggested to contribute to reduced migration and invasion of trophoblast cells , increased migration of innate immune cells to the maternal fetal interface , and , at very high levels , are postulated to play a role in preterm delivery and/or miscarriage [21] , [37] , [38] . Another major role of extravillous trophoblast cells , particularly in the first trimester , is to remodel and invade deep into the maternal endometrium . This process is tightly regulated , and failure to invade to the appropriate degree has been associated with the development of a number of gestational diseases , most importantly preeclampsia , preterm birth and low birth weight [39] , [40] . Failure to identify any difference in the cellular metabolic activity ( MTT assay as a surrogate for cell proliferation ) between the untreated and SEA-exposed cells supports the idea that SEA is not simply cytotoxic to HTR8 cells . Rather , HTR8 cells display increased production of TIMP-1 , inhibition of cellular migration and decreased levels of invasion through matrigel when exposed to SEA . Of the fibrosis-associated molecules measured , TIMP-1 is arguably the most relevant to trophoblast migration/invasion [41] , [42] . These data suggest that the process of placentation could be compromised in pregnancies complicated by schistosomiasis during the first trimester . To our knowledge , there have been no studies examining the effect , if any , of schistosomiasis on gestational diseases such as preeclampsia . Our data are consistent with previous work from our laboratory regarding schistosome induced pro-inflammatory cytokine production across different models of trophoblast cells [19] . Outside of pregnancy , we and others have related these responses to nutritional , hepatic and hematologic morbidities in infected individuals [43]–[46] . Surprisingly , maternal schistosomiasis during pregnancy elicits a pro-inflammatory response detectable in the neonate [8] , and neonates exposed to maternal schistosomiasis during pregnancy display a more robust response to antigenic challenge and have elevated levels of antibodies against schistosome antigens at birth than their unexposed counterparts [13] , [47] . Together , these results suggest that maternal schistosomiasis may influence the outcome of initial pediatric schistosome infections acquired during early childhood . The finding that SEA may modify invasion of extravillous trophoblasts and alter the cytokine milieu at the maternal fetal interface lends support to an aggressive treatment approach for women of reproductive age , such that they enter pregnancies infection free . It should also be noted that studies which have , and will , evaluate the efficacy of praziquantel given after the first trimester ( ClinicalTrials . gov , registered study number NCT00486863 ) , may not capture the full benefit of treatment as it relates to early placentation processes . Studies regarding the incidence of gestational diseases such as preeclampsia in the context of high schistosome prevalence are warranted and may shed additional light on the impact of schistosomiasis on the early development of the human placenta .
|
Approximately 40 million women of childbearing age suffer from schistosome infection globally at any given time . Multiple studies in rodent models , as well as a few reports in humans , suggest that schistosome infection results in poor pregnancy outcomes . We have previously shown that antigens released from schistosome eggs result in a pronounced pro-inflammatory response in syncytialized third trimester trophoblasts . Herein , we examine the effect of schistosome egg antigens on a first trimester trophoblast cell line , an accepted model for early placental development . Not only is the pro-inflammatory response recapitulated in this model system , but we also observed a decrease in migration and invasion of trophoblast cells after exposure to these antigens . Both migration and invasion are key aspects in early placental development , and inadequate invasion has been implicated in pregnancy-related diseases such as growth restriction and preeclampsia . This study is the first to examine the impact of schistosome antigens on early placental development , and may have implications for the subsequent health of both the pregnancy and the child .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"schistosomiasis",
"obstetrics",
"and",
"gynecology",
"pregnancy",
"global",
"health",
"neglected",
"tropical",
"diseases"
] |
2013
|
Schistosoma japonicum Soluble Egg Antigens Attenuate Invasion in a First Trimester Human Placental Trophoblast Model
|
Emerging evidence indicates that gene products implicated in human cancers often cluster together in “hot spots” in protein-protein interaction ( PPI ) networks . Additionally , small sub-networks within PPI networks that demonstrate synergistic differential expression with respect to tumorigenic phenotypes were recently shown to be more accurate classifiers of disease progression when compared to single targets identified by traditional approaches . However , many of these studies rely exclusively on mRNA expression data , a useful but limited measure of cellular activity . Proteomic profiling experiments provide information at the post-translational level , yet they generally screen only a limited fraction of the proteome . Here , we demonstrate that integration of these complementary data sources with a “proteomics-first” approach can enhance the discovery of candidate sub-networks in cancer that are well-suited for mechanistic validation in disease . We propose that small changes in the mRNA expression of multiple genes in the neighborhood of a protein-hub can be synergistically associated with significant changes in the activity of that protein and its network neighbors . Further , we hypothesize that proteomic targets with significant fold change between phenotype and control may be used to “seed” a search for small PPI sub-networks that are functionally associated with these targets . To test this hypothesis , we select proteomic targets having significant expression changes in human colorectal cancer ( CRC ) from two independent 2-D gel-based screens . Then , we use random walk based models of network crosstalk and develop novel reference models to identify sub-networks that are statistically significant in terms of their functional association with these proteomic targets . Subsequently , using an information-theoretic measure , we evaluate synergistic changes in the activity of identified sub-networks based on genome-wide screens of mRNA expression in CRC . Cross-classification experiments to predict disease class show excellent performance using only a few sub-networks , underwriting the strength of the proposed approach in discovering relevant and reproducible sub-networks .
Colorectal cancer ( CRC ) is the second leading cause of cancer death in adult Americans [1] . Interest in this complex disease is represented by a very mature body of research , much of it at the genomic level . Yet the identification and verification of proteins that have a functional role in the patho-physiology of CRC remains an important goal as proteins directly mediate the functions dysregulated in the disease . Modern , high-throughput proteomic methods provide one way of profiling the significant changes in protein expression of tumor samples with respect to control , using tissue biopsies obtained from patients diagnosed with this disease [2]–[5] . Proteomic screening techniques are particularly useful for furthering the understanding of the mechanisms that underlie complex phenotypes like CRC , in that they provide information at the post-translational level . However , due to various biological and experimental constraints ( e . g . , ascertainment bias and physical properties of proteins ) , proteomic methods may screen only a limited fraction of proteins and protein isoforms present in cells and tissues . We propose that this limitation may be mitigated through the integration of proteomic data with genome scale data sources , such as measurements of gene expression . In addition , protein-protein interaction ( PPI ) databases , which are rapidly growing in terms of both the quality and quantity of their annotations , provide another source of genome scale data integration [6] . Such integrative approaches can potentially lead to functional inference at the systems level , through identification of pathways and molecular sub-networks that are implicated in CRC . In support of this approach , a recent review by Ideker and Sharan [7] summarizes studies that indicate that genes with a role in cancer tend to cluster together on well-connected sub-networks of protein-protein interactions . This suggests a hypothesis that the synergistic expression of multiple cancer-related genes at the level of mRNA can co-regulate the expression of proteins in their immediate “network neighborhood” . These differentially expressed proteins may be captured by expression proteomics experiments , thus their network neighborhood should provide an ideal starting place to search for sub-networks with a possible role in the disease . The effectiveness of network-based approaches to the identification of multiple disease markers has been demonstrated in the context of various diseases , including Huntington's disease [8] , the inflammatory response [9] , and human breast cancer [10] . Furthermore , it was recently shown that “differentially expressed sub-network markers” were more accurate predictors of metastasis in breast cancer ( compared to single gene markers ) [11] . However , existing approaches are generally limited to mRNA expression data in terms of quantification of molecular expression , which captures post-transcriptional activity only to a limited extent [12] , [13] . Consequently , inclusion of protein expression data in the search for sub-network markers has the potential to improve the effectiveness of systems biology approaches [14] . However , it remains largely unknown how a network-based approach may be enhanced when starting with proteomic data . In this paper , we propose a novel computational approach that takes into account certain topological features of the interactome , namely connectivity and proximity , for searching the neighborhoods of proteomic targets to find significant sub-networks implicated in CRC . In doing so , we partly overcome ( i ) the bias inherent in proteomic profiling experiments , particularly those that are gel-based , which are typically limited to capturing changes only in relatively abundant proteins and ( ii ) the noise , missing data , and ascertainment bias in PPI data . This is accomplished by assessing the functional association between proteins based on the quantification of the statistical significance of network crosstalk through information-flow based modeling of the PPI network and development of a reference model that takes into account the network connectivity of proteomic targets . We hypothesize that identification of candidate sub-networks with a significant association to proteomic targets can reveal proteins that are not detected to be differentially expressed at the level of the proteome , but whose activity in the network may play a key role in maintaining the phenotype . Consequently , the proposed framework provides a means for expanding proteome expression data to infer a role for proteins that exhibit significant crosstalk to the proteomic targets . The flow of the proposed computational framework is illustrated in Figure 1 . A key objective of this study is to systematically elaborate a proteomics-driven approach as a sound method for inferring small sub-networks implicated in complex phenotypes , and ultimately make these methods practically available to a wider community of researchers working in this area . For this purpose , we ground our approach on the hypothesis that the observed fold change of the proteomic targets may be associated with the synergistic dysregulation of their interacting partners at the level of mRNA . From a computational perspective , our hypothesis is based on the premise that sub-networks which exhibit significant association with the proteomic targets should also show a significant change in activity between control and cancer . To test this hypothesis , we first score each protein in the network based on their crosstalk with the proteomic targets . In order to account for noise , incompleteness of data , and ascertainment bias , we also develop novel methods for assessing the significance of these “crosstalk scores” . Then , for each proteomic target , we identify a candidate sub-network that is composed of its interacting partners with significant crosstalk scores . Subsequently , using an information theoretic measure , we evaluate the synergistic differential expression of these candidate sub-networks between control and disease , based on changes in mRNA expression obtained from microarray experiments performed on tissue biopsies collected from a cohort of patients with CRC . Finally , using the sub-networks that exhibit significant synergistic dysregulation as features , we develop classifiers to predict disease class across different data sets . The proposed computational approach for assessing functional association between proteomic targets and other proteins uses a random-walk based algorithm . Recently , Kohler et al . [15] and Chen et al . [16] used similar network algorithms to prioritize candidate disease genes implicated by linkage analysis in a variety of human diseases . Vanunu and Sharan [17] developed a global , propagation-based method that exploits information on known causal disease genes and PPI confidence scores . Their method more accurately recovered known disease gene relationships compared to several other extant methods . In contrast to these applications and rather than using raw scores obtained by such information flow based algorithms , we develop reference models to assess the statistical significance of these scores , with a view to identifying proteins that are significantly associated with proteomic targets . Furthermore , our biological hypothesis , which drives our approach , is that targets ( proteomic or genomic ) significant for the CRC phenotype may reside in or near cancer hotspots in the network , and thus present an ideal starting place to search for high-value sub-networks associated with the disease . Therefore , our computational approach does not rely on canonical disease-related genes or proteins; rather , it is a global , unbiased search that tries to identify network interactions statistically significant with respect to all targets in an experimentally-derived set . Our previous work in this area [5] was limited in scope due to the lack of access to the topology of the commercial PPI we employed . This prevented us from assessing the importance of topology for sub-network generation , which is the primary focus of our computational approach in this study . Likewise , our network scoring and statistical hypothesis testing were all greatly limited in the previous work due to incomplete access to an unpublished microarray data . For the same reason we were practically prevented from iteratively adjusting network search parameters in the commercial software that would have generated a large list of candidate sub-networks for scoring . Here we describe a new network search method for finding high-value candidate sub-networks associated with CRC . To overcome the limitations of the previous study and to permit independent evaluation of our methods , we utilize a public PPI ( HPRD ) and public microarrays ( Gene Expression Omnibus ) to evaluate performance using two independent sets of proteomic targets obtained by 2D-PAGE that are also publically available . We compare this result to that obtained using a set of CRC driver gene mutants as seeds for the network search . The basis for this test is the hypothesis that if mutated gene products map to cancer hotspots on the network , they would be similarly useful as seeds for our network search algorithm . To reveal the practical utility of our integrative approach , and to extend it beyond merely a theoretical computational framework , we validate by western blot several targets in a sub-network predicted by our method to be dysregulated , using a cohort of tissue biopsies not used in the original proteomic screen . Finally , we employ a cross-validation approach to compare the disease classification performance of the proteomic-versus genomic-derived sub-networks . Our results show that the proposed proteomics-driven approach , as it integrates a variety of biologically relevant data , can identify significant sub-networks implicated in a complex phenotype , i . e . CRC . The definition of terminology frequently used in this paper is provided in Table 1 .
We evaluated the individual differential gene expression of each crosstalker identified using the Nibbe and Friedman proteomic seeds using two microarray datasets obtained from GEO ( GSE10950 & GSE8671 ) . GSE8671 represents 64 experiments using mRNA isolated from tissue biopsies obtained from 32 patients ( matched tumor and adjacent normal mucosa ) performed on an Affymetrix GeneChip ( Human U133 Plus 2 . 0 ) . Similarly , GSE10950 represents 48 experiments on matched tissue biopsies ( 24 patients ) performed on an Illumina array ( Human ref-8 , v2 . 0 ) . The cumulative distribution of individual differential expression scores for proteomic seeds , ( and a seed of CRC driver genes discussed later ) , as well as all proteins in the network computed as described in the Materials and Methods section , is shown in Figure 2 ( please see the Materials and Methods section for details on how differential expression is quantified ) . As seen in the figure , we found no significant difference in the distribution of individual differential expression of the crosstalkers , as compared to the distribution of differential expression of all proteins in the HPRD network . This observation indicates that at the level of individual genes , significant network crosstalk with proteomic seeds in CRC is not associated with transcriptomic dysregulation in CRC . For the purpose of discussion we will refer to a sub-network by the proteomic seed that induced the sub-network ( e . g . TCP1 ) . For each version of each sub-network we computed the mutual information ( MI ) of each sub-network between control and tumor using the mRNA expression data from microarrays GSE10950 and GSE8671 ( see Computational Methods ) , and we used this score to estimate the significance of the various networks in differentiating the phenotype ( Figure 1 ) . The comparison of mutual information for the two versions of each sub-network associated with the Nibbe seed is shown in Figure 3 . We plotted the results only for those ( crosstalker ) sub-networks where the mutual information exceeded 0 . 35 ( approximately 1σ from random mean ) . The purpose of this analysis is to understand how the synergy of each crosstalker sub-network compares to that of its corresponding interactor sub-network . The MI and significance scores for all sub-networks can be found in Supplemental Table S1 . Of the 46 candidate sub-networks associated with Nibbe proteomic seeds , 10 unique interactor sub-networks ( green squares ) exhibited significant MI scores . For five of these sub-networks ( CCT2 , TCP1 , SYNCRIP , HNRPF and HNRPH1 ) the crosstalker version of the sub-networks was found to have enhanced MI on one or the other microarray datasets . Two crosstalker sub-networks ( red diamonds ) , CCT2 and TCP1 , show improvement over their corresponding interactor sub-network on both arrays . Notably , on GSE10950 , the mutual information score of the TPI1 crosstalker sub-network is significant , while the corresponding interactor sub-network failed to show significance . Figure 4 shows the corresponding plots for the Friedman proteomic seeds . Here , seven unique interactor sub-networks have significant MI scores; two of them ( ANXA3 and PSMA6 ) were common to both sets of microarray data . For the Friedman seeds , the crosstalkers for candidate sub-network TUBA1B showed dramatically increased mutual information compared to its interactor network . Furthermore , four other crosstalker sub-networks ( associated with MYL9 , GARS , ANXA3 and GSTP1 ) all revealed much higher synergy compared to their corresponding interactor sub-networks , two of which ( MYL9 , GSTP1 ) failed to show significance on either array . We discuss a possible explanation for these findings in the Discussion section . Figures 5a and 5b show unions of crosstalker sub-networks associated with the Friedman and Nibbe seeds , respectively , for which the synergy was higher than the corresponding interactor sub-network . The graphs reveal that many proteomic seeds reside within or near dense sub-networks of crosstalkers . We observed that several of the sub-networks generated using the two proteomic seed sets contained proteins in common . In particular , certain sub-units of the TCP1 complex exhibited marked crosstalk in the sub-network induced by CCT2 in the Nibbe seed , and TUBA1B in the Friedman seed ( Figure 4 ) . In addition , we had previously shown [5] that certain sub-units of this complex ( CCT3 , CCT5 , and CCT7 ) were also significant for the late-stage CRC phenotype , as revealed by a similar network scoring methodology but using a commercial PPI unrelated to HPRD . TCP1 ( or TCPα ) is a hetero-oligomeric complex comprised of two stacked ring structures , each composed of eight known subunits and plays a functional role in maintaining the CRC phenotype . Specifically , it was shown [18] to be required for the proper biogenesis of PLK1 , a kinase that has a critical role in cytokinesis . However , other than their role as sub-units in the formation of the TCP complex little is known about the independent role , if any , of these sub-units in CRC [19] . Consequently , these targets present an opportunity for follow-on mechanistic studies . For this reason , we verified the protein expression of TCP1 , CCT3 , CCT5 , CCT7 , and PLK1 by western blot in a separate cohort of three patient sample pairs not used in screening phase , and compared this to the average expression at the level of mRNA ( Figure 6 ) . Consistent with our hypothesis , the data indicate co-regulation at the level of mRNA and protein , but also reveal the wide variability of expression of these targets among individual patients . CCT3 and CCT7 were dramatically over-expressed in two patients ( 507 and 534 ) , but less so in patient 540 , which was similar to the pattern for PLK1 . Although these data show that proteomic seeds are well-suited for identifying synergistically dysregulated sub-networks , we wished to investigate the power of genetically identified seed sets in discovering significant sub-networks . As CRC is commonly thought to be caused by the accumulation of somatic mutations , a number of cancer research labs have collaborated to conduct whole genome sequencing to identify the genes thought to be “drivers” in cancer , i . e . those represented by the set of genes that appeared most frequently mutated in a robust cohort of clinical biopsies . The results of one such study on human breast and colon cancer were recently reported by Sjöblom et al . [20] . We hypothesized that the gene products of the CRC driver genes reported in this study would be located at hotspots in the interactome . Further , if the mutations lead to dysregulation of neighboring genes at the level of mRNA , then the seed should reveal significant sub-networks using our method . Additionally , since there is less bias in PCR sequencing and high genome coverage , at least as compared to proteomic profiling , we supposed that driver gene seeds ( n = 42 ) might be superior both in terms of the number and significance of the sub-networks identified . As shown in Figure 7 , when scored by GSE8671 , only four significant sub-networks were found . Strikingly , for every one of them , only the crosstalker sub-networks were significant . Using GSE10950 , seven sub-networks of crosstalkers were significant , including all four found on GSE8671 . For all but two of the sub-networks ( P2RX7 , OBSCN ) , the crosstalkers show substantially higher synergistic differential expression as compared to their interactor counterparts . Notably , APC , a tumor suppressor gene widely viewed as the “gate-keeper” in CRC , was associated with a significantly dysregulated sub-network with respect to both arrays , and of all the genes in the driver seed it was found to be mutated in the highest percentage ( 90% ) of the clinical samples . This expected finding may be viewed as a positive control for our analytical method . In terms of the overall number of significant sub-networks identified , however , there was no apparent improvement using the driver gene seed set versus either proteomic seed set . Additionally , a number of the significant crosstalk sub-networks identified by the proteomic seeds show markedly higher synergy ( MI>0 . 60 ) than all but one ( EVL ) of the sub-networks found by the driver gene seed . We evaluated the quality of the crosstalker versus interactor sub-networks in terms of their ability to classify tumor versus control on the microarrays , using an SVM-based classifier in a cross-validation approach ( see Materials and Methods ) . The significant sub-networks in each group were first ranked by MI , and the features were valued by superposing the mRNA expression values of each gene in the sub-network . When trained on GSE10950 and validated on GSE8671 , proteomic crosstalkers outperformed the interactor sub-networks ( both proteomic and genomic ) when the number of features used to train the classifier was three or less . Beyond three features , both the proteomic interactor and CAN ( candidate CRC driver genes ) crosstalker sub-networks outperformed the proteomic crosstalkers ( Figure 8a ) . Performance was similar when the training and validation sets were reversed , although the performance of proteomic crosstalkers dropped when more than two sub-networks were used for classification ( Figure 8b ) . The raw classification data are provided in Supplemental Table S1 .
We have shown that proteomic targets showing significant expression changes for a complex phenotype , such as CRC , provide valuable inputs for our algorithms designed to discover phenotypically significant sub-networks with connectivity and proximity to these targets . In addition , certain crosstalker sub-networks , when scored with respect to phenotype by the measure of mutual information , display significant differential synergistic expression at the level of mRNA with respect to the seed targets . When these implicated sub-networks contain proteins with no known role in the disease , they present new opportunities for follow-on mechanistic experiments to verify the in silico inference of biological significance in the disease . This point cannot be over-emphasized , because in our view the promotion of a candidate , disease-associated sub-network to an functional sub-network with a validated role in disease must be accomplished by wet lab experiments . As mentioned in the previous section , with respect to the proteomic seeds , a number of the same sub-networks showed significance ( >1σ from background ) when scored by either GSE10950 or GSE8671 . With respect to the driver gene seed , every sub-network that showed significance when scored by the GSE8671 array was also found to be significant when scored by the GSE10950 array . One explanation for why the sub-networks with respect to a given set of proteomic seeds did not show complete redundancy between arrays is that the microarrays represent experiments performed on different pathologic stages of CRC tumors , very early stage in the case of GSE8671 ( adenoma ) versus a more established tumor in GSE10950 ( primary ) . The pathologic stage of the proteomic samples in the Nibbe seed was homogenous late stage CRC ( Duke's D ) while the Friedman seed was a mix of mid to late stage samples ( Duke's B–D ) . This highlights a potential limitation of an integrated –omics approach , namely , it is often difficult to establish an optimal match of the biology underlying the measures made at the level of the proteome and transcriptome . However , in our case , if the sub-networks become dysregulated early in the disease and have a role in maintaining the phenotype through later stages , this limitation can turn into an opportunity for development of hypotheses regarding the mechanisms of the progression of CRC . In particular , the complete overlap of crosstalk sub-networks between arrays observed with the driver gene seed indicates the synergistic activity of these sub-networks may be independent of pathologic stage . We also noted that only a relatively small fraction of the seeds induced significant sub-networks , either interactors or crosstalkers , and this was the case for both the proteomic and the genomic seeds . One potential explanation for this observation is that current human PPI networks capture only a very small fraction of all protein relationships in the human interactome [21] , and therefore cannot be expected to reveal a significant sub-network for every experimentally determined seed . As these networks improve , we expect their value in uncovering interesting biology will only grow . The classification performance indicates that experimentally-derived proteomic disease targets combined with our network search algorithm can discover high-valued sub-networks for mechanistic in vivo verification . This was consistent with our hypothesis , and supports the claim that a proteomic seed can identify sub-networks that provide additional pathways of interest ( e . g CCT2 , TCP1 ) . To strengthen this claim , in an independent cohort of patient biopsies , we validated the differential expression of several targets in the TCP-1 sub-network , predicted by our model to be coordinately dysregulated . The genomic seed showed excellent classification performance , and crosstalkers were superior in most instances to their corresponding interactor sub-networks , consistent with our computational hypotheses . When three or more features were used to train the classifier they were also better than the proteomic crosstalkers . However , this result is not entirely unexpected as the proteomic data has low coverage and may lack key seeds and thus may lack important sub-networks . However , the favorable classification performance of the genomic-derived sub-networks may be viewed as a positive control for this experimental approach . Alternatively , it is unlikely that all relevant sub-networks are regulated at the level of transcription , and this may reduce the number of significant sub-networks discoverable by our approach . Never-the-less , the approach can be generalized to many proteomics expression data sets to discover novel sub-networks dysregulated in many complex diseases . In many classification applications , high dimensionality is an important problem and it is often desirable to be able to choose a small number of features that will provide reasonable performance ( to overcome “curse of dimensionality” ) . In this respect , the classification performance provided by only a few sub-networks is indeed very promising , in that “crosstalk to proteomic targets” may actually provide a shortcut to the identification of a compact set of useful sub-network features . As our classification experiments were carried out in a cross-classification setting , the high accuracy of classification using up to three sub-networks indicates that the most significant crosstalker sub-networks were highly reproducible . Reproducibility is an important concern in classification applications , since if the sub-network features that are used are not reproducible across datasets , this will result in over-fitting . In this regard , the use of proteomic data can also be considered a tool for obtaining useful biological insights for feature selection .
The computational framework for integrating proteomic , transcriptomic , and interactomic data to discover sub-networks implicated in complex phenotypes is shown in Figure 1 . As seen in the figure , we first identify disease targets with significant differential expression with respect to control , via proteomic screening as described above . Once these targets , called proteomic seeds , are identified , we map these seeds on the PPI network obtained from HPRD to identify proteins that are functionally associated with the proteomic seeds . In order to develop biologically sound measures to quantify the functional association between proteins , we develop information flow based algorithms to compute crosstalk scores , which capture network proximity and connectivity to proteomic seeds . We discuss this procedure in Subsections A and B . In order to account for experimental artifacts , incompleteness of data , and ascertainment bias , we use Monte Carlo simulations to assess the significance of the crosstalk scores computed by these algorithms . Our statistical evaluation scheme is based on a reference model that captures the basic characteristics of the proteomic seeds , in terms of the number of seeds and their degree distribution . This procedure is described in Subsection C . Subsequently , for each proteomic seed , we construct two “candidate sub-networks”: ( i ) sub-network induced by all interacting partners of the seed protein , ( ii ) sub-network induced by the interacting partners that have significant crosstalk scores ( in our experiments , we use a p-value cut-off of 0 . 001 to determine “significant crosstalkers” ) . Finally , we evaluate the mutual information score of each candidate sub-network with respect to the phenotype of interest ( in this paper , CRC ) , using mRNA expression data for test and control samples . For this purpose , we use an established information-theoretic scheme that quantifies synergistic differential expression in terms of the mutual information between the aggregate expression of the sub-network and disease classes across samples . This procedure is explained in Subsection D . In order to assess the statistical significance of synergistic differential expression , we also use Monte Carlo simulations based on reference models that accurately capture the basic topological characteristics of each sub-network . This procedure is explained in Subsection E . We then use identified sub-networks to develop classifiers for predicting disease class in CRC . This procedure is explained in Subsection F . Systematic studies of differentially expressed genes in certain phenotype classes show that these genes are related to each other in molecular networks , composed of protein-protein interactions , transcriptional regulatory interactions , and metabolic interactions [22] . In one of the early algorithmic studies , Ideker et al . [23] develop a method for identifying differentially expressed metabolic sub-networks with respect to GAL80 deletion in yeast . This method is based on searching for connected groups of enzymes within the yeast metabolic network , such that the aggregate differential expression of genes coding these enzymes is statistically significant . Variations of this method prove useful in identifying multiple gene markers implicated in a variety of diseases , including prostate cancer [24] , melanoma [25] , and diabetes [26] . Building on these results , information theoretic schemes for assessing synergistic differential expression are also shown to be effective in network based disease classification [11] , [27] . While differential network analysis is effective in identifying multiple gene markers , most of the existing methods utilize network information to primarily find the genes that are connected , hence potentially related to each other . In other words , these approaches do not take into account network topology , connectivity patterns , or degree of connectivity between proteins . This is because ( i ) much of the available network information is noisy and incomplete [28] , therefore , connectivity patterns cannot be interpreted as well-defined wiring schemes , and ( ii ) network models ( particularly , high-throughput protein-protein interactions ) provide only a high-level qualitative description of the information flow in the cell . However , several studies show that variations in molecular expression can be interpreted in terms of network topology ( e . g , subunits of a protein complex are co-expressed significantly over a time course [29] , functional similarity of proteins correlates with proximity in a network of interactions [30] , [31] . Motivated by these considerations , we develop network-based scoring schemes to quantify the crosstalk between proteomic seeds and the rest of the proteins in a network of interactions . Based on the premise that synergistic changes in transcriptional expression may be associated with significant changes in proteomic activity , we expect that proteins that demonstrate significant crosstalk with proteomic seeds will be good candidates for being implicated in the phenotype of interest . In order to assess the crosstalk between a group of proteomic targets and any other protein in the network accurately , we develop information flow based algorithms , as discussed in the next section . Let G = ( V , E ) be a network of protein interactions , where V consists of the proteins in the network , and an undirected edge uv∈E represents an interaction between proteins u∈V and v∈V . For convenience , we also define N ( v ) as the set of interacting partners of protein v∈V , i . e . , N ( v ) = {u∈V: uv∈E} . Let S⊆V be the set of proteomic seeds , i . e . , the proteins that are identified by proteomic studies to exhibit significant fold change with respect to the phenotype of interest . Our objective is to compute a score α ( v ) for each protein v∈V , to quantify the network crosstalk between v and the proteins in S . Here , network crosstalk is used as an indicator of functional association between proteins . In order to develop a biologically sound measure of network crosstalk , we rely on the following observations: ( i ) Functional similarity between two proteins , as measured by semantic similarity of Gene Ontology annotations [32] , is significantly correlated with their network proximity , as measured by the shortest path ( number of hops ) between these proteins [30] , [31] . ( ii ) Existence of multiple alternate paths between two proteins is an indicator of their functional association , since functional multiple paths are often conserved through evolution owing to their contribution to robustness against perturbations , as well as amplification of signals [33] . To incorporate both the number of hops and multiple alternate paths into the assessment of crosstalk between proteins , we use an information flow based algorithm based on random walks with restarts [34] . This algorithm can be considered a generalization of Google's well-known page-rank algorithm [35] . Furthermore , a special case of the proposed crosstalk score , when |S| = 1 , is a network proximity measure [34] known to be closely related to commute distance and effective resistance [36] in graphs . Similar graph-theoretic measures are also used to identify functional modules in PPI networks [37] , annotation of protein function [38] , and prioritization of disease genes [15]–[17] . We assign crosstalk scores to all proteins in the network for a given S by simulating a random walk as follows . The random walk starts at a randomly chosen protein in S . At each step , when the random walk is at some protein v , it either moves to an interacting partner of v with probability 1−r , or it restarts at a protein in S with probability r . Here , the parameter 0≤r≤1 is called the restart probability ( in our experiments , we use r = 0 . 5 ) . For each move , the interacting partner to be moved to is selected uniformly at random from N ( v ) . However , the move probabilities can also be adjusted to reflect the confidence of each interaction , so that more reliable interactions contribute more to the quantification of crosstalk . In other words , one can define the probability of a move from v to u as P ( u , v ) = w ( u , v ) /Σu′∈N ( v ) w ( u′ , v ) if u∈N ( v ) , 0 otherwise . Here , w ( u , v ) denotes the reliability of the interaction between u and v . Similarly , for each restart , the protein to be restarted is selected uniformly at random from S . These probabilities can also be adjusted to reflect the significance of the fold change of each protein in S , so that proteins with more significant fold change are considered as more reliable seed proteins . In other words , one can define the probability of restart at u∈V as ρ ( u ) = zP ( u ) /Σu′∈S zP ( u′ ) if u∈S and 0 otherwise . Here , zP ( u ) denotes the z-score of the fold change of u with respect to the phenotype of interest , based on proteomic screening . Based on this random walk model , we define the crosstalk between the proteins in S and each protein v∈V as the relative amount of time spent at v by such an infinite random walk , or equivalently , the probability that the random walk will be at protein v at a randomly chosen time step after the random walk proceeds for a sufficiently long time . More precisely , let αt denote a |V|-dimensional vector , such that αt ( v ) is equal to the probability that the random walk will be at protein v at step t , where ∥αt∥1 = 1 ( here , ∥ . ∥1 denotes the 1-norm of a vector , defined as the sum of magnitudes of its elements ) . Let P denote the stochastic matrix derived from network G = ( V , E ) , i . e . , P ( u , v ) = 1/|N ( v ) | if uv∈E , 0 otherwise . Then , we have ( 1 ) where ρ denotes the restart vector with ρ ( u ) = 1/|S| for u∈S , and 0 otherwise . Then , letting α0 = ρ , the vector containing the crosstalk scores for each node in the network is given by α = limt→∞ αt . Observe that this formulation lends itself to an iterative algorithm to compute crosstalk scores efficiently , where each iteration requires O ( |E| ) time , since P is a sparse matrix with 2|E| non-zero entries . Note that , when r = 0 , α is equal to the eigenvector of P that corresponds to its largest eigenvalue ( with numerical value 1 ) , i . e . , α ( v ) is exactly equal to the page rank of v in G for all v∈V . Therefore , the crosstalk score of a protein is not only an indicator of its connectivity and proximity to seed proteins , but it is also influenced by the centrality of the protein in the network . In order to account for such sources of bias , as well as the choice of parameter r ( in our experiments , we use r = 0 . 5 ) , we adjust the crosstalk scores statistically as we discuss in the next section . Due to variability in physical properties of proteins and other experimental artifacts , it is likely that there will be significant ascertainment bias in the selection of proteomic seeds , as well as the availability of interaction data for each protein [39] . Indeed , our results show that the seed proteins extracted by proteomic screening are likely to be highly connected in the PPI network derived from HPRD . More specifically , the 60 proteins that are identified to have significant fold change ( p<0 . 01 ) in late stages of human colorectal cancer have 24 . 1 interactions in HPRD on an average , while the average degree of a protein in the HPRD network is 9 . 1 . Consequently , highly connected proteins in the network are likely to be assigned artificially high crosstalk scores just by chance . Since available network data is often incomplete and prone to ascertainment bias , these effects are likely to amplify the ascertainment bias and skew the results toward well-studied proteins . However , we are very interested in finding those proteins that are relatively less characterized but may provide novel insights into phenotype . Therefore , the crosstalk scores described above need to be assigned significance scores based on reliable statistical models . In order to deal with such experimental and data-related sources of bias , we use a reference model that captures the degree distribution of seed proteins accurately . Namely , for a given seed set S , we generate a random instance S ( i ) representative of S as follows . For every protein u∈S , we create a bucket B ( u ) of proteins in the network , such that ∪u∈S B ( u ) = V and B ( u ) ∩B ( u′ ) = ∅ for all u , u′∈S . Here , protein v∈V is assigned to bucket B ( u ) if |N ( v ) −N ( u ) |≤|N ( v ) −N ( u′ ) | for all u′∈S and ties are broken randomly . Then , we construct S ( i ) by choosing one protein from each bucket uniformly at random , so that |S ( i ) | = |S| . Observe that each bucket consists of proteins that have similar number of interactions with a particular seed protein; therefore , each seed protein is represented in S ( i ) by exactly one protein in terms of its number of interactions . Consequently , the expected total degree of the proteins in S ( i ) is likely to be very close to the total degree of the proteins in S . Once a random instance S ( i ) is generated , we compute the corresponding crosstalk vector α ( i ) by letting ρ ( i ) ( u ) = 1/|S ( i ) | for u∈S ( i ) , and 0 otherwise . Repeating this procedure n times , where n is sufficiently large ( we use n = 1000 in our experiments ) , we obtain a sampling {α ( 1 ) , α ( 2 ) , … , α ( n ) } of the null distribution of crosstalk scores , with respect to seed sets that are representative of S in terms of their size and degree distribution . We then estimate the mean μS = Σ1≤i≤nα ( i ) /n and standard deviation σS2 = Σ1≤i≤n ( α ( i ) −μS ) 2/ ( n−1 ) of the null distribution of crosstalk scores for S using this sample . Subsequently , we compute adjusted crosstalk scores ( 2 ) for each protein v∈V . These adjusted crosstalk scores represent the statistical significance of the crosstalk between each protein and the proteins in the seed set , accounting for the centrality of the protein the network , as well as the degree distribution of seed proteins . Once all proteins in the network are scored according to their crosstalk with proteomic seeds , we construct candidate sub-networks as follows: Formally , the set of candidate sub-networks is defined as C ( S ) = {N ( u ) :u∈S}∪{N* ( u ) :u∈S} , where N* ( u ) = {v∈N ( u ) : zS ( v ) >z*} . Here , z* denotes the cut-off for adjusted crosstalk scores to be considered significant . In our experiments , we use z* = 3 . 45 , to reflect a p-value cut-off of 0 . 001 , under the assumption of normally distributed crosstalk scores . For each candidate sub-network Q in C ( S ) , we quantify the synergistic expression of the proteins in Q using an information-theoretic scheme developed by Chuang et al . [11] . Namely , for protein v∈V , let e ( v ) denote the properly normalized m-dimensional mRNA expression vector , provided by genome-scale transcriptomic screening of m disease and control samples . Let c denote an m-dimensional binary vector indicating the phenotype class of each sample , such that c ( i ) = 1 if the ith sample is diagnosed with the disease , 0 otherwise . Furthermore , define the aggregate expression vector e ( Q ) for the sub-network induced by set of proteins Q as ( 3 ) Then , the synergistic differential expression φ ( Q ) of the genes coding for proteins in Q with respect to the phenotype of interest is given by the mutual information between e ( Q ) and c , i . e . , ( 4 ) Here , e ( Q ) denotes a discrete-valued vector obtained by quantizing e ( Q ) into k bins , H ( x ) denotes the entropy of a discrete-valued vector x over a finite alphabet A , i . e . , H ( x ) = Σa∈A−p ( a ) log ( p ( a ) ) , and p ( a ) = |{i:x ( i ) = a}|/m ( in the context of our problem , A represents the set of bins ) . In this paper , we use k = 6 , since this value of k was found to provide reasonable estimates for mutual information in our experiments . Finally , we assess the statistical significance of synergistic differential expression for each candidate sub-network . In order to do so , for a given Q∈C ( S ) , we generate a null distribution for synergistic differential expression of sub-networks that reflect the topological properties of Q . Since Q is composed of proteins that are connected to each other via a single protein ( that is , the corresponding proteomic seed ) , the null distribution should also be derived from sub-networks that consist of the same number of proteins in Q , which are connected to each other through a single protein in the network . Therefore , we first construct a bag D of proteins in the network with degree at least |Q| , i . e , D = {v∈V:|N ( v ) |≥|Q|} . Subsequently , we choose a protein v from D uniformly at random . Finally , we choose |Q| proteins uniformly at random from N ( v ) to construct a random instance Q ( i ) representative of Q . Repeating this procedure n times ( in our experiments , we use n = 1000 ) and computing φ ( Q ( i ) ) , we obtain a null distribution of synergistic differential expression for sub-networks similar to Q . Observe that , only the size of Q ( i ) depends on Q in this procedure . For this reason , in our experiments , we do not explicitly generate a null distribution for each Q∈C ( S ) . Rather , we generate a null distribution for sub-networks of size 2 , 4 , 8 , 16 , 32 , 64 . Then we interpolate the mean and standard deviation of synergistic differential expression for these distributions , to obtain a curve that characterizes the behavior of synergistic differential expression with respect to sub-network size . In order to assess the reproducibility of discovered subnetworks across different data sets and evaluate the potential of the proposed framework for feature selection in classification of CRC , we perform cross-classification experiments . In these experiments , we use the aggregate expression profiles ( e ( Q ) ) of crosstalker and interactor subnetworks associated with Nibbe and CAN seeds as features for classification . For this purpose , in each experiment , we select the crosstalker ( or interactor ) subnetworks with synergistic differential expression ( φ ( Q ) ) one standard deviation above random mean , according to a specific mRNA expression data set ( e . g . , GSE8671 ) . Assume that there are K such subnetworks . Then , for each k≤K , we use the k subnetworks with maximum φ ( Q ) to train an SVM classifier on the same data set ( GSE8671 ) , using Matlab's svmtrain function . Subsequently , we use this classifier to predict the class ( tumor vs . normal ) of each sample on a different data set ( e . g . , GSE10950 ) , using Matlab's svmclassify function . We evaluate the performance of the classifier using the harmonic mean of precision ( selectivity ) and recall ( sensitivity ) , known as the F-measure , defined asHere , precision is the fraction of true positives among all samples classified as tumor and recall is the fraction of tumor samples called accurately by the classifier among all tumor samples .
|
Intensive research on cancer has led to an understanding of many individual genes that may be important for the initiation and progression of tumors . However , since cancer is a progressive disease that results from accumulation of multiple mutations likely acting in concert , individual markers can only provide limited insights into cellular mechanisms that underlie tumorigenesis . For this reason , recent studies focus on identification of “sub-network markers” , that is , functionally associated genes that exhibit coordinate changes in molecular expression during cancer progression . However , expression of genes is most frequently interrogated at the mRNA level , which captures functional activity of genes only to a limited extent . Screening of protein expression , on the other hand , provides information on the abundance of functional gene products , but its scale is often limited compared to screening of mRNA expression . In this article , we develop a proteomics-driven computational method that searches for sub-network markers in human colorectal cancer , based on a seed of differentially expressed proteins identified by proteomic screening . Our results show that significant changes in the expression of these proteins is likely to be associated with coordinate changes in the expression of the genes whose products are functionally associated with these proteins . This analysis leads to novel insights in the synergistic processes that underlie tumorigenesis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"computational",
"biology/systems",
"biology",
"oncology/gastrointestinal",
"cancers",
"biotechnology/protein",
"chemistry",
"and",
"proteomics"
] |
2010
|
An Integrative -omics Approach to Identify Functional Sub-Networks in Human Colorectal Cancer
|
Visceral leishmaniasis ( VL ) is a vector-borne disease affecting humans and domestic animals that constitutes a serious public health problem in many countries . Although many antigens have been examined so far as protein- or DNA-based vaccines , none of them conferred complete long-term protection . The use of the lizard non-pathogenic to humans Leishmania ( L . ) tarentolae species as a live vaccine vector to deliver specific Leishmania antigens is a recent approach that needs to be explored further . In this study , we evaluated the effectiveness of live vaccination in protecting BALB/c mice against L . infantum infection using prime-boost regimens , namely Live/Live and DNA/Live . As a live vaccine , we used recombinant L . tarentolae expressing the L . donovani A2 antigen along with cysteine proteinases ( CPA and CPB without its unusual C-terminal extension ( CPB-CTE ) ) as a tri-fusion gene . For DNA priming , the tri-fusion gene was encoded in pcDNA formulated with cationic solid lipid nanoparticles ( cSLN ) acting as an adjuvant . At different time points post-challenge , parasite burden and histopathological changes as well as humoral and cellular immune responses were assessed . Our results showed that immunization with both prime-boost A2-CPA-CPB-CTE-recombinant L . tarentolae protects BALB/c mice against L . infantum challenge . This protective immunity is associated with a Th1-type immune response due to high levels of IFN-γ production prior and after challenge and with lower levels of IL-10 production after challenge , leading to a significantly higher IFN-γ/IL-10 ratio compared to the control groups . Moreover , this immunization elicited high IgG1 and IgG2a humoral immune responses . Protection in mice was also correlated with a high nitric oxide production and low parasite burden . Altogether , these results indicate the promise of the A2-CPA-CPB-CTE-recombinant L . tarentolae as a safe live vaccine candidate against VL .
Leishmaniasis is a vector-borne disease caused by different Leishmania species that ranges from self-limiting cutaneous leishmaniasis to fatal visceral leishmaniasis ( VL ) and is endemic in 88 tropical and subtropical countries [1] . VL is caused by members of the L . donovani complex with a wide and growing prevalence and incidence ( http://www . who . int/en/ ) [2] and is considered as the most severe form of leishmaniasis and is often fatal , if left untreated [3] , [4] . L . infantum is responsible for VL in the Mediterranean basin , extending to several Middle East and Asian countries [5] . VL has emerged as an opportunistic infection in HIV-1 infected patients in many parts of the world [6] , [7] , [8] . Currently , prophylactic or therapeutic vaccines are not available and the control of the disease depends exclusively on chemotherapy . However , drug-resistant forms have developed from current chemotherapeutic interventions [9] , [10] . Therefore , much attention has been given to improve vaccination strategies . Although induction of lifelong protection against reinfection in recovered individuals demonstrates that a protective vaccine can be achieved , an effective vaccine against human leishmaniasis has not been yet developed [11] . Various vaccination strategies have been explored against experimental leishmaniasis , with particular emphasis on their efficacy against CL rather than VL [12] , [13] . First-generation anti-leishmanial vaccines based on live parasites ( leishmanization ) are the only successful intervention against leishmaniasis [14] , [15] . However , leishmanization was largely abandoned due to safety issues . The development of second-generation vaccines for Leishmania included recombinant proteins , polyproteins , DNA vaccines , liposomal formulation , and dendritic cell vaccine delivery systems [16] , [17] , [18] . Also , multicomponent vaccines have been shown to protect against VL in experimental infection systems [17] , [19] , [20] , [21] , [22] , [23] . Furthermore , it has been reported that persistence of a small number of live parasites is essential for maintaining durable immunity [24] , [25] . The only way to meet this requirement is by using attenuated live vaccines . Attenuated strains based either on long-term in vitro culturing [26] or culturing under drug pressure [27] or on selection for temperature sensitivity [28] and chemical mutagenesis [29] are not easily applicable to human use because there is always a risk of reversion of the organism to its virulent state . Alternatively , approaches based on the genetic attenuation of Leishmania genes encoding virulence factors or enzymes responsible for their synthesis and genes essential for intracellular survival have been reported [30] , [31] , [32] , [33] , [34] , [35] , [36] . Other approaches to develop live attenuated parasites as VL vaccines have utilized nonpathogenic Leishmania species , an approach comparable to the use of BCG as a vaccine against Mycobacterium tuberculosis infections . Among different species of Leishmania , the lizard protozoan parasite L . tarentolae has never been found associated with any form of human leishmaniasis and is therefore considered nonpathogenic [37] . Whilst L . tarentolae is capable of infecting mammalian cells and transforming into amastigote-like forms , it is not however able to persist long enough within macrophages [38] , [39] . The use of L . tarentolae as a vaccine vector to deliver specific Leishmania antigens mimicking live infection has also been explored . In a previous study , the L . donovani amastigote-specific A2 antigen was expressed in L . tarentolae , which lacks this protein [40] , [41] and used as a vaccine strain in an experimental mouse model . Vaccination protected susceptible mice against L . infantum challenge and was associated with the production of high levels of IFN-γ production [42] . It has been reported that sera from either cured or active cases of cutaneous and visceral leishmaniasis patients recognize the recombinant cysteine proteinases CPA ( rCPA ) and CPB ( rCPB ) of L . major and L . infantum [43] , [44] that are members of the papain superfamily [45] . CPA ( type II ) and CPB ( type I ) are expressed at higher levels in amastigotes [45] and stationary-phase promastigotes [46] . An unusual C-terminal extension ( CTE ) of 110 amino acids distinguishes CPB enzyme from the other CPs in the papain superfamily [47] , [48] . Immunization with CTE also displayed both type 1 and 2 immune signatures in experimental murine model of L . infantum infection and therefore is not protective as a vaccine candidate [49] . Furthermore , we have demonstrated that the combination of CPA/CPB and CPA/CPB-CTE is more protective against visceral and cutaneous leishmanial infections than the individual forms [50] , [51] , [52] , [53] , [54] . Despite the proven antigenicity and immunogenicity of these DNA vaccine candidates , their largest drawback is the inefficient intracellular delivery of pDNA causing low levels of gene expression , which in turn limits the resulting immune responses [55] . In our previous studies , cationic solid-lipid nanoparticles ( cSLN ) as an effective delivery system has exhibited considerable low cytotoxicity , and it was able to protect pDNA in a DNase I challenge assay [54] , [56] . Here , we use the A2-CPA-CPB-CTE tri-gene fusion as a DNA vaccine formulated with cSLN and also a recombinant L . tarentolae expressing the tri-gene fusion as a live vaccination strategy against visceral leishmaniasis in two-modalities , namely DNA/Live and Live/Live vaccination in BALB/c mice . We demonstrate that prime-boost strategies harboring recombinant L . tarentolae-based vaccines represent a promising immunization approach against Leishmania infections .
All solutions were prepared using MilliQ ultrapure ( Milli-QSystem , Millipore , Molsheim , France ) and apyrogenic water to avoid surface-active impurities . Cetyl palmitate , Tween 80 and cholesterol were purchased from Merck ( Darmstadt , Germany ) . G418 , N-[1- ( 2 , 3-Dioleoyloxy ) propyl]-N , N , Ntrimethylammonium chloride ( DOTAP ) , Sodium dodecyl sulfate ( SDS ) were purchased from Sigma-Aldrich ( Sigma , Deisenhofen , Germany ) . The materials applied for PCR , enzymatic digestion and agarose gel electrophoresis were acquired from Roche Applied Sciences ( Mannheim , Germany ) . Cell culture reagents including M199 medium , HEPES , L-glutamine , adenosine , hemin , gentamicin , RPMI and Schneider were sourced from Sigma ( Darmstadt , Germany ) and Gibco ( Gibco , Life Technologies GmbH , Karlsruhe , Germany ) , respectively . Fetal Calf Sera ( FCS ) was purchased from ( Gibco , Life Technologies GmbH , Karlsruhe , Germany ) . The A2 gene ( with Kozak sequence ) was digested from pUC57 vector ( synthesized by Shine Gene Molecular Biotech , Inc ) with EcoRI and HindIII restriction sites . After sequence confirmation , the A2 fragment was subcloned into the EcoRI and HindIII sites of vector pGEM7zf ( + ) ( Promega ) . The CPA fragment was amplified from pGEM-CPA using Taq DNA Polymerase ( Roche , Germany ) and the following primers: ( forward , 5′-GTTAAGCTTCGCCCCCAGTGGTGT-3′ ) including HindIII restriction site ( underlined ) ; and ( reverse , 5′-TTTGCTAGCCTAGGCCGTTGTCGT-3′ ) including NheI restriction site ( underlined ) . Then , the PCR-amplified CPA gene was cloned into HindIII and NheI sites of pAT153 vector ( Boca Scientific ) . The CPB-CTE fragment was amplified from pGEM-CPB using Taq DNA Polymerase ( Roche , Germany ) and the following primers: ( forward , 5′-AATGCTAGCGATGCGGTGGACTGG-3′ ) harboring NheI restriction site ( underlined ) ; and ( reverse , 5′-ACTGGATCCCACATGCGCGGA-3′ ) including BamHI restriction site ( underlined ) . The PCR-amplified CPB-CTE gene was cloned into the NheI and BamHI sites of pAT153 vector downstream the CPA gene . Then CPA-CPB-CTE fusion gene was digested with HindIII and BamHI restriction sites and cloned downstream of the A2 gene in the pGEM7zf ( + ) vector . After sequence confirmation , the A2-CPA-CPB-CTE fusion gene was subcloned into the EcoRI and BamHI sites of vector pEGFP-N3 upstream of the GFP gene to generate pEGFP-A2-CPA-CPB-CTE , and the correct insert orientation was confirmed by restriction analysis . Then , the A2-CPA-CPB-CTE-GFP fragment was subcloned into the BglII and NotI sites of vector pLEXSY-NEO2 ( EGE-233 , Jena Bioscience , Germany ) to generate pLEXSY-A2-CPA-CPB-CTE-GFP . Also the A2-CPA-CPB-CTE fragment was subcloned into the EcoRI and BamHI sites of vector pcDNA3 . 1 ( − ) ( Invitrogen , Germany ) to generate pcDNA-A2-CPA-CPB-CTE as a DNA vaccine . pcDNA-A2-CPA-CPB-CTE plasmid was purified by ion exchange chromatography with Endofree Mega kit ( QIAGEN , Germany ) . The L . tarentolae Tar II ( ATCC 30267 ) strain was grown at pH 7 . 2 and 26°C in M199 medium ( Sigma ) supplemented with 5% heat-inactivated fetal calf serum ( FCS , Gibco ) , 40 mM HEPES , 0 . 1 mM adenosine , 5 µg/ml hemin and 50 µg/ml gentamicin . For transfection , 4×107 log-phase parasites were washed and re-suspended in 300 µl of electroporation buffer ( 21 mM HEPES , 137 mM NaCl , 5 mM KCl , 0 . 7 mM Na2HPO4 , 6 mM glucose; pH 7 . 5 ) and mixed with 50 µl H2O containing 15 µg of linearized pLEXSY-A2-CPA-CPB-CTE-GFP with SwaI , stored on ice for 10 min , and electroporated ( Bio-Rad Gene Pulser Ecell , Germany ) at 450 V and 500 µF as described previously [57] . In brief , the electroporated promastigotes were then incubated for 24 h in M199 10% medium at 26°C without any drug ( Neomycin or G418 , Gibco , Germany ) , and plated on solid media ( 2% of Noble agar and 2XM199 10% ( vol/vol ) , Sigma , Germany ) containing 50 µg/ml of G418 . The growth of cells highly resistant to Neomycin was observed after 7–10 days . Clones were selected on Noble agar plates and further propagated in liquid M199 10% medium in the absence of G418 . Expression of EGFP in Leishmania promastigotes was evaluated by Epi-fluorescent microscopy for up to 3 months ( Nikon , E 200 , ACT-1 software , Digital sight Camera , Japan ) . Integration of the expression cassette into the ssu locus was confirmed by diagnostic PCR using genomic DNA of transgenic strains as a template extracted by GF-1 Genomic DNA extraction kit ( Vivantis , Malaysia ) . We performed diagnostic PCR ( annealing temperature 60°C ) with ssu forward primer F3001 ( 5′-GATCTGGTTGATTCTGCCAGTAG-3′ ) and reverse primer A1715 hybridizing within the 5′UTR of the target gene ( 5′-TATTCGTTGTCAGATGGCGCAC-3′ ) according to the LEXSY Kit protocol ( Jena bioscience , Germany ) . Primer pairs including one primer hybridizing within the expression cassette and one primer hybridizing to the ssu sequence not present in the plasmid were used . Integration of the expression cassette into the ssu locus yielded a 1 kb fragment that was not obtained in the control reactions with the genomic DNA of L . tarentolae wild type . For Southern blot analysis , 5 µg of transgenic L . tarentolae-GFP and wild type L . tarentolae genomic DNA were digested with the appropriate restriction enzymes ( BglII/NotI ) . DNA was then resolved on 0 . 7% agarose before being separated and transferred onto membrane according to standard procedures [58] . The membrane was then UV-crosslinked prehybridized with Church mix buffer ( 7% SDS , 0 . 5 M NaPi , 1 mM EDTA , 1% BSA ) for 1 to 2 h . For probe synthesis , 100 ng of the GFP ORF was used by incorporating radiolabeled dCTP using Klenow enzyme . The reaction was finished by addition of 1 µl EDTA ( 0 . 5 M ) and subsequently , the probe was purified by passing through a Sephadex resin . The membrane was then washed once with Wash buffer 1 ( 2× SSC , 0 . 5% SDS ) at 25°C for 30 min , then 2 times at 65°C with Wash buffer 2 ( 1× SSC , 0 . 1% SDS ) for 15 min . The membrane was then exposed for overnight and developed by a Konicka Minolta developer . RNA samples were extracted from promastigote forms using RNeasy kit ( Qiagen ) and treated with RNase-free DNase for 30 min at 37°C to eliminate any remaining DNA . cDNA synthesis was performed using the Qiagen Omniscript RT Kit from 1 µg of RNA . To detect the A2 , CPA , CPB-CTE and GFP cDNAs , PCR reactions were carried out using specific primer pairs to amplify each gene separately . Promastigote forms of L . tarentolae-A2-CPA-CPB-CTE-GFP were examined for GFP expression by Epi-fluorescence microscopy . Promastigotes were centrifuged in 3000 rpm for 15 min and after washing once with PBS , cells were re-suspended in PBS and mounted on microscope slides . Expression of EGFP protein was evaluated by Epi-fluorescent microscopy ( Nikon , E 200 , ACT-1 software , Digital sight Camera , Japan ) . Also wild type ( as negative control ) and GFP expressing promastigote forms were analyzed for EGFP expression using flow cytometry . Parasites at two different growth phases ( logarithmic and stationary phases ) were centrifuged at 3000 rpm for 15 min , washed and then re-suspended at 106 cell/ml in PBS and stored on ice . Cells were analyzed on a FACS caliber flow cytometer ( BD: Becton Dickinson , Franklin Lakes , NJ ) equipped with a 15 mV , 488 nm , air-cooled argon ion laser . 50 , 000 events were recorded and EGFP expression in transgenic L . tarentolae was measured in comparison with wild type ( WT ) parasites and L . tarentolae-GFP expressing parasites . Promastigote forms of L . tarentolae-A2-CPA-CPB-CTE-GFP were harvested by centrifugation at 3000 rpm for 15 min and washed in PBS . The pellets were immediately lysed in 2× SDS-PAGE sample buffer ( 4 . 5 mM Tris-HCl , pH 6 . 8 , 10% , v/v glycerol , 2% , w/v SDS , 5% , v/v 2-mercaptoethanol , 0 . 05% , w/v bromophenol blue ) on ice and then boiled for 5 min . Samples were then loaded on a 15% SDS-PAGE . The gels were transferred onto a nitrocellulose membrane using a Bio-Rad wet blotting system and incubated with blocking solution ( PBS with 0 . 1% Tween 20 and 2 . 5% BSA ) for 1 h . Washing was performed 3 times with 0 . 1% Tween 20 in PBS , and blots were incubated overnight with previously prepared rabbit anti-CPB polyclonal antibodies [43] , [52] as the first antibody at 1∶50 dilution . The membranes were washed three times and incubated for 90 min with peroxidase-conjugated goat anti mouse IgG ( 1∶5000 , Sigma ) assecondary antibodies . Unbound secondary antibodies were removed by washing as described above . Diaminobenzidine tetrahydrochloride ( DAB , Sigma ) were used as the substrate to detect the desired bands of the protein . Six-week-old female BALB/c mice were obtained from the breeding stock maintained at the Pasteur Institute of Iran . The L . infantum strain MCAN/ES/98/LLM-877 was kindly provided by WHO collaborating center for leishmaniasis , Servicio de Parasitología , Centro Nacional de Microbiología , Instituto de Salud Carlos III , Madrid , Spain and kept virulent by continuous passage in hamsters . Amastigotes were isolated from the spleen of infected hamsters and cultured in NNN media in presence of 100 µg/ml of gentamicin . Stationary-phase promastigotes were harvested after 5–6 days by centrifugation ( 270×g , 5 min , 4°C ) , washed three times in PBS ( 8 mM Na2HPO4 , 1 . 75 mM KH2PO4 , 0 . 25 mM KCl and 137 mM NaCl ) and re-suspended at a concentration of 2×108 parasites/ml . This preparation was frozen and thawed ( F/T ) 10 times using liquid N2 and a 37°C water-bath and protein concentration was determined by bicinchoninic acid reagent ( BCA , PIERCE , Chemical Co . Rochford III ) . For infection , virulent promastigotes were harvested in the stationary phase , washed in PBS and injected ( 107 ) by the lateral tail vein into BALB/c mice . All mouse experiments including maintenance , animals' handling program and blood sample collection were approved by Institutional Animal Care and Research Advisory Committee of Pasteur Institute of Iran ( Education Office dated January , 2008 ) , based on the Specific National Ethical Guidelines for Biomedical Research issued by the Research and Technology Deputy of Ministry of Health and Medicinal Education ( MOHME ) of Iran ( issued in 2005 ) . Two independent immunization experiments were carried out in five groups of mice ( n = 15 per treatment at each time point ) and all tests were done in duplicate or triplicate ( number of mice per group/time point n = 2–3 ) . Results are shown as mean±S . E . of measures obtained from 4–6 mice in different groups . Group 1 ( DNA cSLN/Live ) immunized with pcDNA-A2-CPA-CPB-CTE-cSLN ( 50 µg of pcDNA-A2-CPA-CPB-CTE formulated by cSLN nanoparticles as a chemical delivery as previously described [59] as a prime and with 2×107 recombinant L . tarentolae A2-CPA-CPB-CTE as a boost; group 2 ( L . tarentolae Live A2-CPA-CPB-CTE/L . tarentolae Live A2-CPA-CPB-CTE ) vaccinated with 2×107 recombinant L . tarentolae-A2-CPA-CPB-CTE as prime and boost; group 3 ( PBS as a control ) ; group 4 [ ( empty vector pcDNA-cSLN ( prime ) /Live L . tarentolae wild type ( boost ) as a control ) ]; and group 5 ( L . tarentolae Live/L . tarentolae Live ) vaccinated with 2×107L . tarentolae wild type as prime and boost and used as a control . All groups were immunized via footpad . Booster immunization was carried out 3 weeks following the prime immunization . Three weeks after the last immunization , all animals were challenged with 107 stationary phase L . infantum promastigotes by lateral tail vein . Serum samples were analyzed by ELISA for specific antibodies including IgG1 and IgG2a against either rA2 , rCPs or Leishmania F/T at two different time points: before challenge and 5 weeks after challenge . Briefly , 96-well plates ( Greiner ) were coated with rA2 , rCPA and rCPB or L . tarentolae-A2-CPA-CPB-CTE or L . infantum F/T , all at10 µg/ml , overnight at 4°C . Plates were blocked with 100 µl of 1% BSA in PBS at 37°C for 2 h to prevent nonspecific binding . Sera were added ( with dilution of 1∶100 ) and incubated 2 h at 37°C . After three washes , Goat Anti-Mouse IgG1-HPR ( 1∶10 , 000 , Southern Biotech , Canada ) or Goat Anti-Mouse IgG2a -HPR ( 1∶10 , 000 , Southern Biotech , Canada ) were added and incubated for 2 h at 37°C . After four washes , plates were incubated for 30 min at 37°C with Peroxidase Substrate System ( KPL , ABTS ) as substrate . Reactions were stopped with 1% SDS and the absorbance was measured at 405 nm . Three mice from each group were sacrificed before challenge and also at 4 and 8 weeks after challenge and spleens were homogenized . After red blood cell lysis using ACK lysis buffer ( 0 . 15 M NH4Cl , 10 mM KHCO3 and 0 . 1 mM Na2EDTA ) , splenocytes were washed and re-suspended in complete RPMI medium ( supplemented with 5% FCS , 1% l-glutamine , 1% HEPES , 0 . 1% 2ME , 0 . 1% gentamicin ) . Cells were then seeded at a density of 3 . 5×106 cells/ml in the presence of rA2 ( 10 µg/ml ) , rCPA ( 10 µg/ml ) and rCPB ( 10 µg/ml ) , or L . infantum F/T ( 25 µg/ml ) , or L . tarentolae-A2-CPA-CPB-CTE F/T ( 25 µg/ml ) , or medium alone . Concanavalin A ( Con A; 5 µg/ml ) was also used in all experiments as the positive control . Plates were incubated for 24 h for IL-2 measurement and 5 days for IFN-γ and IL-10 measurement at 37°C in 5% CO2 humidified atmosphere . The IL-2 , IFN-γ and IL-10 production in supernatants of splenocyte cultures was measured by sandwich ELISA kits ( R&D , Minneapolis , MN , USA ) , according to the manufacturer's instructions . The minimum detectable dose of mouse IFN-γ , IL-10 and IL-2 is typically less than 2 , 4 and 7 pg/mL , respectively . All experiments were run in duplicates . Nitrite release was determined in 5-day stimulated splenocytes supernatant at 8 weeks after challenge . In this case , as described in Section 2 . 13 , 100 µl of 5-day incubated culture splenocytes supernatant was collected from each well and subsequently mixed with an equal volume of Griess reagent [0 . 1 N ( 1-napthyl ) ethylenediamine dihydrochloride , 1% sulfanil amide in 5% H3PO4] incubated 10 min at RT . Absorbance of the colored complex was determined at 550 nm . The nitric oxide concentration of each corresponding sample was extrapolated from the standard curve plotted with sodium nitrite serial dilution in culture medium . Two mice from each group were sacrificed at 2 , 4 , 8 and 12 weeks after challenge and parasite burden was determined as follows . A piece of spleen and liver was excised , weighed and then homogenized with a tissue grinder in 2 ml of Schneider's Drosophila medium ( Sigma , Germany ) supplemented with 20% heat-inactivated fetal calf serum and gentamicin ( 0 . 1% ) . Under sterile conditions , serial dilutions ranging from 1 to 10−20 were prepared in wells of 96 well microtitration plates . After 7 and 14 days of incubation at 26°C , plates were examined with an inverted microscope at a magnification of 40× . The presence or absence of mobile promastigotes was recorded in each well . The final titer was the last dilution for which the well contained at least one motile parasite . The number of parasites per gram was calculated in the following way: parasite burden = −log10 ( parasite dilution/tissue weight ) [60] , [61] . Furthermore , real time PCR was used to quantify parasite burden in spleen and liver 4 weeks after challenge for vaccinated groups ( G1 and G2 ) and PBS control group ( G3 ) . Two mice from each group were sacrificed and genomic DNA was extracted from 10 mg of spleen and 30 mg of liver tissues using DNeasy Blood & Tissue kit ( Qiagen ) . Two set of primers targeting a region of kinetoplastid minicircle DNA of L . infantum named as RV1 and RV2 ( forward: 5′-CTTTTCTGGTCCCGCGGGTAGG-3′; reverse: 5′-CCACCTGGCCTATTTTACACCA-3′ ) were used [62] . Absolute copy number of the target sequence was measured using Applied Biosystem 7500 real time PCR system . L . infantum genomic DNA was used in 10-fold dilutions corresponding to 2×105 parasites and used in real time PCR to draw the standard curve . For quantification of parasites in tissues , 300 ng of DNA was subjected to the reaction containing 5 pmol of each forward and reverse primers , 12 . 5 µl Qiagen QuantiFast SYBR Green Master Mix in total volume of 25 µl . All reactions were performed in duplicate . Conditions for PCR amplification were as follows: 95°C for 10 min; 40 cycles consisting of 95°C for 15 s , 58°C for 30 s , and 72°C for 40 s . Specific amplification of the target region was confirmed by gel electrophoresis of the PCR products . Liver , spleen and bone marrow tissues of two animals from each group at 4 , 8 , and 14 weeks after challenge were fixed by 10% neutral-buffered formalin for 24 h , dehydrated by immersion in increasing concentrations of ethanol ( 70% , 95% , and then 100% ) and then xylene was added before embedding in paraffin wax . Four-micrometer-thick slides were prepared from paraffin blocks and were stained with hematoxylin and eosin ( H&E ) method . The slides were examined with an Olympus microscope ( BX41 ) and photos were prepared by a DP11 digital camera ( Olympus ) . The slides were reviewed by a pathologist who was not aware of the original treatment of the groups . The parenchyma of the liver was assessed for hepatocyte damage including fatty change , hydropic changes , cholestasis , liver cell necrosis and regenerative changes . The portals were assessed for degree of inflammation and interface hepatitis . Both areas were evaluated for presence of granuloma , neutrophils , plasma cells and lymphocytes as well as intracellular parasites . The micro-architecture of spleen composed of lymphoid follicles , splenic cord , and parafollicular area were evaluated . Presence of granuloma , giant cells and neutrophils were also assessed . Statistics were performed using Graph-Pad Prism 5 . 0 for Windows ( Graphpad Software Inc 2007 , San Diego , Calif . , USA ) as well as SPSS version 18 . All the data were analyzed with one way ANOVA ( Multiple-comparison Tukey post Hoc test ) and when required with a Student's t-test . The correlation between the ratio of IFN-γ/IL-10 production and differences in parasite burden at weeks 4 and 8 were calculated using Spearman Correlation method ( 2 tailed ) . The area under the curve ( AUC ) of parasite burden graphs was calculated using Graph-Pad Prism 5 . 0 software program . The results were considered statistically significant when p<0 . 05 .
Recombinant L . tarentolae stably expressing the L . donovani A2 antigen along with L . infantum cysteine proteinases CPA and CPB without its unusual C-terminal extension ( CPB-CTE ) as a tri-fusion gene ( A2-CPA-CPB-CTE ) together with the EGFP gene were generated by introducing the linearized pLEXSY-A2-CPA-CPB-CTE-EGFP vector into the 18S rRNA ssu locus of L . tarentolae ( Fig . S1A ) , as indicated in Materials and Methods . Specific targeting of the expression cassette into the ssu locus was confirmed both by genomic PCR ( Fig . S1B ) and by Southern blot analysis using an EGFP-specific probe ( Fig . S1C ) . Amplification of each gene from cDNA resulted in a PCR product of the expected size , hence confirming the expression of A2-CPA-CPB-CTE-EGFP by L . tarentolae at the mRNA level ( Fig . S1D ) . Expression of EGFP in L . tarentolae-A2-CPA-CPB-CTE-EGFP and L . tarentolae-EGFP parasites was confirmed by fluorescence microscopy ( Fig . 1A ) and also by fluorescence-activated cell sorting ( FACS ) analysis ( Fig . 1B ) . The percentage of L . tarentolae-A2-CPA-CPB-CTE-EGFP-expressing parasites was decreased in logarithmic phase ( 72% ) as compared to L . tarentolae-EGFP ( 93% ) with decreasing in mean fluorescent intensities ( MFI , Fig . 1B ) that may be due to the addition of the tri-fusion genes before EGFP gene which possibly alters translation of EGFP . The A2-CPA-CPB-CTE-EGFP expression was assessed by western blot analysis . As shown in Figure 1C , an immunoreactive band of 102 . 56 kDa was detected in L . tarentolae transgenic parasites using an anti-CPB antibody . Altogether , these results are consistent both with a proper integration of the fused A2-CPA-CPB-CTE-EGFP gene into L . tarentolae and with its constitutive expression . We first examined whether immunization with the A2-CPA-CPB-CTE recombinant L . tarentolae was protective against L . infantum infectious challenge . For DNA vaccination with rA2-CPA-CPB-CTE , the pcDNA-A2-CPA-CPB-CTE was formulated into cationic lipid particles with nanometer range ( ∼240–250 nm ) . It was shown previously that this formulation facilitates the uptake by dendritic cells and macrophages , thereby enhancing antigen expression , processing and presentation and resulting in stronger immune effects [59] . Five groups of mice were considered for immunization with two subsequent repeats as described in Materials and Methods . The results are shown as mean±S . E . of measures obtained from these two independent experiments . The degree of protection against infection was determined by weekly measurement of the parasite burden in the spleen and the liver at 2 , 4 , 8 and 12 weeks post-challenge and by comparing the area under the curves ( AUC ) , as explained in the statistical analysis section in Materials and Methods . As shown in Figure 2 , immunization using a DNA A2-CPA-CPB-CTE/Live L . tar A2-CPA-CPB-CTE ( G1 ) and L . tar A2-CPA-CPB-CTE/L . tar A2-CPA-CPB-CTE ( G2 ) prime-boosting regiments drastically ( p<0 . 01 ) reduced the infection levels in both the liver ( Fig . 2A and 2B ) and the spleen ( Fig . 2D and 2E ) at 4 weeks after L . infantum challenge in contrast to the control groups ( G3 to G5 , n = 4 at each point/organ ) . The liver parasite load ( Fig . 2A and 2B ) of all control groups increased early following infection , reaching its maximum at 4 weeks after challenge to rapidly decline . However , the parasite burden in the vaccinated groups G1 and G2 peaked with a 4 week delay compared to the control groups . We observed a minor increase in the parasite burden in the spleen of G1 and G2 groups at week 8 but this level remained stable up to 12 weeks after infection ( Fig . 2D and 2E ) . In the liver , there was no significant difference between all groups at 12 weeks post-infection which resulted in the no significant difference in the liver AUC of DNA/Live ( G1 ) and Live/Live ( G2 ) regimens with their related controls G4 and G5 , respectively ( Fig . 2C ) . Carrion et al . demonstrated that after intravenous injection of BALB/c mice with 103 , 105 or 106 promastigotes of L . infantum , VL infection would be established but the development of quantifiable immunohistological features like parasite persistence were dependent on the inoculum size [63] . According to Carrion et al . herein control of the hepatic infection in all groups did not result into complete clearance of the parasite in the liver ( at week 12 there were still few detectable parasites present ) due most likely to high inoculum ( 107 ) . In the spleen , the highest parasite burden in control groups ( G3 , G4 and G5 ) was observed at 12 weeks after challenge and the organ remained chronically infected ( Fig . 2D and 2E ) . Interestingly , only the vaccinated groups G1 and especially G2 , which was immunized with Live L . tar A2-CPA-CPB-CTE/Live L . tar A2-CPA-CPB-CTE vaccine regimen , were able to control the infection as illustrated by the significant differences observed at 12 weeks after challenge in comparison to the controls ( G1 with G4 and G2 with G3 and G5 ) . Thus , the G2 vaccination regimen protects against infection better than G5 ( Live L . tar/Live L . tar ) . Also , the spleen AUC of DNA/Live ( G1 ) and Live/Live ( G2 ) parasites indicate significant differences with the controls G4 and G5 ( Fig . 2F ) . As shown by real time PCR , the parasite load in G3 is significantly higher than in the two vaccinated groups ( G1 and G2 ) both in the liver and the spleen at 4 weeks after challenge ( Fig . S2A and S2B ) . The above experiments demonstrate that immunization with recombinant L . tarentolae expressing A2-CPA-CPB-CTE confers a significant protection against L . infantum infection in mice . Given that production of IFN-γ is considered an important requirement for protection against L . infantum and that the presence of a small amount of IL-10 could promote the induction of type-1 immunity [64] , we determined the levels of IFN-γ and IL-10 at different times pre- and post-challenge in cultures of splenocytes isolated from immunized and control mice in response to rA2-rCPA-rCPB , F/T L . tar A2-CPA-CPB-CTE-EGFP and F/T L . infantum as recall antigens ( Fig . 3A–C ) . In addition , we analyzed the levels of IL-2 production before challenge and also at 4 and 8 weeks after challenge as well as the concentration of nitrite levels ( NO2 ) only 8 weeks after challenge in the spleen of all five groups following stimulation with rA2-rCPA-rCPB , F/T L . tarA2-CPA-CPB-CTE-EGFP and F/T L . infantum antigens ( n = 6 at each mentioned period Fig . 3D–E ) . As shown in Figure 3A , stimulation of splenocytes isolated from vaccinated groups G1 and G2 prior and after challenge with all three recall antigens ( e . g . rA2-rCPA-rCPB , L . tar A2-CPA-CPB-CTE-EGFP lysate and L . infantum lysate ) elicited a significantly higher IFN-γ production than control groups G3 , G4 and G5 . The production of IFN-γ in response to rA2-rCPA-rCPB antigens reached the highest level ( 524±16 . 4 pg/ml ) at 4 weeks after challenge in G2 ( Live/Live ) . This was significantly ( p<0 . 05 ) different from G1 ( DNA/Live ) ( 404±19 . 9 pg/ml ) . IFN-γ levels were maintained high even 8 weeks post-challenge in both G1 ( 404 . 5±101 pg/ml ) and G2 ( 331±27 pg/ml ) ( Fig . 3A ) . Stimulation with F/T L . tarA2-CPA-CPB-CTE-GFP recall antigen elicited also high IFN-γ production in G1 ( 600±22 . 8 pg/ml ) before challenge but this was not significantly different from G2 ( Fig . 3A ) . IFN-γ levels remained high ( although slightly less than before challenge ) after 4 and 8 weeks post-challenge in G1 but to a lesser extent in G2 ( Fig . 3A ) . The production of IL-10 upon antigen stimulation at 4-week and especially at 8 weeks after challenge was lower in G1 and G2 vaccinated groups in comparison to the control groups ( G3 to G5 ) ( Fig . 3B ) . We further calculated the IFN-γ to IL-10 ratio for each vaccinated group as a clear indicator of successful immunization [33] ( Fig . 3C ) . Indeed , the Leishmania-specific IFN-γ/IL-10 ratios upon stimulation with rA2-rCPA-rCPB and L . tar A2-CPA-CPB-CTE-GFP recall antigens were higher in G1 and G2 compared to the control G3 , G4 and G5 mice both at 4 and 8 weeks after challenge . Moreover , a clear Spearman correlation was observed between the ratio of IFN-γ/IL-10 production and differences in parasite burden in liver ( − . 729* ( p = . 017 ) ) and spleen cells ( − . 640* ( p = . 046 ) ) stimulated with F/T L . infantum at 4th week after challenge . This means that higher amount of IFN-γ/IL-10 ratio resulted in lower parasite burden in both spleen and liver at 4th week after challenge . The IFN-γ/IL-10 ratio was drawn against the parasite burden in both spleen and liver at 4th week after challenge with individual values of each group ( five groups ) in liver and spleen as well as their mean , and the negative slops in both graphs are in concordance with this fact ( Fig . S3A and S3B ) . IL-2 production that is important for lymphocyte proliferation was higher in G1 and G2 than in control groups ( Fig . 3D ) . Significant differences were seen in the amount of IL-2 production for G2 at all time points in response to different recall antigens . Similar levels of cytokines were produced following exposure to ConA in all groups ( data not shown ) . The highest amount of nitric oxide , which is essential for killing parasites inside infected macrophages , was observed with the Live/Live modality vaccinated mice ( G2 ) upon stimulation with rA2-rCPA-rCPB , F/T L . tarA2-CPA-CPB-CTE-GFP and F/T L . infantum antigens at 8 weeks after challenge ( Fig . 3E ) . To compare IgG isotypes in different groups , all sera were assayed by ELISA before ( Fig . 4A–B ) and 5 weeks after ( Fig . 4C ) challenge . As shown in Figures 4A and 4C , both rA2-rCPA-rCPB and F/T L . infantum specific IgG1 and IgG2a were higher in vaccinated groups G1 and G2 compared to the control groups . Although G1 shows higher amount of rA2-rCPA-rCPB specific IgG2a , there were no significant differences with control groups . Also , increased amount of F/T L . infantum specific IgG1 was seen in G5 at 5 weeks after challenge . Interestingly , in G2 vaccinated mice with Live L . tar A2-CPA-CPB-CTE-GFP/Live L . tar A2-CPA-CPB-CTE-GFP , a higher amount of F/T L . tar A2-CPA-CPB-CTE-GFP specific IgG1 and IgG2a was detected in comparison to G1 and G4 vaccinated with DNA/Live modality ( Fig . 4B ) . Two animals from all groups at each round of experiment were sacrificed at specific time points ( 4 , 8 , and 14 weeks ) after challenge to collect liver , spleen and bone marrow samples for histopathological analysis . In spleen , splenic cords were normal and somewhat thickened in vaccinated groups but were thin and in some cases were disappeared in other groups . Normal follicle formation of the spleen was noted in G1 and G2 in 4th and 8th week but the normal follicles were disappeared in the other groups ( G3 , G4 and G5 ) ( Fig . 5A ) . Absence of follicles and derangement of splenic cords disfigured the total architecture of the spleen . In total of 18 non-vaccinated mice ( G3 , G4 and G5 ) during 4 , 8 and 14 weeks after challenge , the architecture was completely distorted in 12 , minimally changed in 3 and unchanged in 3 mice . However , in vaccinated groups ( G1 and G2 , total of 12 mice ) , the architecture was relatively unchanged in 9 and minimally changed in 3 mice ( Fig . 5A ) . Spleen granuloma formation was noted only in 3 mice of non-vaccinated cases in 4th week and not seen later . We could not find parasites in the spleen in vaccinated groups ( G1 and G2 ) but they could be seen in nearly all non-vaccinated groups ( G3 , G4 and G5 , Fig . 5B ) . In the liver , mononuclear cell infiltration in portals was more prominent in non-vaccinated groups at week 14th . Plasma cells were only seen in these groups . Interface hepatitis was not a significant finding . Parasites in the portals were only seen in non-vaccinated groups ( G3 and G5 ) in 8th week ( data not shown ) . In all groups , lobular inflammation was seen at 4th week , increased significantly in 8th week and decreased at week 14 . The severity of this inflammatory response at 4th week was higher in G3 compared to the other groups ( 15–16/10 hpf ( high power field ) v . s 2–4/10 hpf of vaccinated groups , p<0 . 05 ) ( Fig . 6 ) . Inflammatory cell infiltration was seen in the liver parenchyma in groups 1 , 2 , 3 , 4 and 5 at 4th week after challenge ( Fig . S4 ) . Portal granuloma formation ( either mature or immature ) was seen as early as 8th week in non vaccinated group , but in parenchyma of all groups they were seen at this time and persisted until 14th week ( Fig . 5C and 5D ) . Parasite in the parenchyma was visualized only in one out of 12 mice of vaccinated groups but it was seen in 11 out of 18 non-vaccinated mice . Occurrence of cholestasis was noted at 8th week with no predilection to any group ( data not shown ) . There was no significant difference in neutrophil infiltration , fatty changes , giant cell formation , and kupffer cell hyperplasia between groups ( data not shown ) . Liver cell necrosis was minimal and no significant regenerative changes were noted ( data not shown ) . No parasites were found in bone marrow macrophages and no significant changes in bone marrow of all examined mice were observed ( data not shown ) .
Although preventive vaccines are recognized as the best and most cost-effective protection measure against pathogens , no effective vaccines are available to control leishmaniasis . Leishmania vaccine development has been proven to be a difficult and challenging task and is hampered by an inadequate knowledge of disease pathogenesis , the complexity of immune responses needed for protection , and the cost of vaccine development [16] , [65] . Immunization against experimental visceral leishmaniasis in murine models has been reported to be more difficult to achieve than for the cutaneous form . This may be due to the more complex situation in immunopathology of murine infection with VL species . In particular , the outcome of VL infection in mice does not depend on the Th1 versus Th2 subset expansion [66] , [67] . Opposite to L . major murine model , this dichotomy seems inadequate in human and murine forms of VL . Type 1 immune responses are suppressed by IL-10 and TGF-β [68] . Based on recent data about T reg expansion during infection , Foxp3 expressing CD4+CD25+ T cells were found responsible for TGF-β and CD4+CD25−Foxp3− T cells for IL-10 production [69] . In some areas , the inoculation of infectious material isolated from cutaneous lesions or cultured virulent promastigotes ( a process called leishmanization ) in hidden parts of uninfected individuals has been used to prevent further lesions [70] . These individuals develop a strong immunity to reinfection , sometimes only after unpleasant clinical episodes . Thus , there is a general consensus among Leishmania vaccine researchers that parasite persistence and exposure to complex antigens in the right context over time may be important for effective protective response and could be achieved by live attenuated parasites immunization . However , reversion of these parasites to the virulent form restricts their use [71] . Early attempts for the development of attenuated strains ( e . g . obtained through long-term in vitro culture , γ-irradiation , selection for temperature sensitivity , or chemical mutagenesis ) did not elicit a protective immunity [72] . On the other hand , manipulation of the Leishmania genome through discovery of new genes involved in parasite growth and survival revives the potential of a live attenuated parasite vaccine with reduced risk of reversion . In this context , several successful attenuated lines of L . donovani , such as a partial knockout of A2-A2rel gene clusters with attenuated virulence in mice [73] , a biopterin transporter knockout with reduced infectivity and induced protection against challenge [35] and a centrin ( LdCen1−/− ) knockout with reduced parasite survival in macrophages [74] have been generated . Using these live attenuated Leishmania vaccines , which most closely mimic the natural course of infection , could be very beneficial . On the other hand , return back to virulence may occur , and there is an increasing need to develop new safer live vaccine vectors that are capable of enhancing antigen presentation and eliciting potent immune responses without the risk of disease development in humans . Consequently , a live nonpathogenic to humans parasitic vector consisting of a lizard parasitic protozoan , L . tarentolae , has been introduced by Breton et al . as a candidate vaccine against visceral leishmaniasis [38] . We have demonstrated previously that a recombinant L . tarentolae strain expressing the L . donovani A2 gene elicited a strong protective immunity against virulent L . infantum challenge [42] . We have also showed that vaccination with L . infantum cysteine proteinases type I and II as part of a prime-boost strategy induced a strong parasite-specific Th1 response and conferred protection against parasite challenge [51] . As a follow-up of our previous work , the objective of this study was to generate a recombinant L . tarentolae strain stably expressing three candidate antigens as part of a tri-gene fusion , including the L . donovani A2 and L . infantum CPA and CPB genes and to evaluate its potential as a candidate live vaccine against VL using different prime-boost vaccine modalities . Immunization of BALB/c mice with recombinant L . tarentolae A2-CPA-CPB-CTE as a DNA/Live or a Live/Live prime-boost regimen ( groups G1 and G2 ) elicited a significant protective immunity against a high dose virulent L . infantum challenge . Moreover , mice immunized with these modalities did control the parasite propagation both in the liver and the spleen at 4 weeks but also 12 weeks after challenge . It is worth mentioning that tissue parasite loads were measured with the limiting dilution assay that is semi-quantitative and may give inconsistent results due to medium and incubation conditions . Hence , for more accurate measurements of parasite burden considering both LDA and qPCR is highly recommended . Therefore at week 4 after infectious challenge , we also confirmed the quantity differences between vaccinated groups ( G1 and G2 ) and PBS control group ( G3 ) by qPCR in both liver and spleen . Interestingly , the obtained results for this specific time point are in accordance with semi-quantitative LDA . In our experimental system , immunization of BALB/c mice with recombinant L . tarentolae A2-CPA-CPB-CTE revealed the induction and secretion of IFN-γ and IL-10 , although IFN-γ was much higher than IL-10 . Determining the main source of IL-10 production has some beneficial results for understanding mechanisms of resistance and susceptibility to VL . Here , in immunized G1 and G2 groups , we detected high IL-10 production , especially at 8 weeks after challenge , upon stimulation with L . infantum lysate as a recall antigen . The sources of IL-10 in human VL have not been defined yet . IL-10 can be produced by several cell types such as monocytes , macrophages , B cells and CD4+ T cells . It has been described that several subpopulations of IL-10 producing CD4+ T cells have the ability to inhibit the response of other T cells . One of the important sources of IL-10 in the suppression of anti-leishmanial immunity in human VL is T regulatory ( T reg ) cells that arise from CD25−Foxp3− T cells [75] . Also , there are some CD3+ CD25− depleted cells that produce both IFN-γ and IL-10 cytokines . Simultaneous production of IFN-γ and IL-10 by human T cell clones can be induced by IL-12 [76] . In addition , IL-10 produced by adaptive T reg cells may be preferentially required to control tissue damage in sites of strong inflammation associated with antimicrobial responses [77] . The IFN-γ/IL-10 ratio in splenocytes stimulated with rA2-rCPA-rCPB from mice immunized with the recombinant L . tarentolae A2-CPA-CPB-CTE was ∼5 . 1- to 11 . 2-fold higher at 4 and 8 weeks post-challenge than mice immunized with wild type L . tarentolae only . The same result was obtained with splenocytes stimulated with other F/T antigens . Also , histopathological studies confirmed the Spearman correlation between the ratio of IFN-γ/IL-10 production and differences in parasite burden in liver and spleen cells as no parasites were seen in the liver parenchyma and spleen at 4th week post-challenge in the vaccinated groups G1 and G2 , whereas they were easily detected in the non-vaccinated control groups . In addition , the level of both IgG1 and IgG2a at different time points before and after challenge was not significantly different between the vaccinated groups . Thus , cytokine data along with antibody titration indicated that vaccination with DNA/Live L . tarentolae A2-CPA-CPB-CTE and Live/Live L . tarentolae A2-CPA-CPB-CTE induce a clearly stronger Th1 response . Although it has been suggested that the control of splenic parasite burden is likely to be mediated through a NO-dependent or a NO-independent pathway [78] , in this report we showed that L . tarentolae A2-CPA-CPB-CTE , especially when administered as a Live/Live modality , induces a strong humoral and cellular immune response in addition to NO generation in response to rA2-rCPA-rCPB , F/T L . tarentolae A2-CPA-CPB-CTE and F/T L . infantum specific stimulators at 8 weeks after infectious challenge with L . infantum . Possibly , the cumulative effect of IFN-γ along with iNOs-mediated parasite killing is responsible for this response . Furthermore , it has been reported that one of the major factors contributing to healing of leishmaniasis is the development of strong CMI ( cell-mediated immunity ) response like IFN-γ and NO production [66] , [68] , [79] . Therefore , higher IFN-γ and NO production in vaccinated groups G1 and G2 suggests a fine correlation between CMI and resistance to VL infection . The generation of NO in stimulated cells with rA2-rCPA-rCPB , F/T L . tarentolae A2-CPA-CPB-CTE and F/T L . infantum also supports the up-regulation of inducible NO synthase by TH1 cells and confirms that the NO-mediated macrophage effector mechanism is critical in the control of parasite replication in these animals . Our results showed that the nitrite ( NO2 ) levels are in direct correlation with IFN-γ production in these groups and that higher production of IFN-γ was obtained after 56 days of infection , indicating a possible control of the parasites' persistence and suggesting that the control of L . infantum infection may occur by a NO-dependent pathway . It has been described previously that during the early stages of visceral infection in BALB/c mice , parasites multiply in large numbers in the liver; however , once the infection becomes chronic , hepatic parasite loads tend to decrease , while parasitism in the spleen and BM tends to increase [63] . Pathology is mediated by the direct loss of specific cell populations and changes to the local tissue microenvironments that ultimately decrease the induction of effective immune responses and by the inability of splenic macrophage populations to generate leishmanicidal mechanisms or to recruit appropriate cells for eliminating the parasite . Furthermore , the intensity of pathological changes in the visceral organs of BALB/c mice can vary depending on the initial inoculum size [63] , [80] . Pathological studies showed that after infection with 107 L . infantum promastigotes , control groups exhibited severe histopathological alterations in both the spleen and liver at the peak of parasite burden . Once in the liver , the development of cell-mediated immune responses is essential for the clearance of L . infantum parasites . In contrast , the spleen ultimately becomes the site of parasite persistence [63] , [81] , [82] , suggesting that it is more susceptible to L . infantum infection than the liver [83] . Interestingly , the leishmanicidal efficacy of hepatic granulomas is dependent on their degree of maturation [63] , [84] , [85] . Among these alterations , we detected the appearance of granulomas in different maturation stages and giant cell granulomas in amastigotes in the liver of all groups infected with L . infantum , which results in liver parasite clearance . However , disruption of the splenic architecture accompanied by lymphoid depletion only in control groups results in spleen parasite persistence . Overall , our results on cytokine production , humoral responses , parasite burden and histopathological studies support that immunization with the novel recombinant L . tarentolae A2-CPA-CPB-CTE candidate vaccine protects mice against visceral leishmaniasis when administered as a prime-boost modality more than L . tarentolae [38] or L . tarentolae A2 [42] . The next step will be to determine the long-term memory protection in mice or hamsters and to evaluate the effectiveness of this promising live vaccine against L . infantum in dogs as an important outbreed animal model for VL research .
|
Visceral leishmaniasis ( VL ) is the most severe form of leishmaniasis and has emerged as an opportunistic infection in HIV-1 infected patients in many parts of the world . Drug-resistant forms have developed so emergence and increased the need for advanced preventive strategies . Using live avirulent organisms as a vaccine has been proven to be more effective than other regimens . The lizard protozoan parasite Leishmania tarentolae is considered as nonpathogenic to humans . In our previous work , a recombinant L . tarentolae strain expressing the amastigote-specific L . donovani A2 antigen as a vaccine candidate elicited protection against L . infantum challenge in mice . Furthermore , combinations of CPA/CPB cysteine proteinases were more protective against visceral and cutaneous Leishmania infections than the individual forms . Herein , we used DNA/Live and Live/Live prime-boost vaccination strategies against visceral leishmaniasis in BALB/c mice consisting of the A2-CPA-CPB-CTE tri-fusion genes formulated with cationic solid lipid nanoparticles and a recombinant L . tarentolae expressing the tri-fusion . Assessments of cytokine production , humoral responses , parasite burden and histopathological studies support that the recombinant L . tarentolae A2-CPA-CPB-CTE candidate vaccine elicits a protective response against visceral leishmaniasis in mice and represents an important step forward in the development of new vaccine combinations against Leishmania infections .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"biology",
"veterinary",
"science"
] |
2013
|
Development of Novel Prime-Boost Strategies Based on a Tri-Gene Fusion Recombinant L. tarentolae Vaccine against Experimental Murine Visceral Leishmaniasis
|
Large-scale gene expression datasets are providing an increasing understanding of the location of cis-eQTLs in the human genome and their role in disease . However , little is currently known regarding the extent of regulatory site-sharing between genes . This is despite it having potentially wide-ranging implications , from the determination of the way in which genetic variants may shape multiple phenotypes to the understanding of the evolution of human gene order . By first identifying the location of non-redundant cis-eQTLs , we show that regulatory site-sharing is a relatively common phenomenon in the human genome , with over 10% of non-redundant regulatory variants linked to the expression of multiple nearby genes . We show that these shared , local regulatory sites are linked to high levels of chromatin looping between the regulatory sites and their associated genes . In addition , these co-regulated gene modules are found to be strongly conserved across mammalian species , suggesting that shared regulatory sites have played an important role in shaping human gene order . The association of these shared cis-eQTLs with multiple genes means they also appear to be unusually important in understanding the genetics of human phenotypes and pleiotropy , with shared regulatory sites more often linked to multiple human phenotypes than other regulatory variants . This study shows that regulatory site-sharing is likely an underappreciated aspect of gene regulation and has important implications for the understanding of various biological phenomena , including how the two and three dimensional structures of the genome have been shaped and the potential causes of disease pleiotropy outside coding regions .
It has been almost 30 years since the locus control region at the human β-globin cluster was identified [1] , one of the first well-defined mammalian examples of a regulatory site linked to the regulation of multiple , nearby genes [2] . Subsequent studies across species have suggested that such regions may be a common feature of eukaryote gene expression [2] . However , despite their potential importance , only a handful of such regions have been identified to date . How common such master cis-regulatory sites are in the human genome remains largely unclear . This is partly because the study of regulatory sites has focused primarily on their effect on individual genes . The study of expression quantitative trait loci ( eQTLs ) , for example , has shown itself to be a powerful tool in the understanding of the genetic basis of gene regulation [3] . Genetic variants linked to gene expression variation ( eVariants [4] ) point towards the location of regulatory elements in the genome . However , traditionally , eQTL studies adopt a single variant to single gene testing approach; that is testing variants against genes one by one [3] . These analyses provide little information on the extent to which regulatory sites are shared across genes . High levels of linkage disequilibrium between variants can confound where two eVariants are tagging distinct regulatory loci , or where multiple genes are in fact linked to a single regulatory site . Supporting the idea of substantial co-regulation in the human genome is the well characterised observation that nearby genes often display similar expression patterns across tissues [5] . A number of hypotheses have been proposed to explain this , one of which is that such genes are directly co-regulated . It has also been proposed that the similarities in expression profiles of nearby genes are simply though an artefact of their shared chromatin environment [6 , 7] . The latter is supported by observations that co-expressed genes are more likely to move apart over evolutionary time [6 , 8] , suggesting that co-expression is an unwanted artefact of proximity and that selection acts to reduce the interference between the regulation of nearby genes . This apparent contradiction between studies is potentially due to the inability to distinguish between artefactual co-expression and directed co-regulation when only the expression levels of genes are analysed . Where co-regulation does exist , the mechanism by which master regulatory sites may co-regulate multiple genes is also poorly understood . A number of models have been proposed to explain how a single site could regulate multiple genes [2 , 9] . These include: that transcription factor binding at the locus leads to remodelling of the chromatin across the entire locus making it more conducive for transcription; that transcription factors bind to the regulatory site which then track along to the various genes; or the promoters of the genes interact with the regulatory site via chromatin looping . Finally , the importance of regulatory site sharing in pleiotropy , the phenomenon by which a single gene or variant is associated with more than one phenotype [10] , has been largely unexplored . Traditionally pleiotropy has been thought of in terms of a particular gene being linked to multiple traits [10] but there is growing evidence that this view may be outdated . The fact that 88% of variants linked to a disease map outside transcribed regions [11] suggests that pleiotropy may also be driven by non-coding regions . Genetic variants at shared regulatory sites affect the expression levels of multiple genes and , as a consequence , multiple downstream phenotypes . In summary , despite their potential importance to a variety of key biological processes , from understanding disease risk , to pleiotropy and the forces shaping genome evolution , little is currently known about the locations and extent of master regulatory sites in the human genome . To begin to address this we first identified shared , non-redundant , cis-regulatory sites in the human genome , which we term master-eQTLs . We then investigated the links between these sites and disease , their potential mechanisms of action and how they appear to have shaped the evolution of our genome .
To determine the location of master-eQTLs , we first identified the location of cis-eVariants in the human genome using a dataset of 379 European derived lymphoblastoid cell lines [12] . Genetic variants within 500kb of each TSS in the genome were tested for an association with the gene’s expression using linear regression , while controlling for any sex and population of origin effects . Robust permutation-adjusted P values were subsequently calculated for all eVariants reaching a nominal significance of p < 0 . 00001 . A set of 441 , 723 cis-eVariants identified at a false discovery rate of 0 . 05 . A final pruning step based on forward regression ( see methods ) was used to reduce the redundancy among the set of eVariants for each gene , and identify a final set of 5 , 254 conditionally independent regulatory cis-eQTLs . This pruned , non-redundant set of eVariants is provided as S1 Table . To assess the quality of this set of cis-eQTLs and the impact of pruning , we explored the replication of these eQTLs across tissues using independent data from the GTEx consortium [13] . For each eVariant defined in this study , we obtained its corresponding GTEx p value in 44 tissues , including matching independent lymphoblastoid cell lines . As illustrated in Fig 1 , the cis-eVariants identified in this study displayed a high level of reproducibility in related tissues such as spleen , the primary storage area of lymphocytes . In particular the conditionally independent set of cis-eVariants , each expected to correspond to a separate eQTL , showed the highest reproducibility among the corresponding dataset of lymphoblastoid cell lines , illustrating that these independent eVariants comprise a high quality collection of regulatory variants that replicate across analysis approaches and datasets . In comparison to other tissues , pruning had a comparatively modest impact on replication in the matching set of GTEx lymphoblastoid cell lines ( Fig 1 , bottom panel ) . This is broadly consistent with pruning having reduced spurious cross-tissue replication , while not substantially affecting replication across the same tissue type ( see S1 Fig for further details ) . This pruned set of non-redundant eQTLs was therefore used in all downstream analyses . After having identified this set of non-redundant cis-eQTLs , we assessed the extent of eQTL sharing , that is where the expression of multiple genes is linked to a single eVariant . In total 534 eQTLs , 11 . 8% of the total , were found to be linked to the expression of two or more genes . Fig 2 shows examples of putative shared regulatory sites at the COLCA1 and COLCA2 locus on chromosome 11 . Of 4 eQTL variants independently associated with these genes , two are shared by both , consistent with previous studies demonstrating co-regulation between these genes [14] . S2 Fig shows that the probability of a variant being a non-redundant eVariant is related to its allele frequency and distance from the corresponding gene . Independent eVariants are generally of a higher minor allele frequency and in close proximity to the gene they are linked to . This suggests regions of high gene density may show a high degree of eQTL overlap by chance . To provide a baseline frequency of eQTL sharing , we used these frequencies to determine the probabilities of eQTLs being linked to multiple genes based solely on the allele frequencies and distance to nearby genes of eVariants ( see methods ) . Comparatively few eQTLs are expected to overlap by chance based solely on these frequencies , suggesting a bias towards shared regulatory sites in the genome ( Fig 3 ) . Although these master-eQTLs may indeed correspond to shared regulatory sites , an alternative explanation is that they are in fact two independent eQTLs in high LD . The tight genetic correlation of many eVariants means that their effects cannot be easily teased apart . However , two independent regulatory variants , affecting both distinct regulatory sites and downstream genes , will not necessarily show correlated effects on the expression of the genes across different tissues . Transcription factor binding , histone modifications and other chromatin factors at these distinct sites can change independently of one another and the underlying genotypes at these variants , leading to divergence in the expression levels of the genes across tissues . In contrast changes in transcription factor binding across tissues at shared regulatory sites are shared by each co-regulated gene , suggesting the effect of variants at such sites will have more correlated effects on associated genes across tissues than two independent , but linked , variants . This theory is illustrated in Fig 4A and 4B . Although the two eVariants , rs13247029 and rs35121828 , are in high LD ( R2 = 0 . 88 ) the estimated size and direction of their effects on the genes they regulate are not correlated across tissues . The expression levels of the genes change independently across cell types , and show different associations with the distinct eVariants . In contrast the size and direction of effect of the master-eQTL eVariant rs36209093 on the two genes it is linked to , are highly correlated across tissues . We tested to see if this is a general phenomenon and a feature of the master-eQTLs defined here , that is whether multi-eQTLs show a greater correlation in their size and direction of effect on their linked genes across tissues than would be expected if multi-eQTLs are in fact tagging two independent regulatory sites . To test this , we obtained from the independent GTEx dataset the coefficients across 44 tissues for each of our independent eVariants . Each eQTL-gene pair’s set of 44 tissue coefficients represents the size and direction of effect of the eQTL across tissues . The Pearson’s correlation was then calculated between the sets of coefficients associated with each pair of eQTL-gene sets in the genome ( rtiss ) . A high rtiss indicates that the size and direction of effect of the eQTL or eQTLs on their associated genes is highly correlated . rtiss is generally low for randomly selected pairs of eQTLs ( Fig 4C ) , i . e . the size and direction of effect of distinct eQTLs on different genes across tissues are generally uncorrelated . A similar pattern is observed if nearby gene-pairs are randomly selected to approximately match the distribution of inter-gene distances observed between genes sharing a master-eQTL . The coefficients associated with these gene pairs and their distinct eQTLs also show little correlation ( low rtiss ) , suggesting two genes simply being in close proximity does not necessarily increase their level of co-regulation across tissues . However , rtiss associated with master-eQTLs and associated genes are unusually high . The sets of coefficients associated with master-eQTLs and each co-regulated gene were generally more highly correlated than expected . As the genotypes of genetically linked variants are correlated , this may partly explain correlations between associated sets of eQTL coefficients . As shown in Fig 4D , there is indeed a weak relationship between the correlations observed between the sets of coefficients associated with pairs of eQTLs ( rtiss ) and increasing levels of LD between the variants . The higher the LD between two eQTLs the greater the correlation between their size and direction of effect on the expression of their associated genes across tissues . However , extrapolating out this relationship in Fig 4D , master-eQTLs display correlations between their associated sets of coefficient ( rtiss ) that are substantially higher than would be expected from two independent master-eQTLs in perfect LD . These data suggest that master-eQTLs are not simply the result of independent regulatory variants in perfect LD , or because the genes simply share the same broader chromatin environment , but rather are likely enriched with true shared regulatory variants . To investigate the potential mechanisms by which multiple , often distant , genes may potentially be co-regulated , we tested for evidence of chromatin looping between master-eQTLs and their associated genes [15] . Two complementary Hi-C chromosome conformation datasets were used . One traditional genome-wide Hi-C analysis [16] characterising chromatin interactions genome-wide , and one higher resolution but targeted study of loops specifically associated with gene promoters [17] . In total 175 of the 1039 gene pairs sharing an eQTL ( 16 . 7% ) displayed evidence of both genes looping towards the shared regulatory site ( S2 Table ) . A high degree of overlap was observed between both chromatin conformation datasets ( S3 Fig ) . To assess the significance of this result we adopted a circular permutation approach [18] . This involved maintaining the size , order and interactions between Hi-C target and anchor regions , but shifting the locations of all the Hi-C regions the same random distance along a circularised version of the genome , to test how often looping between genes and regulatory regions is expected by chance ( see methods for more details ) . Although genes that are closer together are expected to be more often linked by chromatin interactions , this circular permutation approach ensures that factors such as the distance between genes are controlled for . Using this approach the rate of chromatin looping between master eQTLs and their linked genes was observed to be approximately 2 . 6 times higher in the real data than the average of the permutations ( permutation p<0 . 001 , Fig 5A and S4 Fig ) , with none of the 1000 permutations showing as much looping between co-regulated genes and shared eQTLs as observed in the real data . Some genes were observed to map to the same Hi-C chromatin domain , so that the eQTL is brought into both their proximity via one higher order chromatin loop . However , distinct chromatin looping between regions harbouring an eQTL and each of the regulated genes was also observed significantly more often than expected ( S4 Fig ) . This suggests that co-regulated genes often appear to be brought into the vicinity of a shared eQTL via multiple distinct chromatin looping events . The promoters of genes sharing an eQTL are also found closer together in three-dimensional chromatin space than expected by chance given their genomic distance apart , ( permutation p<0 . 001 , Fig 5A and S5 Fig ) . Consequently master-eQTLs and their associated genes often appear to be forming higher order multi-way chromatin interactions in the cell . An example of a co-regulated gene cluster is shown in Fig 5B . The master-eQTL at this locus on chromosome 6 is significantly associated with the expression of 6 genes . Five of these genes show evidence of chromatin looping to the region harbouring the putative regulatory variant , with the final gene residing directly upstream of the eQTL itself . The promoters of these genes were also found to loop towards one-another in three dimensional chromatin space ( S3 Table ) , supporting a multi-way interaction at this locus . Although , as far as we are aware , no master regulatory site has previously been described in this region containing three SERPINB genes , a locus control region has been characterised at the SERPINA cluster on chromosome 14 [19] , suggesting that both SERPIN loci are potentially under the influence of master regulators . Consequently there is evidence that master-eQTLs and associated regulated genes are often forming higher order multi-way chromatin interaction modules in the cell . An active role of chromatin looping in gene co-regulation suggests that such clusters may be expected to be under evolutionary constraint to be conserved together across time; co-regulated gene pairs being disproportionately found within unusually close proximity , and over 21% being within 10kb ( S6 Fig ) . Previous studies of co-expression in mammals have argued against co-expressed gene clusters being maintained together and that rather they have generally moved apart over evolutionary time [6] . A potential limitation of these studies may be the focus on co-expression . Using our set of gene pairs showing evidence for genetic co-regulation , we tested to see if the sharing of an eQTL and chromatin interactions are linked to the conservation of gene order across species . Controlling for their distance apart in the human genome , genes that loop together towards a shared genomic target were found to have been maintained at a more similar distance apart in both the chimpanzee and mouse genomes than those gene pairs with no evidence of being linked by chromatin looping ( Fig 6 ) . Where genes loop towards the same region that also harbours a shared eQTL , there appears to be an unusually strong constraint on their inter-gene distance . The average change in inter-gene distance of such gene pairs is not significantly different from 0 in both the human and chimpanzee analyses . Despite the general increase in inter-gene distance of other gene pairs , these genes linked by chromatin looping to a shared regulatory site have been maintained at largely the same genomic distance apart across these comparatively long evolutionary timescales ( Fig 6 ) . Comparing these gene pair groups to each other while controlling for their distance apart in the human genome via multiple linear regression , reaffirms gene pairs linked by chromatin looping to a shared eQTL display the smallest increases in inter-gene distance ( Table 1 ) . Chromatin looping , that has been shown to be relatively well conserved across mammals [20] , appears to be associated with constraints on genome evolution . However , the combination of chromatin interactions and shared eQTLs is associated with the strongest conservation of inter-gene distances across species . Co-regulated gene modules linked by chromatin looping and a master-eQTL appear to be under unusually strong constraint to maintain them together over evolutionary time . Shared regulatory sites may be expected to be particularly important in shaping downstream phenotypes due to their potential link to multiple biological pathways . Previous studies have shown that eVariants are also often a GWAS variant [12] , but whether master-eQTLs are more likely to be linked to multiple phenotypes has not previously been examined . To investigate this we first identified the total set of variants in LD with each eQTL at different thresholds ( R2 > = 0 . 6 , > = 0 . 8 , = 1 ) , and using information from the GWAS catalogue [21] determined the total number of phenotypes linked to each set . Where an eQTL was in LD with multiple GWAS variants associated with the same or similar phenotype they were collapsed into one , providing a non-redundant count of phenotypes linked to the variants in LD with each eQTL . Depending on the LD threshold used , master-eQTLs were observed to be between 2 . 5 and 2 . 6 times more likely to be linked to one or more GWAS phenotype than eQTLs linked to only one gene ( chi-squared p < 5x10-5 at all three LD thresholds ) . Notably , this broadly tallies with master-eQTLs being associated with 2 . 34 genes on average . Multi-eQTLs were observed to be associated with a modest , albeit significant , increase in the number of variants in LD ( Fig 7 ) . However , accounting for this general enrichment of linked variants using logistic regression , the sets of variants in LD with master-eQTLs were found to be generally associated with more distinct phenotypes than variants in LD with single-eQTLs . Each extra phenotype with which an eQTL is associated , increases the odds of it being a multi-eQTL by a factor of between 1 . 31 to 1 . 51 depending on the LD threshold used ( Fig 7 ) . Amongst all eQTLs there is a significant correlation between the number of genes whose expression it is linked to and the number of non-redundant GWAS variants in LD ( R2 threshold of > = 0 . 6 . Partial correlation Kendall’s tau: 0 . 094; p = 4 . 34x10-21 when accounting for the total number of variants in LD with each eQTL ) . Consequently eQTLs linked to multiple genes are more likely to be in LD with multiple , non-redundant GWAS variants suggesting the regions of shared regulatory sites are key hubs of disease risk and human phenotypes .
There is an increasing focus on understanding the genetics of gene regulation , with studies such as GTEx [13] assaying the links between genetic variants and gene expression in hundreds of individuals and across multiple tissues . Despite the inarguable utility of these analyses , their focus on producing lists of individual gene to individual variant associations perhaps oversimplifies genetic regulation . In this study we identify independent cis-eQTLs and show that the sharing of non-redundant regulatory variants is a relatively common phenomenon . Shared regulatory variants are not restricted to a small handful of sites , but found across over 10% of all pruned regulatory sites . Although distinguishing between a single master regulatory site , and multiple regulatory sites in perfect , or near perfect , LD is difficult [22] , suggesting this estimate is likely an upper bound , we have shown that genes sharing a master-eQTL show correlated changes in expression linked to the variant’s genotype across tissues . This therefore supports the idea of substantial levels of co-regulation in the human genome . Likewise the patterns of chromatin looping linked to master-eQTLs adds further support to the function of these sites . The observation that genes sharing a regulatory site are less likely to move apart over evolutionary time also suggests that regulatory site sharing is not simply a result of regulation “leaking” from one gene to its neighbours as has previously been suggested [6] . Selection is expected to break up gene pairs if regulation of one is adversely affecting its neighbours [6] . In contrast , such gene groups are unusually well conserved , suggesting the opposite is likely to be the case , and that their proximity is being maintained over relatively long evolutionary timescales . A limitation of eQTL studies is that they are dependent on a polymorphism falling within the regulatory site . Many , if not the majority , of regulatory sites are therefore likely to be missed in any study based on genetic regulatory variants . However , we can extrapolate from the numbers observed in this study . Although we identified just 534 shared eQTLs , only 3422 genes were linked to an eQTL in this analysis , with the majority of regulatory sites unlikely to be tagged by an eQTL . With as many as 50 , 000 ( coding and non-coding ) genes in the genome [23] , this suggests there may be several thousand shared regulatory sites in total . As defining the locations of eQTLs is a common post-GWAS prioritisation approach , understanding the distribution and location of these sites has important implications for interpreting disease loci . This study has shown that master-eQTLs are more likely to be linked to several phenotypes than other eQTLs . Pleiotropy is receiving increasing attention , in part due to its potential for increasing power in association studies [22] , and these results provide a potential mechanism for interpreting pleiotropy outside the most commonly studied coding regions .
Ethical , legal and social implication ( ELSI ) statements for the GEUVADIS and GTEx datasets used in this study can be found at: http://www . geuvadis . org/web/geuvadis/resources/elsi and https://www . ncbi . nlm . nih . gov/pmc/articles/PMC4010069/#S11title . Cis-eVariants were defined using the 379 European Individuals in the 1000 genomes project that have matching GEUVADIS gene expression data available [12] . All 1000 genomes phase 3 variants [24] within 500kb of the transcription start site of each gene were first tested for an association with the gene’s expression levels using linear regression , fitting sex and 1000 genomes population of origin as covariates . Permutation p values were then calculated for those cis-eVariants reaching a nominal significance of p < 0 . 00001 using the lmPerm R package . A maximum of 1 million iterations were run for each variant , with iterations terminating if the estimated standard deviation of the p value fell below 0 . 1 of the estimated p value [25] . Forward stepwise regression was used to reduce the redundancy among these variants and identify conditionally independent eQTLs . The lmPerm R package again was used to determine conditioned permutation p values ( with a maximum iteration number of 10 million ) . As these independent eVariants , each expected to correspond to a distinct eQTL , were defined for each gene independently , two different variants could be associated with different genes , despite being redundant and tagging the same functional variant . To try and remove this redundancy , a final step was undertaken where redundant eQTLs were merged across genes . For a given gene , gene x , we determined those eVariants that were dropped following pruning but that corresponded to the same eVariant as an eQTL that survived pruning for another gene , gene y . If this eVariant was equally effective at explaining variation in gene x’s gene expression as the current eQTL of gene x , the two variants were marked as potentially redundant ( if an ANOVA F-test p value comparing linear models with both or just one SNP fitted against the gene’s expression was > 0 . 05 ) . The variant with the larger chromatin enrichment in the immediate vicinity ( see below for calculation ) was subsequently chosen to be the eQTL assigned to both genes x and y if equally effective in explaining each gene’s variation in expression . In total 7 . 8% of non-redundant eVariants were replaced with another representative in this way . A final round of forward regression with this final set of pruned eVariants was undertaken to ensure this process did not affect the redundancy among eQTLs . Finally any eVariants not in V6 of GTEx were excluded from all analyses , leaving the final set of 5254 non-redundant eQTLs . A comparison of the numbers of eQTLs linked to multiple genes prior to and after this final merging step is shown in S8 Fig . To determine the expected amount of eQTL sharing among genes purely by chance we determined the frequency at which variants of a particular minor allele frequency ( MAF , grouped into 2% frequency bins ) and distance from the TSS of the respective gene ( 10kb bins ) were called an independent eVariant . These frequencies equate to the probability of a variant corresponding to an eQTL of a gene given its MAF and distance to the gene’s TSS . When calculating these frequencies we also conservatively included all eVariants in perfect LD with each independent eQTL , to account for any increased eQTL sharing resulting from the same variant being picked for each gene when multiple variants were of equal significance . These frequencies were used in 100 permutations to assess the expected amount of eQTL sharing by chance , based purely on this observed distribution of regulatory variants in the genome . In each permutation all variants in the final set of eQTLs were first randomly assigned to just one of the genes of which it was an eQTL . Using the probabilities in S2 Fig the probability of this eQTL being associated with each of the remaining genes within 500kb was then determined . If this probability was greater than a randomly drawn number from a uniform distribution between 0 and 1 , this variant was deemed to be an eQTL for the corresponding gene in this permutation . In this way the frequency of observing eQTLs linked to multiple genes based on their minor allele frequency and distance to nearby genes could be determined . The average number and 95% interval ranges of eQTLs linked to different numbers of genes across permutations was finally determined to compare to the observed counts in Fig 3 . In our approach to define master-eQTLs described above we had undertaken a final step merging redundant eQTLs across genes . This had the effect of increasing eQTL sharing . However , as shown in S8 Fig , even without this final merging step eQTL sharing was substantially higher than expected based purely on the distribution of eQTLs in the genome represented in S2 Fig . Consequently , even without collapsing eVariants in high LD into one eQTL , there is an enrichment of eQTL sharing in the genome . To test the cross-dataset and cross-tissue reproducibility of eVariants ( i . e . whether the eVariants are significant in other tissues ) , GTEx P values for the eVariants defined in this study were obtained from http://www . gtexportal . org/home/ . As discussed any variant not tested by GTEx was excluded from all analyses . To test whether the defined master-eQTLs demonstrate unusually high correlations between their size and direction of effects on genes across tissues , we also obtained GTEx eQTL coefficients from the GTEx website . These coefficients represent the degree to which the gene’s expression changes upon changes in the eVariant’s genotype . Consequently a high correlation ( rtiss ) between two sets of coefficients suggests the expression of the two corresponding genes changes in a similar way across tissues upon changes in the genotype of the eVariant ( s ) . In this analysis we calculated the Pearson’s correlation ( rtiss ) between all sets of coefficients associated with our pruned eQTLs having excluded eQTLs for which a coefficient was not available in at least three quarters of the 44 tissues studied . These correlations therefore represent the degree to which two different eQTL-gene pairs show similar sizes and direction of effects across tissues ( see examples in Fig 4A and 4B ) . From the total set of correlations a subset corresponding to the correlations between the coefficients associated with a master-eQTL and two associated genes were first extracted ( “Share a master eQTL” group ) . The remaining group ( “others” ) was composed of those correlations where the eQTL was not tagged by the same variant i . e . the two sets of coefficients corresponded to two different eQTL-gene pairs . We then sampled 100 times from this group subsets of correlations corresponding to pairs of genes whose inter-gene distances approximately matched ( same 50kb bin ) the distribution of inter-gene distances of genes sharing a master-eQTL ( “close proximity” group ) . To investigate the link between LD and the correlations between these tissue-level coefficients we calculated LD between all pairs of independent eQTL variants within 500kb of each other using PLINK [26] . Interacting regions of the human genome were obtained from Jin et al . [16] and Mifsud et al . [17] . The former is a genome-wide study of looping in the human genome and the latter a higher resolution study of loops specifically associated with gene promoters . To assess the significance of the numbers of eQTLs and gene pairs found within genomic regions interacting in the genome-wide study of Jin et al . [16] we adopted a circular permutation approach . All autosomes being concatenated and in each permutation all genomic regions defined in the Hi-C dataset being moved the same random distance along the genome ( shifts being between 20kb and the total length of all the autosomes ) . Any regions falling off the end of the concatenated genome were added back on to the beginning . In this way we maintained the relationship between genomic regions , as well as their relative distances and sizes , but broke their relationship with eQTLs and gene locations . If the relationship between eQTLs and the TSS they are linked to is unrelated to chromatin interactions , we would have frequently expected to see as many eQTLs and TSSs linked by chromatin looping in these permuted Hi-C datasets . Our ( one tailed ) p value was calculated as the proportion of permutations where the number of eQTL-gene pair associations linked by Hi-C loops was greater than or equal to the number in the unpermuted data . Links between TSS pairs were assessed in the same way i . e . the Hi-C dataset was shifted a random distance along the circularised genome and the proportion of permutations determined where the number of gene pairs linked by chromatin looping was greater than or equal to the number in the real data . The orthologues of human genes in the mouse and chimpanzee genomes were obtained from Ensembl [23] . Only orthologues with a one to one relationship between species were kept . All inter-gene distances were calculated between the TSS of genes found on the same chromosome in both species i . e . the measure is the change in distance between genes due to insertions , deletions , inversions etc . but not translocations . LD levels between independent eQTL variants defined in this study and all variants within 500kb were calculated using the 1000 genomes phase 3 European genotype dataset [24] and VCFtools [27] . The GWAS catalogue [21] was used to determine the number of phenotypes that have been associated with the set of variants in LD with each eQTL , and redundancy among phenotypes was removed by manually assessing the phenotypes linked to each eQTL and collapsing duplicates or analogous phenotypes ( e . g . weight and BMI ) . Logistic regression was used to test for an association between master-eQTLs and their associated number of phenotypes . The response variable being 0 for eQTLs linked to only one gene and 1 for master-eQTLs . The total number of variants in LD with the corresponding eQTL was fitted as a covariate alongside the total number of associated phenotypes to account for any differences in the extent of LD around master-eQTLs and single gene eQTLs .
|
Where a gene’s regulatory site is disrupted by a genetic variant , its expression levels will vary between individuals depending on the version of the variant they carry . Such genetic loci , termed eQTLs , have been found to be disproportionately associated with disease and have proven to be a powerful tool for identifying the location of regulatory variation in the human genome . Despite the widespread study of eQTLs , the focus has , to date , largely been on the effect of individual variants on individual genes . The extent to which individual variants are linked to the expression of multiple nearby genes ( suggestive of such variants tagging shared regulatory sites ) is largely unknown . By first removing redundancy among eQTLs and then examining their effects across tissues and how they relate to how the genome folds , we show that cis-eQTL sharing is a relatively common phenomenon . We show that these master-eQTLs are not only linked to the way in which the human genome has evolved , but are also relevant to the understanding of the co-occurrence of diseases in individuals . These variants are more often linked to multiple diseases and phenotypes , which suggests that these locations are key hubs for understanding human disease risk .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"genome",
"evolution",
"gene",
"regulation",
"human",
"genomics",
"mathematics",
"epigenetics",
"mammalian",
"genomics",
"chromatin",
"discrete",
"mathematics",
"combinatorics",
"chromosome",
"biology",
"gene",
"expression",
"molecular",
"evolution",
"evolutionary",
"genetics",
"animal",
"genomics",
"permutation",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"genomics",
"evolutionary",
"biology",
"computational",
"biology"
] |
2017
|
Shared regulatory sites are abundant in the human genome and shed light on genome evolution and disease pleiotropy
|
Chronic obstructive pulmonary disease ( COPD ) is a leading cause of global morbidity and mortality and , whilst smoking remains the single most important risk factor , COPD risk is heritable . Of 26 independent genomic regions showing association with lung function in genome-wide association studies , eleven have been reported to show association with airflow obstruction . Although the main risk factor for COPD is smoking , some individuals are observed to have a high forced expired volume in 1 second ( FEV1 ) despite many years of heavy smoking . We hypothesised that these “resistant smokers” may harbour variants which protect against lung function decline caused by smoking and provide insight into the genetic determinants of lung health . We undertook whole exome re-sequencing of 100 heavy smokers who had healthy lung function given their age , sex , height and smoking history and applied three complementary approaches to explore the genetic architecture of smoking resistance . Firstly , we identified novel functional variants in the “resistant smokers” and looked for enrichment of these novel variants within biological pathways . Secondly , we undertook association testing of all exonic variants individually with two independent control sets . Thirdly , we undertook gene-based association testing of all exonic variants . Our strongest signal of association with smoking resistance for a non-synonymous SNP was for rs10859974 ( P = 2 . 34×10−4 ) in CCDC38 , a gene which has previously been reported to show association with FEV1/FVC , and we demonstrate moderate expression of CCDC38 in bronchial epithelial cells . We identified an enrichment of novel putatively functional variants in genes related to cilia structure and function in resistant smokers . Ciliary function abnormalities are known to be associated with both smoking and reduced mucociliary clearance in patients with COPD . We suggest that genetic influences on the development or function of cilia in the bronchial epithelium may affect growth of cilia or the extent of damage caused by tobacco smoke .
Chronic obstructive pulmonary disease ( COPD ) is a leading cause of global morbidity and mortality [1] and whilst smoking remains the single most important risk factor , it is also clear that COPD risk is heritable [2] . The genetics underlying COPD are still not fully understood although genome-wide association studies have identified 26 genomic regions showing robust association with lung function [3]–[6] and , of these , 11 have also now shown association with airflow obstruction [7]–[9] . However , the proportion of the variance accounted for by the 26 common genetic variants representing these regions remains modest ( ∼7 . 5% for the ratio of forced expired volume in 1 second ( FEV1 ) to forced vital capacity ( FVC ) ) [5] . Although over a quarter of the population with a significant smoking history go on to develop COPD [10] , some individuals are observed to have preserved lung function as measured by a normal or high FEV1 despite many years of heavy smoking . We hypothesised that these “resistant smokers” may harbour rare variants with large effect sizes which protect against lung function decline caused by smoking . Identification of these variants , and the genes that harbour them , could provide further insight into the mechanisms underlying airflow obstruction . We undertook whole exome re-sequencing of 100 heavy smokers ( >20 pack years of smoking ) who had healthy lung function when age , sex , height and amount smoked were taken into account . We employed 3 complementary approaches to investigate the genetic architecture of the resistant smoker genotype ( Figure 1 ) . Firstly , we screened these 100 “resistant smokers” for novel rare variants ( i . e . not previously identified and deposited in a public database ) with a putatively functional effect on protein product and tested for enrichment of these novel variants in functionally related genes and pathways . Secondly , using a comparision with two independent control sets with exome re-sequencing data , we looked for signals of association with the resistant smoker phenotype for individual variants ( including variants of all minor allele frequencies ) . Thirdly , we looked for association of the resistant smoker phenotype with the combined effects of multiple rare and common variants within genes . We found the strongest evidence of association with resistance to smoking for a non-synonymous variant in CCDC38 , a gene encoding a coiled-coil domain protein with a role in motor activity , previously identified as showing an association with lung function . We also show evidence of cytoplasmic expression of CCDC38 in bronchial columnar epithelial cells . In addition , we found evidence for an enrichment of novel rare functional variants in resistant smokers in gene pathways related to cilia structure and function . Given that abnormalities of ciliary function are already known to play a role in reduced mucociliary clearance in COPD sufferers and smokers , these data suggest that genetic factors may play a significant role in determining the ciliary response of the airway to inhaled tobacco smoke .
100 individuals from the Gedling [11] , [12] and Nottingham Smokers cohorts with good lung function ( FEV1/FVC>0 . 7 and % predicted FEV1>80% ) when age , sex , height and smoking history ( >20 pack years ) were taken into account were selected as “resistant smoker” cases . Characteristics of the 100 resistant smoker case samples are shown in Table S1A and Figure S1 . Exome re-sequencing and alignment was undertaken as described in the methods . Two independent control sets were used; the robustness of findings using the primary control set ( n = 166 ) were further assessed using a secondary control set ( n = 230 ) . We searched for novel variants among the resistant smokers , i . e . genetic variants which were not observed in either control set and which were not documented in public databases . Bioinformatic tools allow for scoring of likely functional impact , including whether a variant is likely to be “deleterious”; here we use the term “putatively functional” since some variants which have a deleterious effect on the function of a given gene may result in a protective phenotype . A total of 24 , 098 variants which were not already in databases of known variants or within segmental duplications were identified with high confidence using two independent calling algorithms . A total of 6587 coding SNPs were scored using CAROL ( including non-synonymous , loss/gain of stop codon , synonymous and splice site/UTR variants ) and 1722 were predicted as being putatively functional ( CAROL score>0 . 98 ) and were within 1533 genes . 16 of these 1533 genes each contained three novel putatively functional variants ( Table S2 ) ( no gene contained more than three such variants ) . GBF1 contained three novel putatively functional variants of which one , chr10:104117872 , was identified in two case samples . A further 157 genes each contained 2 novel putatively functional variants and the remaining 1360 genes contained one novel putatively functional variant . In the resistant smokers , there was no overall enrichment or depletion of novel putatively functional variants among the 26 regions reported to be associated with lung function [5] , ( 16 were observed , the same number would have been predicted based on the sequence length of exons ) and no novel putatively functional variants were identified within the CHRNA3/5 region which has been previously associated with smoking [13] and airflow obstruction [9] . Eight of the 1722 novel variants predicted to be putatively functional were identified in >1 case sample . These are listed in Table S3 . ATAD3C contained a novel putatively functional variant for which six case samples were heterozygous , SHANK2 contained a novel putatively functional variant for which three cases were heterozygous , and the remaining six genes each contained such a variant for which two cases were heterozygous . One hundred and ninety two Gene Ontology ( GO ) terms reached nominal significance for the set of 1533 genes containing novel putatively functional variants in resistant smoker cases . Of these , 22 high level GO terms were significant after Bonferroni correction for multiple testing and are listed in Table 1 [14] . The most significant GO term was the molecular function term “motor activity” which describes molecules involved in catalysis of movement along a polymeric molecule such as a microfilament or microtubule , coupled to the hydrolysis of a nucleoside triphosphate . Other related GO terms also feature amongst the significant signals from this analysis ( including “cytoskeleton” , “microtubule motor activity” , “myosin complex” , “axoneme” , “cilium” and “cilium part” ) ( Table 1 and Table S4 ) . We tested for association of known and novel exonic variants with the resistant smoker phenotype . After exclusion of variants which were missing in >5% of either cases or controls , 269 , 822 ( of which 215 , 747 were listed in dbSNP137 ) variants remained . Of the 269 , 822 variants , 94 , 138 were exonic and included in further analyses . Similar distributions of variants across the minor allele frequency spectrum were observed for the cases , primary , and secondary controls ( results not shown ) . After testing for association with resistant smoker status using primary controls , no SNPs reached genome-wide significance ( P<5×10−7 , based on Bonferroni correction for 94 , 138 tests ) . Substantial under-inflation of the test statistics was seen ( lambda = 0 . 6 ) ( Figure 2A ) , possibly due to the large number of rare variants ( lambda = 0 . 92 if only variants with MAF>5% [n = 25 , 646] were considered , Figure 2B ) . Twenty exonic SNPs showed nominal evidence of association with P<10−3 ( Table 2 ) . The strongest signal from a non-synonymous SNP was within a region previously identified as being associated with lung function [5] . The non-synonymous SNP in CCDC38 ( rs10859974 , OR = 2 . 36 , P = 2 . 34×10−4 ) is 17 . 43 kb away from , but statistically independent of , rs1036429 ( intronic , r2 = 0 . 064 ) which has previously shown genome-wide significant association with FEV1/FVC [5] . SNP rs10859974 itself has shown weak evidence of association with FEV1/FVC ( P = 0 . 032 ) [5] . This SNP is predicted to cause a methionine to valine substitution at protein position 227; the valine allele is predicted to be protective . Investigations into CCDC38 expression in bronchial tissue via immunohistochemistry identified moderate staining of CCDC38 in the cytoplasm of columnar epithelial cells , with weak staining in the sub-epithelial layer ( Figure 3 ) . We found no evidence that rs10859974 or any of its proxy SNPs ( r2>0 . 3 ) were lung eQTLs for CCDC38 itself , although rs11108320 which is intronic in CCDC38 and in strong LD with rs10859974 ( r2 = 1 ) is an eQTL for nearby gene NTN4 ( significant at 10% False Discovery Rate ( FDR ) threshold ) . Many additional SNPs located near or within CCDC38 and SNRPF were eQTLs for NTN4 ( Table S5 ) . Nearby CCDC38 intronic SNPs in weaker LD ( r2 = 0 . 3 ) with rs10859974 were eQTLs for SNRPF ( Table S5 ) . The strongest signal of association in the single-variant analysis was from a synonymous SNP , rs1287467 , in SH3BP5 ( OR = 0 . 47 , P = 1 . 47×10−4 ) ( Table 2 ) . A SNP downstream of SH3BP5 ( rs1318937 , 1000G CEU MAF = 0 . 108 , 16 kb from rs1287467 , r2 = 0 . 018 ) has shown evidence of association with alcohol dependence and alcohol and nicotine co-dependence [15] . Synonymous SNP rs2303296 in ITSN2 was the second strongest signal of association ( OR = 0 . 45 , P = 2 . 31×10−4 ) and had previously shown weak evidence of association with FEV1 ( P = 0 . 02 ) [5] but was not near to any previously identified genome-wide significant associations with lung function and has not shown evidence of association with COPD [9] . Another SNP in ITSN2 , rs6707600 ( intronic , 1000 G CEU MAF = 0 . 017 , 89 Kb from rs2303296 , r2 = 0 . 02 ) , has shown some evidence of association with antipsychotic response in schizophrenia patients [16] . The second strongest signal from a non-synonymous SNP was rs4850 in UQCRC2 ( OR = 4 . 87 , P = 2 . 4×10−4 ) . There were no nearby associations reported with any other trait for this gene . The third strongest signal from a non-synonymous SNP was rs2297950 ( OR = 0 . 51 , P = 6 . 65×10−4 ) in CHIT1 which encodes chitinase 1 ( Chit1 ) . The chitinase pathway has been implicated in asthma and lung function [17] and lung function decline in COPD patients [18] . Chit1 expression in mice has been shown to be correlated with severity of bleomycin-induced pulmonary fibrosis ( with overexpression leading to increased severity and Chit1−/− mice exhibiting reduced pulmonary fibrosis ) [19] . A non-synonymous SNP in LOXL3 , rs17010021 , was the only SNP with an association P<10−3 regardless of whether the primary or the secondary controls were used ( Table S6 ) . This variant had a minor allele frequency of 0 . 048 and 0 . 061 in the primary and secondary control sets respectively , but the minor allele was not observed in any of the resistant smoker cases . Synonymous SNP rs1051730 , in CHRNA3 ( 15q25 . 1 ) , has previously shown very strong evidence of association with smoking behaviour ( particularly cigarettes per day ) [13] , [20] , [21] . This SNP showed weak evidence of association with the resistant smoker phenotype in our study ( P = 0 . 03 when the secondary control set was used and P = 0 . 06 when primary control set was used ) . Association results for SNPs within 500 Kb of rs1051730 are in Table S7 . No nominally significant enrichment of association signals in known pathways was identified in the exome-wide results of the single-variant analysis using MAGENTA [22] . SKAT [23] and AMELIA [24] analyses were undertaken to assess whether multiple variants within a gene collectively showed evidence of association; these tests are agnostic to whether a given variant is previously known . Quantile-Quantile plots for SKAT and AMELIA analyses are shown in Figure 4 . Genes with nominally significant association ( P<10−3 ) for SKAT or AMELIA analysis using the primary controls are shown in Table 3 ( results of SKAT and AMELIA analyses using the secondary controls are shown in Table S8 ) . No genes showed significant association after Bonferroni correction for multiple testing ( P<0 . 05/18000 = 2 . 8×10−6 ) for either analysis ( Table 3 and Table S8 ) . Since the genes are likely to be correlated ( through LD structure or overlapping reading frames ) , SKAT provides a resampling function to control Family Wise Error Rate ( FWER ) . No genes were significant after controlling FWER = 0 . 05 . None of the genes in Table 3 and Table S8 were within any of the 26 lung function associated regions [3]–[5] , the CHRNA3/5 smoking-associated region [13] or SERPINA1 ( mutations in which are known to cause alpha-1-antitrypsin deficiency ) [25] . We also checked overlap between the gene-based association testing and single-variant tests . A signal in TMEM252 ( which showed P<10−3 in the SKAT analysis regardless of which control set was used ) was driven by rs117451470 , a non-synonymous SNP , which had P = 2 . 2×10−3 in the single-variant association analysis ( the other SNP in TMEM252 , a singleton novel synonymous variant , had P = 0 . 38 in the single-variant analysis ) . Signals in UQCRC2 ( strongest signal using SKAT ) , SPATA3D1 , PGAP3 and ADCK2 were also driven by variants with P<10−3 in the single-variant analysis . Signals from TMX3 , IMPG2 and TCOF1 were not driven by single-variant signals ( all SNPs within these genes had P>0 . 01 in the single variant analysis ) . IMPG2 was the strongest signal from the AMELIA analysis and all 8 SNPs within IMPG2 had no evidence of association in the single-variant analysis ( P> = 0 . 18 ) . We tested for enrichment of GO terms within the set of genes showing association with P<0 . 01 in the SKAT analyses . Ten high level GO terms reached nominal significance ( P<0 . 05 ) for the set of 150 genes identified using SKAT but none were significant after Bonferroni correction for multiple testing [14] .
Understanding why some heavy smokers seem to show resistance to the detrimental effects of cigarette smoke on lung function should provide further insight into the genetics of lung function and COPD . We undertook pathway enrichment , single-variant association testing and gene-based association testing analyses on whole exome re-sequencing data from a set of resistant smokers . Although no individual SNP achieved genome-wide statistical significance ( P<5×10−7 ) , our strongest association signal for a non-synonymous SNP was in CCDC38; a gene which has previously shown strong and robust evidence of association with lung function [5] . The intronic SNP previously shown to be associated with lung function ( FEV1/FVC ) and the non-synonymous SNP showing nominally significant association with the resistant smoker phenotype in this study are located close together ( 17 . 4 kb apart ) but are not well correlated ( the non-synonymous SNP has previously shown nominally significant evidence [P<0 . 05] of association with FEV1/FVC ) . A conditional analysis of these two SNPs was consistent with no statistical correlation between these signals . Although the function of CCDC38 is not yet well understood , members of the coiled-coil domain protein family are known to have a role in cell motor activity ( e . g . myosin ) [26] and cilia assembly [27] , [28] . Expression of CCDC38 has been identified in the human bronchi of two subjects , with strong cytoplasmic staining in the epithelium and moderate staining in the airway smooth muscle ( Human Protein Atlas [http://www . proteinatlas . org] [29]: ENSG00000165972 ) . We experimentally confirmed these findings using immunohistochemistry on lung sections . We observed moderate cytoplasmic CCDC38 staining in bronchial columnar epithelial cells and some potential airway smooth muscle staining . There is no evidence that SNP rs10859974 is an eQTL for CCDC38 itself , although proxies for rs10859974 are eQTLs for a nearby downstream gene , NTN4 , encoding Netrin-4 which may play a role in embryonic lung development [30] . Gene Ontology terms shown to be significantly enriched among the novel putatively functional variants identified only in the resistant smokers also pointed to pathways relating to motor activity and the cytoskeleton , including cilia . Another locus showing association with lung function ( 1p36 . 13 , [5] ) also contains a gene encoding a component of cilia ( CROCC which encodes rootletin , another coiled-coil domain protein ) and Crocc-null mice have been shown to have impaired cilia with pathogenic consequences to the airways [31] . The enrichment of genes involved in cilia function amongst the results of our analyses supports the importance of cilia function in lung health . Cilia abnormalities are known to be associated with smoking [32] , [33] , asthma [34] , and play a role in COPD [35] where reduced cilia function leads to reduced mucus clearance of the airways . Improving mucociliary clearance is one of the aims of drug therapy for chronic bronchitis in COPD patients ( reviewed in [36] ) . Impaired cilia function is known to cause a wide range of diseases ( collectively known as ciliopathies ) many of which include pulmonary symptoms [37] . Primary Ciliary Dyskinesia ( PCD ) is a rare genetic disorder where respiratory tract cilia function is impaired leading to reduced ( or absent ) mucus clearance . Mutations in genes which encode components of the cilia have been found to cause several forms of PCD and include the dynein , axonemal heavy chain encoding genes DNAH11 [38] , [39] and DNAH5 [40] within which resistant smoker-specific novel putatively functional variants were identified in this study ( 2 such variants were discovered in DNAH11 ) . Whilst PCD affects resistance to infection and results in bronchiectasis , abnormal lung function can manifest early in life and progressive airflow obstruction has been observed in later life , although aggressive treatment may prevent the latter [41] . Retinitis pigmentosa is a feature of several ciliopathies , including some with pulmonary involvement ( for example , Alstrom Syndrome ) . Low frequency variants in IMPG2 ( interphotoreceptor matrix proteoglycan 2 ) collectively showed strong evidence of association ( using AMELIA ) . Variants in IMPG2 are associated with a form of retinitis pigmentosa [42] . Another retinitis pigmentosa gene , RP1 , was amongst the 16 genes containing 3 novel putatively functional variants in the resistant smokers . RP1 encodes part of the photoreceptor axoneme [43] , a central component of cilia . A recent study identified modulators of ciliogenesis using a high throughput assay of in vitro RNA interference of 7 , 784 genes in human retinal pigmented epithelial cells ( htRPE ) and identified 36 positive modulators and 13 negative modulators of ciliogenesis [44] . These modulators included many genes which did not encode structural cilia proteins and thus were not obvious candidates for a role in cilia function . None of the genes highlighted by the single-variant or gene-based analyses were confirmed as modulators of ciliogenesis although ITSN2 , which contained one of the top signals in our single-variant analysis , was included in the screen and showed suggestive evidence of a positive role in ciliogenesis but this was not confirmed in a second screen . Two of the genes found to harbour a novel putatively functional variant in the resistant smokers were identified as positive modulators of ciliogenesis: GSN ( gelsolin ) which is a known cilia gene with a role in actin filament organisation and AGTPBP1 ( ATP/GTP binding protein 1 ) which has a role in tubulin modification . Collectively , our data show an enrichment of novel putatively functional variants in genes related to cilia structure and function in resistant smokers . Association between smoking and shorter cilia has been reported [32] . The largest genome-wide association with lung function to date supports the notion that the majority of associated variants , including those associated with COPD risk , affect lung function development rather than decline in lung function in adults [5] . If confirmed in other studies , it would be interesting to assess whether genetic influences on the function of cilia primarily affect growth or whether these affect more directly the extent of damage caused by tobacco smoke . Very large GWAS have identified up to hundreds of common variants each with a modest effect on a variety of phenotypes . However , collectively , these still only explain a very modest proportion of the additive polygenic variance . It has been widely hypothesised that rare variation may account for some of this missing variance [45] . Commercially available SNP arrays have tended to include mostly variants with minor allele frequencies upwards of 5% and rare variants have not been reliably imputed from these . Re-sequencing approaches provide the most accurate platform for the study of exome-wide and genome-wide rare variation . However , there is increasing evidence that rare variants may not account for the missing heritability for all traits [46] . Our study did not find evidence for any individual rare variants with large effects in any of the known lung function associated loci or elsewhere in the exome ( albeit in a modest overall sample size ) , although we did identify significant enrichment of novel rare variants in sets of genes with known functions in pathways which are known to have a role in lung health . For the single variant analyses , we used Fisher's Exact Test . Whilst this is an appropriate test to use for small cell counts ( for example , where minor allele counts are low ) , alternatives have been recently proposed including the Firth test , and although the optimal approach in the size of study we undertook is not clear from the comparisons shown to date , the Fisher's Exact Test can be more conservative than the Firth test and this may have had some impact on the power of the study [47] . Methods for the analysis of rare variant data are continuing to evolve . Although this is the first exome re-sequencing study of resistance to airways obstruction among heavy smokers , our study does have potential limitations . Sample size was limited both by availability of individuals with such an extreme phenotype as that we were able to study , and also by current sequencing costs . We were able to utilise re-sequencing data available to the scientific community as control data and therefore maximise the discovery potential of our resources by re-sequencing to a sufficient sequencing read depth for confident rare variant calling . By doing so , and selecting an extreme phenotype group from our sampling frame , we adopted a suitable design to test whether there was enrichment of rare variants of large effect in resistant smokers . The same limitations also impact on the availability of suitable replication studies . In particular , it would have been desirable to undertake replication to support the statistically significant findings of the pathway analysis . However , in the absence of a suitable replication resource , the prior evidence for the role of cilia in lung health does lend support to our findings . As it becomes possible to sample and re-sequence from very large biobanks it should become possible to circumvent these issues in years to come , particularly if the cost of sequencing falls . As limited information was available on smoking status among the controls , we did not restrict controls to heavy smokers and there is therefore potential for genetic associations to be driven via an effect on smoking behaviour . Nevertheless , our design is also consistent with the detection of association due to primary effect on airways and previous genome-wide association studies of lung function not fully adjusted for smoking have detected loci associated with lung function and COPD which were not associated with smoking behaviour [4] , [5] . We saw only a weak association with variants at the CHRNA3/CHRNA5 locus ( the locus at which variants have shown the strongest effect with smoking behaviour [13] , [20] , [21] ) . Misclassification impacts on power; we would have underestimated the power to detect SNP and gene-based associations if the prevalence of resistance to airways obstruction among heavy smokers was greater than we assumed . In a cross-sectional study of this kind , survivor bias could occur if genetic variants influencing survival were under-represented or over-represented in the resistant smokers , but as the mean age of the resistant smokers was 56 . 4 , any survivor bias , if present , is unlikely to have had a major impact . Finally , although we would expect the allele frequencies of the control sets we used to be representative of a general population control set across the vast majority of the genome , biases could potentially be introduced for any genetic variants related to the ascertainment strategy of the control sets . For the main findings we report in this paper , we also present allele frequencies from a public database ( 1000 Genomes Project ) ; any such bias does not explain our main findings . In the first deep whole exome re-sequencing study of the resistant smoker phenotype , we have identified an association signal in a region that has already shown robust association with lung function ( CCDC38 ) and demonstrate significant enrichment of novel putatively functional variants in genes related to cilia structure . These findings provide insights into the mechanisms underlying preserved lung function in heavy smokers and may reveal mechanisms shared with COPD aetiology .
The Gedling study was approved by the Nottingham City Hospital and Nottingham University Ethics committees ( MREC/99/4/01 ) and written informed consent for genetic study was obtained from participants . The Nottingham Smokers study was approved by Nottingham University Medical School Ethical Committee ( GM129901/ ) and written informed consent for genetic study was obtained from participants . The Edinburgh MR-psychosis sample set was compliant with the UK10K Ethical Governance Framework ( http://www . uk10k . org/ethics . html ) and no restrictions were placed on the use of the genetic data by the scientific community . For TwinsUK , ethics committee approval was obtained from Guy's and St Thomas' Hospital research ethics committee . Tissue for immunohistochemistry was from Nottingham Health Science Biobank ( Nottingham , UK ) with the required ethical approval ( 08/H0407/1 ) . For lung eQTL datasets: At Laval , lung specimens were collected from patients undergoing lung cancer surgery and stored at the “Institut universitaire de cardiologie et de pneumologie de Québec” ( IUCPQ ) site of the Respiratory Health Network Tissue Bank of the “Fonds de recherche du Québec – Santé” ( www . tissuebank . ca ) . Written informed consent was obtained from all subjects and the study was approved by the IUCPQ ethics committee . At Groningen , lung specimens were provided by the local tissue bank of the Department of Pathology and the study protocol was consistent with the Research Code of the University Medical Center Groningen and Dutch national ethical and professional guidelines ( “Code of conduct; Dutch federation of biomedical scientific societies”; http://www . federa . org ) . At Vancouver , the lung specimens were provided by the James Hogg Research Center Biobank at St Paul's Hospital and subjects provided written informed consent . The study was approved by the ethics committees at the UBC-Providence Health Care Research Institute Ethics Board . 100 individuals with prolonged exposure to tobacco smoke and unusually good lung function ( resistant smokers ) were selected from the Gedling and Nottingham Smokers studies , described below . The Gedling cohort is a general population sample recruited in Nottingham in 1991 ( 18 to 70 years of age , n = 2 , 633 ) [11] and individuals were then followed-up in 2000 ( n = 1346 ) when blood samples were taken for DNA extraction , and FEV1 and FVC were measured using a calibrated dry bellows spirometer ( Vitalograph , Buckingham , UK ) , recording the best of three satisfactory attempts [12] . The Nottingham Smokers cohort is an ongoing collection in Nottingham using the following criteria; European ancestry , >40 years of age and smoking history of >10 pack years ( currently n = 538 ) . Lung function measurements ( FEV1 and FVC ) were recorded at enrolment using a MicroLab ML3500 spirometer ( Micro Medical Ltd , UK ) recording the best of three satisfactory attempts . Our inclusion criteria was; aged over 40 with more than 20 pack years of smoking and no known history of asthma . A total of 184 samples were eligible for this project after further exclusion of individuals with either FEV1 , FVC or FEV1/FVC less than the Lower Limit Normal ( LLN ) ( based on age , sex and height ) . We calculated residuals after adjusting % predicted FEV1 for pack years of smoking and selected the 100 samples with the highest residuals for exome re-sequencing ( Figure S1 ) . Primary controls were from the Edinburgh MR-psychosis set ( n = 166 ) of the UK10K project ( http://www . uk10k . org/ ) and consisted of subjects with schizophrenia , autism or other psychoses , and with mental retardation . No additional phenotype information was available for the primary controls . The TwinsUK secondary control samples ( n = 230 ) were all unrelated females selected from the high and low ends of the pain sensitivity distribution of 2500 volunteers from TwinsUK [48] , [49] . Characteristics of the secondary controls are given in Table S1B ( note that phenotype information was only available for a subset of the samples ) . These secondary controls were not included in the main analyses due to the difference in exome coverage . Further phenotype information was not available for either control sample set . For the 100 resistant smoker case samples , DNA was extracted from whole blood and the Agilent SureSelect All Exon 50 Mb kit was used for enrichment . The 100 resistant smoker samples were individually indexed and grouped into 20 pools of 5 samples . Each pool was sequenced in one lane ( 20 sequencing lanes in total ) using an Illumina HiSeq2000 . Sequences were generated as 100 bp paired-end reads . Exome-wide coverage of 97 out of 100 samples was >20 ( Figure S2 ) . Three samples had mean sequence depth coverage <20 , of these , one appeared to have had poor enrichment ( high number of off-target reads ) , one had a low overall sequence yield and high number of duplicate reads and one had a high number of duplicate reads ( but good sequence yield ) . To preserve power , and because there was no evidence that the sequence data quality for these samples was lower than for the other samples , these 3 samples were not excluded from further analyses . A total of 166 exomes from the Edinburgh MR-psychosis study: a subset of the neurodevelopmental disease group of the UK10K project ( http://www . uk10k . org/ ) , were used as primary controls . These were enriched using the Agilent SureSelect All Exon 50 Mb kit and sequenced using an Illumina HiSeq2000 to a mean coverage depth of ∼70x ( 75 bp paired-end reads ) . The sequencing of the secondary controls from the TwinsUK pain study has been described elsewhere [49] . In brief , raw sequence data was available for 230 exomes which had been enriched using the NimbleGen EZ v2 ( 44 Mb ) array and sequenced on an Illumina HiSeq2000 to a mean depth of coverage of 71x ( 90 bp paired-end reads ) . The sequence alignment of the primary control exomes has been described elsewhere ( http://www . uk10k . org/ ) . The 100 resistant smoker case exomes and 230 TwinsUK controls were aligned using BWA v0 . 6 . 1 [50] with -q15 for read-trimming . Samtools v0 . 1 . 18 [51] was used to convert sort , remove duplicates and index the alignment . bam files . GATK v1 . 4-30 [52] was used to undertake local realignment around indels and to recalibrate quality scores for all 3 datasets . In order to identify novel variant calls in the 100 resistant smoker exomes , GATK and SAMtools mpileup were run on a per sample basis for all 100 exomes . Only bases with a base quality score >20 were included . The variants called were then compared with dbSNP137 , 1000 Genomes Project ( 1000G ) and NHLBI Exome Sequencing Project calls and all known variants were excluded in order to identify novel rare variants which were unique to the 100 resistant smoker exomes . The novel GATK variant calls were then excluded if they had a Phred scaled probability ( QUAL ) score <30 , quality by unfiltered depth ( non-REF ) ( QD ) <5 , largest contiguous homopolymer run of the variant allele in either direction >5 , strand bias >−0 . 1 or Phred-scaled P value using Fisher's Exact Test to detect strand bias >60 . The novel SAMtools mpileup variants were excluded if they had a QUAL<30 , mapping quality <25 or genotype quality <25 . Variants called at sites with a depth of coverage less than 4 or greater than 2000 were also excluded . The intersect of variants which were identified and passed filtering using both GATK and SAMtools mpileup was taken forward for further analysis . CAROL ( http://www . sanger . ac . uk/resources/software/carol/ ) was used to predict the consequence of all coding variants . This method combines the results of the functional scoring tools SIFT and Polyphen2 . SNPs were predicted as being putatively functional if they had CAROL score>0 . 98 . Amino acid changes were predicted using ENSEMBL . For each comparison ( resistant smoker cases vs . primary controls and resistant smoker cases vs . secondary controls ) , variant calling was undertaken across cases and controls together using the GATK v1 . 5-20 Unified Genotyper . Only bases with a base quality score >20 were included . Coverage was down-sampled to 30 ( reads are drawn at random where coverage is greater than 30 ) . This was done to improve comparability between cases and controls and to speed up computation . A minimum QUAL of 30 was used as the threshold for calls . The GATK VQSR approach was used to filter variants across all samples . Variants with QUAL<30 and VQSLOD score equivalent to truth< = 99 . 9 were excluded ( VQSLOD score<2 . 2989 ) . Only single nucleotide polymorphism variants ( SNPs ) were called . There was >99% genotype concordance with genotype array data ( Illumina 660k ) for ∼5000 exonic SNPs with MAF>5% in the 100 resistant smokers . Single-variant association testing was undertaken using the Fisher's exact test for a comparison of resistant smoker cases and primary controls . A secondary comparison of the resistant smoker cases and the secondary controls was also undertaken although results were interpreted with caution due to disparity of the exome coverage at the pre-sequencing enrichment stage between the cases and secondary controls . Two approaches to analyse the effect of multiple variants within genes were used: SKAT ( v0 . 92 ) [23] and AMELIA [24] . Variants were assigned to RefSeq genes using Annovar [53]: a total of 16439 genes were identified as containing variants in the resistant smoker cases vs . primary controls analysis . Analysis with SKAT was undertaken using default weighting to account for the assumption that rare variants are likely to have bigger effect sizes . An alternative method , AMELIA [24] , was run using a subset of the variants with MAF<5% . A total of 18182 genes were identified as containing variants ( with MAF<5% ) , of which 7516 contained more than 4 variants and so could be reliably tested by AMELIA . For both SKAT and AMELIA , only variants which were annotated as exonic , 5′UTR or 3′UTR were included . Power estimates for the identification of novel putatively functional variants in cases only , single-variant association tests and SKAT analysis were undertaken . For a given variant unique to , and with a minor allele frequency of 0 . 005 in , resistant smokers , the probability of identifying at least one copy of the minor allele in 100 such individuals is 0 . 63 ( 0 . 86 for a minor allele frequency of 0 . 01 ) . Estimates of power for the single-variant association tests were undertaken for a sample size of 100 cases ( resistant smokers ) and 166 controls assuming a prevalence of the resistant smoker phenotype of 2% in the controls . Power calculations for detecting single variants were undertaken using Quanto and are shown in Figure S3 . As an example , power to detect a variant with an allele frequency of 0 . 01 and an OR of 10 would be 10% for an alpha level of 5×10−8 , and 81% for an alpha level of 0 . 001 . SKAT power calculations were run using the R package SKAT . The simulated dataset that the R package provides based on the coalescent populations genetic model was used to assess LD and MAF . The “Log” option was used to specify that the logOR distribution varies with allele frequency ( logOR increases as minor allele frequency decreases ) , the effect size of each variant is equal to c|log10 ( MAF ) | , where c is estimated assuming that the maximum OR corresponds to a MAF of 10−4 . It was assumed that no logOR for causal variants was negative ( results were broadly consistent if 20% of the causal variants were assumed to have negative logOR , results not shown ) . One thousand simulations were run for a region length of 17 . 7 kb ( median of all gene lengths analysed in the real data ) , maximum OR of all variants analysed ranged from 5 to 10 , significance ( alpha ) thresholds of 2 . 8×10−6 ( Bonferroni correction for testing of 18 , 000 genes ) and 0 . 01 ( nominal significance threshold used to define genes as input to DAVID Gene Ontology analysis ) were used and the percentage of causal variants with MAF<1% ( only variants with MAF<1% were considered causal ) given were 25% and 50% . Power to detect a region of length 17 . 7 kb with a maximum OR of 10 , assuming that 50% of variants with MAF<1% are causal is 53% for a Bonferroni-corrected significance threshold of 2 . 8×10−6 , and 89% for a nominal significance threshold of 0 . 01 ( Figure S4 ) . We tested for enrichment of Gene Ontology terms and enrichment of signals in known biological pathways within the results of the single-variant , gene-based and case-only analyses . A total of 150 genes had P<0 . 01 in the SKAT analyses ( of these , 28 also had P<0 . 01 in the AMELIA analysis but many genes were not analysed using both SKAT and AMELIA and so only SKAT results , which included all SNPs with no MAF cut-off , were included in this analysis ) . A total of 1533 genes contained novel putatively functional variants in the resistant smoker cases . We tested for enrichment of Gene Ontology categories within each of these gene lists using DAVID [14] with an EASE ( modified Fisher's Exact ) P<0 . 05 . We tested for pathway enrichment within the single-variant association results using MAGENTA v2 [22] . Briefly , MAGENTA tests for deviation from a random distribution of strengths of association signals ( P values ) for each pathway and includes all available exome-wide single-variant association results ( n = 94 , 138 ) . Six databases of biological pathways were tested: including Ingenuity Pathway ( June 2008 , number of pathways n = 92 ) , KEGG ( 2010 , n = 186 ) , PANTHER Molecular Function ( January 2010 , n = 276 ) , PANTHER Biological Processes ( January 2010 , n = 254 ) , PANTHER Pathways ( January 2010 , n = 141 ) and Gene Ontology ( April 2010 , n = 9542 ) . Significance thresholds were Bonferroni corrected for each database . Fixed lung tissue was sectioned and mounted . Slides were treated in Histo-Clear and then re-hydrated using 100% ethanol and 95% ethanol washes . Antigen retrieval was carried out by steaming the tissue samples for 30 minutes in sodium citrate buffer ( 2 . 1 g Citric Acid [Fisons - C-6200-53]+13 ml 2M NaOH [Fisher - S-4880/53] in 87 ml H2O ) . Tissue was then treated with peroxidise blocking solution ( Dako - S2023 ) , followed by treatments with a 1 in 50 dilution of rabbit anti-CCDC38 antibody ( Sigma HPA039305; 0 . 2 mg/ml ) or a 1 in 50 dilution of the Rabbit IgG Isotype control ( Invitrogen 10500C , diluted to 0 . 2 mg/ml ) . Secondary antibody staining and DAB treatment was carried out using the EnVision Detection Systems Peroxidase/DAB , Rabbit/Mouse kit ( Dako – K5007 ) . Tissue was then counterstained with Mayers Hematoxylin solution ( Sigma – 51275 ) before being dehydrated using 95% ethanol and 100% ethanol washes . Slides were mounted using Vectamount ( Vector Laboratories - H-5000 ) . The description of the lung eQTL dataset and subject demographics have been published previously [54]–[56] . Briefly , non-tumor lung tissues were collected from patients who underwent lung resection surgery at three participating sites: Laval University ( Quebec City , Canada ) , University of Groningen ( Groningen , The Netherlands ) , and University of British Columbia ( Vancouver , Canada ) . Whole-genome gene expression and genotyping data were obtained from these specimens . Gene expression profiling was performed using an Affymetrix custom array ( GPL10379 ) testing 51 , 627 non-control probe sets and normalized using RMA [57] . Genotyping was performed using the Illumina Human1M-Duo BeadChip array . Genotype imputation was undertaken using the 1000G reference panel . Following standard microarray and genotyping quality controls , 1111 patients were available including 409 from Laval , 363 from Groningen , and 339 from UBC . Lung eQTLs were identified to associate with mRNA expression in either cis ( within 1 Mb of transcript start site ) or in trans ( all other eQTLs ) and meeting the 10% false discovery rate ( FDR ) genome-wide significant threshold .
|
Very large genome-wide association studies in general population cohorts have successfully identified at least 26 genes or gene regions associated with lung function and a number of these also show association with chronic obstructive pulmonary disease ( COPD ) . However , these findings explain a small proportion of the heritability of lung function . Although the main risk factor for COPD is smoking , some individuals have normal or good lung function despite many years of heavy smoking . We hypothesised that studying these individuals might tell us more about the genetics of lung health . Re-sequencing of exomes , where all of the variation in the protein-coding portion of the genome can be measured , is a recent approach for the study of low frequency and rare variants . We undertook re-sequencing of the exomes of “resistant smokers” and used publicly available exome data for comparisons . Our findings implicate CCDC38 , a gene which has previously shown association with lung function in the general population , and genes involved in cilia structure and lung function as having a role in resistance to smoking .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genome-wide",
"association",
"studies",
"chronic",
"obstructive",
"pulmonary",
"disease",
"psychology",
"medicine",
"and",
"health",
"sciences",
"addiction",
"genetic",
"association",
"studies",
"genome",
"analysis",
"recreational",
"drug",
"addiction",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"drug",
"addiction",
"genetics",
"of",
"disease",
"pulmonology",
"human",
"genetics"
] |
2014
|
Whole Exome Re-Sequencing Implicates CCDC38 and Cilia Structure and Function in Resistance to Smoking Related Airflow Obstruction
|
Dengue virus ( DENV ) has emerged as major human pathogen . Despite the serious socio-economic impact of DENV-associated diseases , antiviral therapy is missing . DENV replicates in the cytoplasm of infected cells and induces a membranous replication organelle , formed by invaginations of the endoplasmic reticulum membrane and designated vesicle packets ( VPs ) . Nonstructural protein 1 ( NS1 ) of DENV is a multifunctional protein . It is secreted from cells to counteract antiviral immune responses , but also critically contributes to the severe clinical manifestations of dengue . In addition , NS1 is indispensable for viral RNA replication , but the underlying molecular mechanism remains elusive . In this study , we employed a combination of genetic , biochemical and imaging approaches to dissect the determinants in NS1 contributing to its various functions in the viral replication cycle . Several important observations were made . First , we identified a cluster of amino acid residues in the exposed region of the β-ladder domain of NS1 that are essential for NS1 secretion . Second , we revealed a novel interaction of NS1 with the NS4A-2K-4B cleavage intermediate , but not with mature NS4A or NS4B . This interaction is required for RNA replication , with two residues within the connector region of the NS1 “Wing” domain being crucial for binding of the NS4A-2K-4B precursor . By using a polyprotein expression system allowing the formation of VPs in the absence of viral RNA replication , we show that the NS1 –NS4A-2K-4B interaction is not required for VP formation , arguing that the association between these two proteins plays a more direct role in the RNA amplification process . Third , through analysis of polyproteins containing deletions in NS1 , and employing a trans-complementation assay , we show that both cis and trans acting elements within NS1 contribute to VP formation , with the capability of NS1 mutants to form VPs correlating with their capability to support RNA replication . In conclusion , these results reveal a direct role of NS1 in VP formation that is independent from RNA replication , and argue for a critical function of a previously unrecognized NS4A-2K-NS4B precursor specifically interacting with NS1 and promoting viral RNA replication .
Dengue virus ( DENV ) , the causative agent of dengue fever , is the most prevalent arbovirus infecting humans worldwide . It is estimated that all four serotypes of DENV combined are responsible for ~390 million infections annually , leading to ~20 , 000 deaths [1] . Despite many efforts , no antiviral therapy against DENV is available to date and the only approved vaccine has limited efficacy and depends on baseline serostatus of the vaccine recipient [2] . DENV belongs to the Flavivirus genus in the Flaviviridae family and is a small enveloped virus with a single-stranded RNA genome of positive polarity and a length of ~10 , 700 nucleotides . Upon binding to various attachment factors on the cell surface , DENV enters the cell mainly via clathrin-mediated endocytosis , although other entry routes have been described [3] . Upon fusion with the endosomal membrane , the viral RNA is released into the cytoplasm and translated in a cap-dependent manner . The translation product is a polyprotein that is cleaved co- and post-translationally by viral and cellular proteases into 10 proteins . These are the three structural proteins capsid ( C ) , envelope ( E ) and premembrane ( prM ) and the seven nonstructural proteins NS1 , NS2A , NS2B , NS3 , NS4A , NS4B and NS5 . The nonstructural proteins induce massive remodeling of ER membranes , manifesting as convoluted membranes and vesicle packets ( VPs ) . While the function of convoluted membranes is still not clear , VPs most likely are the site of viral RNA replication . Consistently , VPs are clustered ER membrane invaginations with each vesicle connected to the cytoplasm via a ~11 nm pore [4 , 5] . Several enzymatic functions have been identified amongst the nonstructural proteins . These comprise an RNA-dependent RNA polymerase and a methyltransferase activity for NS5 , a serine protease in the amino-terminal region of NS3 , which requires NS2B as a protease cofactor , and helicase , NTPase and RNA triphosphatase activities in the carboxy-terminal region of NS3 . The functions of the small transmembrane proteins NS2A , NS4A and NS4B , which are all essential for virus replication , are much less understood , although they have been proposed to participate in the modification of intracellular membranes and organelles [6–9] and in counteracting host immune response [10–15] . DENV NS1 exerts an amazing array of different functions . On one hand it is required for RNA replication and assembly/release of virus particles [16–19]; on the other hand NS1 is secreted from infected cells as lipid-containing hexamer [20] . This extracellular form of NS1 plays a critical role in immune evasion of the complement system and contributes to dengue pathogenesis , most likely by triggering the release of vasoactive cytokines from immune cells . These cytokines are thought to induce vascular leakage , which is a hallmark of severe dengue [21] . NS1 is inserted into the ER lumen via a 24 amino acid residues long signal sequence corresponding to the carboxy-terminus of E . Upon removal of the signal sequence by the host cell signalase and cleavage by an unknown protease at the NS1-NS2A junction , NS1 rapidly dimerizes [22] . The structure of the NS1 dimer is composed of three distinct domains ( Fig 1A ) : first , a “β-roll” domain ( amino acids 1 to 29 ) composed of two β-hairpins; second , the “Wing” domain ( amino acids 30 to 180 ) formed by the α/β subdomain ( amino acids 38 to 151 ) and the connector subdomains ( amino acids 30 to 37 and 152 to 180 ) ; and third , the central “β-ladder” domain ( amino acids 181 to 352 ) [23] . The connector subdomain and the β-roll domain create a hydrophobic surface likely allowing NS1 interaction with the ER membrane [23] . Although many studies confirmed that NS1 is indispensable for viral RNA replication , the underlying mechanism remains obscure . NS1 colocalizes with dsRNA in infected cells and clusters closely to VPs , presumably at the ER luminal surface of the VPs [24] , or even inside the vesicles [24 , 25] . Given this localization and the ability of NS1 to remodel liposomes , NS1 was proposed to participate in the formation or stabilization of membranous viral replication organelles , possibly by interaction with NS4B and/or NS4A [19 , 23 , 26 , 27] . In addition , multiple cellular proteins interacting with NS1 have been identified , including ribosomal proteins , subunits of the oligosaccharyltransferase and the chaperonin TRiC/CCT complex [28 , 29] , suggesting that functions executed by NS1 are mediated , at least in part , by recruited cellular proteins . With the aim to decipher the mechanism by which NS1 supports the DENV replication cycle we have previously performed a genetic screen and identified a set of mutants that are either impaired in the production of infectious virus particles or RNA replication [17] . While in this previous study we unraveled how NS1 contributes to virus particle assembly and release , in the present study we investigated replication-impaired NS1 mutants with respect to interaction with viral proteins , NS1 secretion and the formation of VPs . We identified a NS1 –NS4B genetic complementation group and determined the viral NS1 interactome . Moreover , we identified a novel interaction between NS1 and the NS4A-2K-4B cleavage intermediate and characterized the role of this interaction for viral RNA replication , NS1 secretion and VP formation . Overall , our results provide a comprehensive map of NS1 determinants required for the multi-functionality of this protein in the DENV replication cycle and demonstrate the indispensable role of NS1 in the formation of the viral membranous replication organelle .
With the aim to establish a genetic complementation map of DENV NS1 that might inform about its viral interaction partners , we took advantage of 18 alanine substitutions in NS1 that severely impair virus replication [17] . We selected for second-site compensatory mutations by using a selectable subgenomic DENV replicon encoding the hygromycin phosphotransferase gene ( sgDVH2A ) ( Fig 1B , left panel ) . Single amino-acid substitutions in NS1 that inhibit RNA replication were inserted into this replicon and in vitro transcribed RNAs were transfected into VeroE6 cells cultured in the presence of hygromycin B . After three to four weeks , growth of well-isolated single cell clones became apparent for mutants Y32A , R62A , D136A , W150A , Y158A , G161A , W168A , K189A , D197A , W311A , P319A , P320A , E334A and R336A , however in the case of W150A , W311A , E334A and R336A no viable cell clones were successfully propagated . For mutants W8A , L153A , S252A and T317A a high number of colonies was obtained and therefore those mutants were excluded from subsequent analyses ( Fig 1B , right panel ) . To identify second-site mutations , total RNA was extracted from single cell clones and cDNA fragments spanning the complete non-structural genome region were amplified by RT-PCR . The PCR products were sequenced , examining at least two independent cell clones for each mutant ( except Y32A , where only one viable cell clone could be isolated ) . Besides the D136A and P319A mutations that consistently reverted to wildtype , and D197A that reverted in one instance , the original NS1 alanine substitution was retained and a second-site mutation was identified on the same amplicon . The only exception was the W168A mutation for which no additional mutations were found in the sequenced region ( Table 1 ) . Interestingly , the majority of second-site mutations mapped to the NS4B coding region ( Fig 1C ) , an observation that is consistent with the proposed role of the NS1—NS4B interaction in the DENV replication cycle [26] . In addition , we detected second-site mutations in NS4A , NS2A and NS5 ( Table 1 ) . With the aim to determine whether those second-site mutations might compensate the replication defect caused by the respective primary NS1 mutation , we inserted each primary mutation together with the respective second-site mutation into a full length DENV-2 genome encoding the Renilla luciferase ( RLuc ) reporter gene ( DVR2A ) , and measured RLuc activity upon transfection of in vitro transcribed RNA into VeroE6 cells ( Fig 2 ) . Consistent with previous data [17] , all primary NS1 mutants were severely impaired in RNA replication , achieving at most 1% of the wildtype replication level at 72 h post-transfection ( Fig 2 ) . Importantly , the replication defects of some NS1 mutants was compensated by second-site mutations in NS4B , while tested mutations in other viral proteins did not rescue replication ( Fig 2 ) . Although the replication fitness of the double mutants was clearly below wildtype level , we observed a partial restoration in the case of the NS1 mutant R62A by NS4B G122R or M172L or S238F , of the NS1 mutant Y158A by NS4B E167H or V219A , and of the NS1 mutant K189A by NS4B F164L or S238F ( Fig 2 ) . In addition , a partial rescue of RNA replication was also found in case of the primary NS1 mutation D197A that was increased by insertion of the pseudoreversion S228C in NS4B . This genetic map supports the proposed role of the NS1—NS4B interaction in the DENV replication cycle [26] and suggests that defects caused by alanine substitutions in NS1 can be compensated , at least in part , by second site mutations in NS4B . Although our forward genetic approach identified NS4B as primary cooperation partner of NS1 , we hypothesized that NS1 might promote viral RNA replication not only via association with NS4B , but also via interaction with other DENV proteins . Therefore , we determined the viral NS1 interactome by using affinity purification followed by liquid chromatography and tandem mass spectrometry analysis of captured complexes . To this end we took advantage of a previously reported trans-complementation system [17] allowing functional tagging of NS1 in the context of viral infection in VeroE6 cells . We adapted this system to Huh7 cells because these cells are of human origin and have been used extensively by us and others to study the DENV replication cycle in detail . Huh7 cells stably expressing HA-tagged or non-tagged NS1 were infected with a DENV reporter virus containing an in-frame deletion within NS1 ( DVR2ApΔNS1 ) ( Fig 3A ) . In this setting , viral replication could be readily detected at 48 h p . i . demonstrating efficient trans-complementation of NS1 also in Huh7 cells ( Fig 3B ) . Taking advantage of this system we isolated NS1 by HA-specific precipitation and analyzed captured protein complexes by mass spectrometry . The specificity of the pull-down was confirmed by western blot from samples prepared in parallel experiments ( Fig 3C ) . Samples from cells expressing non-tagged NS1 were used as control to exclude proteins binding non-specifically to the resin . In addition to several host proteins ( Fig 3D ) , NS4A and NS4B were the only potential viral interaction partners of NS1 identified with this approach ( Fig 3D and 3E ) , with the NS1 –NS4B interaction being consistent with our results from the forward genetic screen . Focusing our analysis on viral NS1 interaction partners , we confirmed them by evaluating HA-captured protein complexes by immunoblot using NS4B-specific antibodies . As shown in Fig 4A ( left panel , lane 1 and 2 ) , NS4B migrates with an apparent molecular weight ( MW ) of ~25 kDa as a double band with the higher , less pronounced , band likely corresponding to the uncleaved 2K-4B form . Surprisingly , none of these two NS4B species could be detected in HA-NS1 captured immune complexes . Instead , two proteins with apparent molecular weights of ~35 kDa and ~30 kDa , both reacting with the NS4B-specific antibody , were consistently detected ( Fig 4A , left panel , lane 6 ) . Both protein species also reacted with a NS4A-specific antibody ( Fig 4A , right panel , lane 6 ) , while fully processed NS4A with an apparent MW of ~10 kDa was only visible in total cell lysates from infected cells , but did not co-precipitate with NS1 ( Fig 4A , right panel , lane 1 and 2 vs 6 ) . This result suggests that one or both of the 30 and 35 kDa protein species likely corresponds to the uncleaved precursor of NS4B , i . e . NS4A-2K-4B . Notably , this putative precursor protein was highly enriched in HA-NS1 immune captured complexes whereas in lysates of infected cells , mature NS4A and NS4B were the predominant species . In order to assure that the observed interaction was not a result of NS1 overexpression in trans , we inserted the HA tag into NS1 in the context of a full length DENV reporter virus genome ( DVR2A-NS1_HA* ) . Consistent with a previous report [25] , epitope-tagged virus was replication competent , although slightly attenuated when compared to the wildtype ( S1 Fig ) . Strikingly , upon HA-specific pulldown the same protein species cross-reacting with NS4A and NS4B specific antibodies were observed ( Fig 4B ) , while no mature form of NS4A or of NS4B could be detected . To exclude any effect of unequal protein levels caused by the attenuation of the epitope-tagged virus , relative to the wildtype , we expressed HA-tagged NS1 in the context of a NS1 to 5 polyprotein by using a construct encoding all NS proteins under control of the T7 RNA polymerase promoter . Upon expression in Huh7-Lunet_T7 cells [30] , the putative NS4A-2K-NS4B precursor was the only protein species enriched in the NS1_HA-specific complexes ( Fig 4C ) . Taken together these results demonstrate that the NS1 –NS4A-2K-4B interaction is not an artifact arising from the individual overexpression of NS1 in trans and occurs also in DENV-infected cells . Aiming to further characterize the putative NS4A-2K-4B precursor species , we next analyzed the interaction between NS1 and various forms of NS4A and NS4B by using transient expression . To this end , Huh7_T7 cells stably expressing DENV-2 NS2B-3 , which is required for polyprotein cleavage , were transfected with equal amounts of HA-tagged NS1 and NS4B or NS4A expression constructs ( Fig 5A ) and cell lysates were subjected to HA-specific immunoprecipitation . As reported previously [31] , trans cleavage of the polyprotein by the DENV protease is possible , albeit with limited efficiency , giving rise to three forms of NS4B-containing proteins in lysates of Huh7_T7_NS2B-3 cells transfected with the NS4A-2K-4B construct ( Fig 5B , left panel , lane 2 and 3 , MW 25–30 kDa ) . Those 3 forms most likely correspond to the full length NS4A-2K-4B precursor , 2K-4B formed after cleavage by the NS2B-3 protease , and mature NS4B released after additional cleavage by cellular signalase , respectively . Comparable to DENV-infected cells , two protein species with apparent MW of ~30 and ~35 kDa reacting with both NS4A- and NS4B-specific antibodies were detected in HA-NS1 immune purified complexes , together with additional higher molecular weight species ( Fig 5B , lower left panel , lane 9 ) . While the theoretical molecular weights for NS4A , NS4B and the NS4A-2K-4B precursor are 16 , 27 and 45 kDa , respectively , an abnormal migration pattern of those transmembrane proteins has been observed by us and others with the precursor having an apparent MW of ~30–35 kDa [31–33] . This is consistent with the migration pattern observed here and close to the MW of the precursor reported by others for DENV and other flaviviruses [33–35] . Therefore , we conclude that the lower band detected in the NS1 precipitates corresponds to the monomeric uncleaved NS4A-2K-4B precursor . Importantly , no interaction with NS1 was detected upon co-expression of NS1_HA and 2K-NS4B or NS4A_FLAG ( Fig 5B , left panel , lanes 11–14 ) . To confirm that all protein species observed in NS1_HA immune captured complexes indeed represent the uncleaved NS4A-2K-4B polyprotein , we performed the same experiment in the absence of the DENV protease by using parental Huh7_T7 cells . Under those conditions the NS4A-2K-4B polyprotein should not be processed , because NS2B-3-mediated cleavage at the NS4A—2K site is a prerequisite for signalase-mediated cleavage between 2K and NS4B [31–33 , 35] As shown in Fig 5B ( right panel , lane 5 ) , NS1_HA immune purified complexes from Huh7_T7 cells contained all the NS4-reactive protein species observed in the presence of the DENV protease . Moreover , the same protein pattern was detected by using a commercially available NS4B-specific antibody and after treatment with PNGase ( S2 and S3 Figs ) . These results corroborate that the NS4-reactive proteins with an apparent molecular weight lower than 50 kDa correspond to incompletely processed forms of the NS4A-2K-4B polyprotein , with the higher ~35 kDa band likely representing an incompletely denatured form of the lower , ~30 kDa species . The higher molecular weight species ( above 60 kDa ) might correspond to lipid-bound NS4 protein species or a heat-stable dimer of the NS4A-2K-4B precursor . The NS4A-2K-4B cleavage intermediate likely contains at least 6 transmembrane helices ( Fig 1C ) [7 , 36] , which could explain this abnormal migration pattern . In summary , these results show that the NS4A-2K-4B precursor , but not mature NS4A or NS4B , constitutes the main interaction partner of NS1 . Having identified an interaction between NS1 and the NS4-2K-4B precursor , we next sought to address the importance of this interaction for viral replication . We predicted that the point mutations in NS1 abrogating the replication might result in loss of binding between NS1 and this cleavage intermediate . To address this hypothesis NS1 point mutants were analyzed in the transient transfection setting described in Fig 5A . As shown in Fig 6 , the replication-impairing G161A and W168A mutations resulted in almost complete loss of the interaction with the precursor polyprotein , arguing that the association between NS1 and NS4A-2K-4B is required for RNA replication . Consistently , two other mutations in NS1 , Y32A and E334A also reduced the interaction , albeit to a lesser extent . Out of those four mutations decreasing the NS1 –NS4A-2K-4B interaction , second-site mutations in NS4A or NS4B were only identified for the NS1 G161A mutant . These second-site mutations had no impact on NS1 –NS4A-2K-NS4B interaction ( S4 Fig ) , which is consistent with the very low replication rescue provided by those pseudoreversions ( maximum 5% of wildtype levels at 72 h post transfection , Fig 2 ) . Several other replication-inactivating NS1 mutations had no statistically significant effect on NS1—4A-2K-4B interaction ( summarized in Table 2 ) suggesting that those NS1 mutants have a replication defect that is independent from the interaction with this cleavage intermediate . In summary , our results identify two residues in the connector region of the Wing domain of NS1 that play a crucial role for NS1 interaction with the NS4A-2K-4B precursor and indicate a previously unappreciated importance of this cleavage intermediate for DENV replication . Earlier studies suggested that mutations within the connector region of the Wing domain might weaken NS1 association with ER membranes [37] , potentially perturbing its secretion . To ensure that the observed loss of interaction with the precursor was not due to increased NS1 secretion , we examined the secretion efficiency for each NS1 mutant . To this end , cells that had been transfected with constructs encoding for various NS1 mutants , and corresponding culture supernatants , were assessed by quantitative western blot ( Fig 7 ) . Of note , the two mutations completely blocking the NS1 –NS4A-2K-4B interaction , i . e . G161A and W168A , did not affect NS1 secretion , indicating that the effect of these alanine substitutions on precursor binding was not due to lower intracellular abundance of NS1 as a result of enhanced NS1 release . Interestingly , several other point mutations , mostly residing in the carboxy-terminal region of the β-ladder domain , i . e . W311A , P319A , E334A and R336A almost completely blocked NS1 secretion ( Fig 7 ) . While this could be due to structural defects possibly resulting in ER-associated degradation of these NS1 variants , intracellular abundance of these proteins was not strongly affected , arguing that they were not targeted for degradation because of misfolding or destabilization . Instead , they might have a defect in self-interaction or routing to the secretory pathway . We next asked whether the alanine substitutions we had inserted into NS1 might have an impact on DENV polyprotein processing , especially the production of fully cleaved NS4B . Therefore , we inserted the alanine substitutions into an expression construct encoding the complete DENV polyprotein that was expressed by a T7 promoter-based system to allow replication-independent protein production ( Fig 8A ) . Huh7-Lunet_T7 cells were transfected with these constructs and 18 h later cells were lysed and NS1 , NS2B , NS3 , NS4B and NS5 expression was detected by western blot . None of the point mutations in NS1 had an obvious impact on polyprotein processing ( Fig 8B ) , at least under steady state conditions , demonstrating that the inhibitory effects exerted by these mutations ( block of RNA replication , NS4A-2K-4B interaction and NS1 secretion ) did not result from improper polyprotein cleavage . One of the proposed mechanisms by which NS1 supports DENV RNA replication is the formation of membranous replication organelles , i . e . the VPs . Since both NS4A and NS4B are transmembrane proteins with membrane remodeling properties [7 , 33] , interaction between NS1 and the NS4A-2K-4B precursor might be important for establishment or stabilization of those membrane structures . Therefore , we wanted to determine whether the two mutations abrogating NS1 interaction with this cleavage intermediate ( G161A and W168A ) also block VP formation . Thus far , the biogenesis of these membranous structures could only be studied in systems supporting viral replication , making it impossible to determine the impact of replication-impairing mutations on VP formation . To overcome this limitation , we employed a polyprotein expression approach ( Fig 9A ) . In this system , the sole expression of the DENV polyprotein in Huh7-Lunet_T7 cells resulted in the formation of membrane invaginations that closely resemble VPs observed during DENV infection ( Fig 9A , left panel ) in ~25% of the cells ( Fig 9B ) . The diameter of the vesicles was ~75 nm ( Fig 9C ) , which is comparable to VP diameter in infected cells [4] . Taking advantage of this system , we next inserted the G161A or W168A mutations into NS1 of this expression construct and determined abundance and morphology of vesicle formation ( Fig 9 ) . Interestingly , none of the mutations had an effect on number or size of the vesicles , demonstrating that the NS1—NS4A-2K-4B interaction is not involved in VP formation . We next asked whether NS1 is at all required for the biogenesis of VPs . To address this question , we used two polyprotein expression constructs . The first contained an in-frame deletion of 97 codons in the NS1 coding region that can be rescued by ectopic expression of NS1 in the context of DENV infection ( partial deletion of NS1 , pΔNS1 ) [17] , and the second a complete deletion of NS1 ( cΔNS1 ) ( Fig 10A ) . In the cΔNS1 construct the first 5 and last 8 amino acid residues of NS1 were retained to ensure proper polyprotein insertion into the ER membrane and processing , respectively [38] . Huh7-Lunet_T7 cells were transfected with either mutant or the wildtype polyprotein expression constructs and protein expression was determined by western blot ( Fig 10B ) . While viral protein abundance was slightly lower in the case of the two NS1 deletion mutants , the ratio between the viral proteins was similar to the wildtype . Importantly , transfection efficiency was comparable ( ~40% as determined by immunofluorescence; Fig 10C ) and the subcellular distribution of NS4B and NS3 was not affected by the NS1 deletions ( S5 Fig ) . Analysis of the cells by transmission electron microscopy ( EM ) revealed that both NS1 deletions abrogated the formation of VPs ( Fig 10D and 10E ) showing that NS1 is required for VP formation . Next , we sought to determine whether expression of NS1 in trans can rescue the VP defect caused by the deletions within this protein . To this end , we transfected Huh7-Lunet_T7 cells stably expressing mCherry-tagged NS1 ( LunetT7_NS1-mCh ) with the pΔNS1 or cΔNS1 polyprotein expression construct and assessed VP formation ( Fig 11A ) . Production of the viral proteins in transfected cells was confirmed by western blot ( Fig 11B ) . Of note , the defect of VP formation caused by the partial deletion in NS1 could be restored by expression of NS1 in trans ( Fig 11C and 11E ) . By contrast , no such rescue could be observed in cells expressing the polyprotein with the complete deletion in NS1 ( Fig 11D and 11E ) . Remarkably , rescue of VP formation by trans-complementation with NS1 correlated with rescue of viral replication ( Fig 11F ) . While both the partial and the complete NS1 deletion abrogated DENV RNA replication , the provision of NS1 in trans rescued the replication of the partial NS1 deletion , but not of the complete NS1 deletion ( Fig 11F ) . In summary , these results demonstrate that NS1 is indispensable for proper formation of the DENV replication organelle , but this function is not related to interaction of NS1 with the NS4A-2K-4B precursor . The fact that the complete NS1 deletion cannot be rescued by trans-complementation argues that NS1 contains both cis- and trans-acting determinants involved in VP formation .
Despite intensive research efforts the role of NS1 in the flavivirus replication cycle remains elusive . NS1 is not involved in viral entry or RNA translation but is essential for RNA replication . This observation has originally been made during studies conducted with the yellow fever virus ( YFV ) , and later confirmed with several other flaviviruses [16 , 17 , 19 , 26 , 39 , 40] . It is thought that NS1 is required for negative RNA strand synthesis [16 , 19] , but the molecular mechanism is not known . In the present study we employed reverse and forward genetic screening , combined with biochemical assays , to identify domains and regions within NS1 that contribute to its role in DENV RNA replication as well as NS1 secretion . With respect to the latter , we observed a clustering of secretion-inhibiting mutations in the highly conserved carboxy-terminal region of the β-ladder , which contains the almost invariant MEIRP motif comprising amino acid residues 333–337 ( Fig 12A ) [41] . Intracellular retention of these mutants , which are also replication deficient , might result from improper NS1 self-interaction , or altered trafficking due to loss of interaction with cellular transport proteins . Interestingly , the carboxy-terminal tip of the β-ladder domain contains multiple epitopes recognized by NS1-specific antibodies , including some that cross-react with cellular surface proteins [23 , 41 , 42] . Hence , this region might represent an attractive drug target , offering the possibility to block viral replication , NS1 secretion and the induction of antibodies possibly contributing to DENV pathogenesis . The results of our genetic studies provide strong evidence for an interaction between NS1 and NS4B , or NS4B-containing cleavage intermediates . This conclusion is derived from the observation that replication-inactivating mutations in NS1 can be rescued , at least in part , by pseudoreversions residing in NS4B . This genetic NS1 –NS4B association was corroborated by the viral NS1 proteome identifying NS4B and NS4A as predominant interaction partners of NS1 . However , further characterization of the NS1—NS4B association revealed that the main interaction partner of NS1 is the NS4A-2K-4B precursor . This interaction , which has so far not been reported , was consistently detected in complexes isolated from DENV-infected cells and in different expression-based approaches . The NS4A-2K-4B intermediate has been previously detected in YFV-infected cells using pulse-chase experiments , where it was found to be processed post-translationally with a half-life of ~10 min [34] . A subsequent study confirmed the production of NS4A-2K-4B also in the course of DENV infection [35] , however its function in virus replication has not been appreciated so far . Based on our findings , and on the high enrichment of NS4A-2K-4B in the NS1-associated protein complexes , we conclude that delayed cleavage of this polyprotein fragment plays an important role for interaction with NS1 as well as RNA replication . Such regulated cleavage has been described for several other positive-strand RNA viruses and also in those cases it critically determines RNA replication [43–45] . For instance , for the related HCV it was found that constitutive cleavage between NS4B and NS5A by insertion of an IRES at the cleavage site completely abrogates RNA replication [43] . Likewise , in the case of alphaviruses processing intermediates and corresponding mature forms were found to play distinct roles in negative and positive RNA strand synthesis , respectively [44] . The importance of various polyprotein precursors is also well established for picornaviruses and their functions include genome circularization [46] , modulation of enzymatic activities and modification of cellular membranes ( reviewed in [47] ) . We speculate that delayed cleavage of DENV NS4A-2K-4B might be required for proper membrane association of NS4A or NS4B , maturation of these proteins e . g . by post-translational modification [48] or the formation of dimeric or oligomeric complexes . Although the exact mechanism remains to be determined , it is tempting to speculate that NS1 binding to NS4A-2K-4B might regulate its cleavage . Two replication-impairing point mutations in NS1 , G161A and W168A , almost completely abrogated NS4A-2K-4B precursor binding , corroborating the specificity of the observed interaction and its critical role in RNA replication . These mutants were stable and efficiently secreted , demonstrating that the loss of interaction was not due to general defects in protein structure , degradation or enhanced NS1 release . Moreover , both mutants were able to form VPs when expressed in the context of the NS1 to 5 DENV polyprotein , implying that subcellular localization and membrane association of NS1 were not affected by these substitutions . Residue G161 localizes to the so called “greasy finger” loop within the Wing domain of NS1 , which has been proposed to mediate NS1 interaction with the ER membrane [23] ( Fig 12B ) . While it might appear surprising that a glycine to alanine substitution had such drastic effects on viral replication and protein-protein interaction , the G159-X-G161 motif is absolutely conserved in flaviviruses [17 , 49] , which is indicative of a critical function of this region in the viral replication cycle . The aromatic amino acid residue at position 168 is also invariant and resides on the NS1—ER membrane interface , contributing to the formation of a hydrophobic “inner face” of the NS1 dimer ( Fig 12B ) . While several other mutants , most notably Y32A and E334A , also showed a trend towards reduced NS1 –precursor interaction ( Fig 6 and Table 2 ) , their impact was much weaker compared to the G161A and W168A substitutions . We therefore assume that replication impairment caused by the Y32A and E334A substitutions is due to defects other than impaired interaction with NS4A-2K-4B . This is also very likely for mutations in the C-terminal region that result in loss of NS1 secretion . Of note , neither the mutation affecting residue W8 , which resides in the β-roll domain and is assumed to be directly involved in NS1—membrane interaction [50] , nor the mutation affecting Y158 directly adjacent to the greasy finger ( S6 Fig ) abrogated NS1 –cleavage intermediate binding . This result further suggests that the phenotype caused by G161A and W168A is not due to general alteration of the NS1—membrane interaction , but rather to abrogation of specific contact sites between NS1 and the NS4A-2K-4B cleavage intermediate that are formed by the two residues . Four of the primary NS1 mutations ( R62A , Y158A , K189A and D197A ) , all residing in the membrane proximal region of the NS1 dimer ( Fig 12C ) , could be compensated by pseudoreversions in NS4B . These primary mutations did not affect NS4A-2K-4B precursor binding , indicating that the observed phenotype was not related to this interaction . Interestingly , all of the pseudoreversions restoring replication were found in the putative transmembrane domains of NS4B ( Fig 1C ) . Given this localization , the primary mutations might impair the membrane remodeling activity of NS1 that might be restored by structural changes in NS4B , consistent with its ability to alter ER membranes as reported for WNV [51] . However , the rescue was only partial suggesting that additional defects , which cannot be restored by NS4B might be caused by these alanine substitutions in NS1 . DENV replication most likely occurs within invaginations in the ER membrane designated VPs [4] . Owing to the lack of systems to study VP biogenesis in the absence of viral replication , little is known about viral and cellular factors involved in their biogenesis [9] . By using transient expression of a DENV polyprotein we were able to analyze VP formation with replication-deficient NS1 mutants . The following important conclusions can be drawn from our EM-based studies: ( i ) NS1 is essential for the establishment of the membranous DENV replication organelle; ( ii ) this function is not related to NS1 interaction with NS4A-2K-4B , because mutations disrupting this interaction did not affect VP formation; ( iii ) at least part of NS1 acts in cis to allow for VP formation as deduced from the observation that only the partial , but not the complete NS1 deletion could be rescued by trans-complementation . Therefore , NS1 plays a more sophisticated role in viral RNA amplification that goes beyond the structuring of replication complex formation as often proposed [21 , 52] . Since the G161A and the W168A mutation impaired viral replication without affecting VP formation , we conclude that NS1 has at least two distinct functions , i . e . the assembly of the membranous replication organelle and the generation of an active replicase that catalyzes the amplification of the viral RNA , the latter steps possibly linked to the interaction of NS1 with the NS4A-2K-4B cleavage intermediate . Of note , Akey et al . reported that mutations targeting residues 159–161 of NS1 were deleterious to viral RNA replication , but did not affect the ability of NS1 to remodel liposomes [23] . This result is in consistent with our notion that this region in NS1 is involved in an RNA replication-relevant process that is independent from the formation of VPs . Studies conducted with YFV and West Nile virus revealed that trans-complementation of NS1 replication defects depends on the degree of the deletion , with at least 54 carboxy-terminal amino acid residues being required in cis for efficient rescue , whereas bigger deletions could not be complemented [16 , 19 , 39 , 53] . This observation is similar to our results and suggests that the carboxy-terminal region might play an important role in a cis-dominant process such as polyprotein cleavage . Alternatively , based on the recent identification of multiple host ribosomal proteins and chaperones as NS1 interaction partners [28] , it is conceivable that this domain might contain binding sites for host cell factors required for stability or folding of the ( immature ) polyprotein , thus contributing indirectly to the formation of a functional replication complex . This hypothesis would be in agreement with the lower levels of some NS proteins observed upon expression of the DENV polyprotein with the complete NS1 deletion . Alternatively , NS1 or an NS1-2A precursor [34] might be involved in stabilizing other NS proteins . Although impaired polyprotein cleavage , resulting from complete NS1 deletion , cannot be excluded , we did not observe alterations in the ratios of polyprotein cleavage products under steady-state conditions . Consistently , it has been reported that the last 8 amino acid residues of NS1 are sufficient for NS1-2A cleavage [54] . In any case , the observation that the partial deletion in NS1 can be rescued by trans-complementation clearly shows that remaining parts of NS1 are also indispensable for VP biogenesis . Though underlying mechanisms remain to be clarified , observed functions of NS1 might result from an intrinsic membrane-bending ability of the central region in NS1 , its affinity for lipids that are recruited to the site of the viral replicase or interactions with host proteins mediating these or some other function . In summary , our study identifies a novel interaction between NS1 and the NS4A-2K-NS4B cleavage intermediate which plays a critical role in the DENV replication cycle . In addition , we provide evidence for a contribution of NS1 to the formation of the membranous replication organelle that is independent from viral RNA replication . Finally , we establish a comprehensive map of regions and domains involved in the various functions of NS1 . The multitude of roles fulfilled by NS1 , including its contribution to the pathogenesis of dengue [21] , identifies NS1 as a highly promising target for direct acting antivirals aiming to suppress viral replication and severe disease manifestations [55] . Although further investigations will be needed to mechanistically define the various functions NS1 exerts in the viral replication cycle , the genetic map established in the present study offers a starting point for the design of antiviral agents targeting this DENV “Swiss Army Knife” [56] .
All cell lines were maintained in Dulbecco's modified Eagle medium ( DMEM; ThermoFisher Scientific , Darmstadt , Germany ) supplemented with 2 mM L-glutamine , nonessential amino acids , 100 U penicillin/ml , 100 μg streptomycin/ml , and 10% fetal calf serum . VeroE6 cells [57] were obtained from Progen ( Heidelberg Germany ) . Huh7 cells [58] were obtained from the laboratory of Heinz Schaller ( Center for Molecular Biology , Heidelberg ) . These cells served as founder for the production of all Huh7 derived subclones . Huh7_T7 and Huh7-Lunet_T7 cells [30] were generated by lentiviral transduction to allow stable expression of the bacteriophage T7 RNA polymerase . Huh7_T7_NS2B3 cells that in addition stably express full length DENV NS2B-3 were described previously [31] . Huh7_NS1_HA and Huh7_NS1_nt were generated by transduction with lentiviral vectors encoding DENV-2 NS1 with a carboxy-terminal HA-tag or non-tagged NS1 , respectively [17] . Huh7-Lunet_T7_NS1-mCh cells stably expressing NS1 with a carboxy-terminal mCherry tag were generated by transduction with lentiviral vectors encoding NS1-mCherry [17]; control Huh7-Lunet_T7_ cells were generated by transduction with the empty pWPI vector . Huh7_T7 and Huh7-Lunet_T7 cells were maintained in medium containing 5 μg/ml zeocin . Huh7-T7_NS2B3 , Huh7-Lunet-T7_NS1-mCh and Huh7-LunetT7_vector cells were cultured in medium containing 5 μg/ml zeocin and 1 μg/ml puromycin and Huh7_NS1_HA and Huh7_NS1_nt cells in medium containing 1 μg/ml puromycin . All cell lines are routinely tested for mycoplasma contamination using the MycoAlert mycoplasma detection kit ( Lonza , Basel , Switzerland ) . The plasmids pFK_DVR2A containing a DENV genome based on the 16881 strain and encoding a Renilla luciferase ( RLuc ) reporter gene as well as pFK_sgDVH2A containing a hygromycin-B selectable DENV subgenomic replicon were described previously [59] . The plasmid pFK_DVR2ApΔNS1 containing an in-frame deletion of 97 amino acids in NS1 was described previously [17] . The DVR2A construct containing an HA-tag within NS1 ( pFK_DVR2A-NS1_HA* ) was based on the insertion site reported earlier [25] and generated by overlap PCR followed by insertion of the amplicon via KasI and MluI restriction sites into pFK_DVR2A . DVR2A constructs containing pseudoreversions identified in this study were created by overlap PCR and insertion of PCR products into wildtype DVR2A or DVR2A containing a specified primary NS1 mutation . All NS1 constructs used in this study contain the last 72 nucleotides of E ( ET ) immediately upstream of the NS1 sequence to ensure proper insertion into the ER membrane and signalase cleavage . Lentiviral expression constructs pWPI_puro_NS1_HA , pWPI_NS1_mCh and pWPI_puro_NS1_nt containing the DENV-2 ( 16681 ) NS1 sequence with or without carboxy-terminal tag were described previously [17] . Expression constructs pTM_NS1_HA ( containing a carboxy-terminal HA-tag ) and pTM_NS1_nt were generated by PCR using pCDNA_NS1_HA and pCDNA_NS1_nt as template , respectively . XmaI and BamHI restricted amplicons were inserted into the pTM expression vector [60] . Expression constructs pTM_NS4A-2K-4B and pTM_2K-4B were described previously [31] . Plasmid pTM_FLAG_NS4A was generated by PCR using pTM_NS4A-2K-4B as template followed by insertion of the NcoI and BamHI restricted PCR products into the pTM vector . NS1 point mutations were inserted into the construct pSM3-DVs_CAE_NS1-3’ encoding the complete DENV polyprotein under the control of the T7 RNA polymerase promoter by using the DVR2A constructs containing the desired mutations and transfer of MluI—KasI DNA fragments or by overlap PCR in the case of mutations located upstream of the MluI site . The polyprotein construct containing an internally HA-tagged NS1 ( pSM3-DVs_CAE_NS1-3’_NS1_HA* ) was generated by inserting the NS1-NS2A fragment isolated from pFK_DVR2A-NS1_HA* via KasI and MluI sites into pSM3-DVs_CAE_NS1-3’ . The polyprotein construct containing a complete NS1 deletion ( pSM3-DVs_CAE_NS1-3’_cΔNS1 ) was obtained by overlap PCR , generating a DNA fragment containing the last 24 codons of E and the first 5 codons of NS1 directly fused to last 8 codons of NS1 that was inserted via BamHI and KasI sites into pSM3-DVs_CAE_NS1-3’ plasmid . The polyprotein construct containing a partial deletion in NS1 ( pSM3-DVs_CAE_NS1-3’_pΔNS1 ) was created by inserting the NS1-NS2A fragment from pFK_DVR2ApΔNS1 via MluI and KasI sites into pSM3-DVs_CAE_NS1-3’ . The plasmid pFK_DVR2AcΔNS1 was generated by overlap PCR using pSM3-DVs_CAE_NS1-3’_cΔNS1 and pFK_DVR2A as templates and insertion of BamHI and KasI digested PCR products into pSM3-DVs_CAE_NS1-3’ . A complete list of primers used in this study is available upon request . Rabbit antisera raised against various DENV proteins ( NS1 , NS2B , NS3 , NS4A , NS4B and NS5 ) have been described previously [4] and were used at a 1:1 , 000 dilution . In addition , the following primary antibodies were used for immunofluorescence staining: rabbit polyclonal anti-NS4B antibody , mouse monoclonal anti-NS3 antibody , mouse monoclonal anti-NS1 antibody ( all from GeneTex , Irvine , CA , USA ) , mouse monoclonal anti-reticulon 3 antibody ( Santa Cruz , Dallas , TX , USA ) , and anti-protein disulfide isomerase rabbit polyclonal antibody ( Cell Signaling Technology , Danvers , MA , USA ) . The mouse monoclonal anti-GAPDH antibody , as well as anti-rabbit and anti-mouse secondary antibodies conjugated to horseradish peroxidase were purchased from Sigma-Aldrich ( Tufkirchen , Germany ) . For in vitro transcription , 5 or 10 μg of plasmid DNA was linearized using the XbaI restriction enzyme and purified using NucleoSpin Gel and PCR Clean-up ( Macherey-Nagel , Düren , Germany ) . In vitro transcription was carried out using the SP6 RNA polymerase as described previously [59] . In vitro transcripts were purified by phenol-chloroform extraction and resuspended in RNase-free water . RNA integrity was confirmed by agarose gel electrophoresis . For RNA transfection VeroE6 or Huh7-derived cells were trypsinized , harvested in complete DMEM and washed once with PBS . Cells were resuspended in cytomix ( 120 mM KCl , 0 . 15 mM CaCl2 , 10 mM potassium phosphate buffer , 2 mM EGTA , 5 mM MgCl2 , 25 mM HEPES [pH 7 . 6] , 2 mM ATP and 5 mM glutathione , the latter two freshly added ) at a density of 1 . 5x107 Vero cells/ml or 1x107 Huh7-derived cells and 400 μl of the cell suspension were mixed with 5 μg of in vitro transcripts . Cells were transferred into a 0 . 2 cm gap width electroporation cuvette ( BioRad Hercules , CA , USA ) , pulsed once with 166 V and 500 μF , resuspended in pre-warmed complete DMEM and seeded as required for subsequent assays . Stocks of the DVR2ApΔNS1 virus containing an in-frame deletion of 97 amino acids in NS1 were produced in VeroE6 helper cells stably expressing NS1 and titrated by focus forming assay as previously described [17] . Virus stocks of DVR2A and DVR2A-NS1_HA* were produced as described previously [31] and titrated by plaque assay . Viral replication was measured in cells infected or transfected with DVR2A by using RLuc assay as described elsewhere [61] . In brief , cells grown on 24-well plates were lysed in 100 μl RLuc lysis buffer ( 25 mM Glycine-Glycine [pH 7 . 8] , 15 mM MgSO4; 4 mM EGTA , 10% ( v/v ) glycerol , 0 . 1–1% ( v/v ) Triton X-100 , 1 mM DTT added right before use ) at indicated time points after transfection/infection , snap frozen at -70°C and thawed prior to use . RLuc activity was measured using a Mithras LB 940 plate reader ( Berthold technologies , Bad Wildbad , Germany ) after addition of 400 μl assay buffer ( 25 mM Glycine-Glycine [pH 7 . 8] , 15 mM K4PO4 buffer [pH 7 . 8] , 15 mM MgSO4 , 4 mM EGTA ) containing 1 . 43 μM of coelenterazine ( PJK , Kleinblittersdorf , Germany ) . In some cases , RLuc activity was measured with a Lumat LB9507 tube luminometer ( Berthold ) after mixing 20 μl of cell lysate with 100 μl of RLuc buffer . Single substitutions in NS1 were inserted into a selectable subgenomic DENV replicon ( sgDVH2A ) containing a hygromycin phosphotransferase gene downstream of the cis-acting elements of the capsid-coding region ( CS ) . A 2A cleavage sequence at the carboxy-terminus of the hygromycin phosphotransferase gene was inserted to allow proper processing of the DENV polyprotein . VeroE6 cells were electroporated with selectable replicon RNAs and cultured in the presence of 150–250 μg/ml Hygromycin B . After three to four weeks , single cell clones were isolated and expanded and once sufficient cell numbers had been reached , total cellular RNA was extracted using the NucleoSpin RNA II kit ( Macherey-Nagel , Germany ) . Viral RNA was reverse transcribed using the SuperScript III reverse transcriptase ( ThermoFisher Scientific ) and the primer 5’-CGA CCT GAC TTC TAG CCT TGT TTC-3’ . cDNA was used to amplify a DNA fragment spanning the coding region of the DENV-2 non-structural proteins ( from nucleotide 2 , 422 to 10 , 248 ) using the Expand Long Template PCR System ( Roche , Mannheim , Germany ) , forward primer 5’-ATT AGA GCT CGA TAG TGG TTG CGT TGT GAG CT-3’ and reverse primer 5’-ATA ATC TAG ACC ACA GAA CTC CTG CTT CTT CC-3’ . Purity and integrity of amplicons was verified by agarose gel electrophoresis and excised fragments were subjected to nucleotide sequence analysis . Target cells were seeded one day prior to transfection to achieve 90–100% confluency at the time point of transfection . After a medium change 30 min before transfection , cells were transfected using the Trans-IT-LT1 transfection reagent ( Mirus , Madison , WI , USA ) , according to the protocol of the manufacturer , except for 10 cm-diameter culture dishes where 10 μg of DNA , 30 μl of Trans-IT-LT1 reagent and 800 μl transfection medium were used . Reduced serum Opti-MEM medium ( ThermoFisher Scientific ) was used for preparing transfection complexes . In the case of co-transfection of two constructs , equal amounts of each plasmid DNA were used to reach a total DNA amount required for the given format . In the case of EM or immunofluorescence analysis , medium was changed 4 h post transfection . Cells were lysed or fixed 16 to 20 h after transfection and processed for subsequent assays . Infected or transfected cells were washed twice with PBS , lysed in 1 to 2 ml of immunoprecipitation ( IP ) lysis buffer ( 150 mM NaCl , 50 mM Tris-HCl [pH 7 . 4] , 0 . 5% Triton-X100 , freshly supplemented with 1% cOmplete protease inhibitor cocktail ( Roche ) ) , and incubated on ice for 1 h . Lysates were centrifuged at 21 , 000 x g for 1 h . In some cases , cells were collected in PBS and dry cell pellets were stored at -80°C prior to processing for cell lysis . Pre-cleared cell lysates were normalized to the sample with lowest total protein concentration as measured by Bradford assay [62]; 10% of total normalized cell lysate was saved as input and the rest was added to 20 μl of HA-specific agarose beads slurry ( Sigma-Aldrich , St . Louis , MO , USA ) . After incubation for 3 to 5 h at 4°C with gentle agitation , the resin was washed extensively with lysis buffer and samples were eluted once with 3% SDS in PBS and once with PBS . Eluates were collected , pooled and subjected to overnight acetone precipitation . Samples were centrifuged for 1 h at 21 , 000 x g , pellets were resuspended in 2 x SDS-PAGE loading buffer ( 230 mM Tris-HCl [pH 6 , 8] , 120 mM SDS , 200 mM DTT , 3 . 5% glycerol , 0 . 1% bromophenol blue ) and denatured by 5 min incubation at 98°C . In the case of DVR2A_NS1_HA*-infected cells the same procedure was employed but using Pierce anti-HA magnetic beads ( Thermo Fisher Scientific ) and direct processing of eluates without the acetone precipitation step . For mass spectrometry ( MS ) analysis , cells were processed for HA-specific affinity purification as described above . After washing with lysis buffer , proteins bound to the resin were washed additionally 3 times in lysis buffer without detergent and protease inhibitors . Bound proteins were denatured by incubation in 20 μl guanidinium chloride buffer ( 600 mM GdmCl , 1mM Tris[2-carboxyethyl] phosphine–HCl , 4mM chloroacetamide , 100 mM Tris-HCl [pH 8 . 0] ) . After digestion with 1 μg LysC ( WAKO Chemicals USA ) at room temperature for 3 h , the suspension was diluted in 100 mM Tris-HCl [pH 8 . 0] , and the protein solution was digested with trypsin ( Promega ) overnight at room temperature . Peptides were purified on stage tips with three C18 Empore filter discs ( 3M , Maplewood , MN , USA ) and analyzed by liquid chromatography coupled to MS on an Orbitrap XL instrument ( Thermo Fisher Scientific ) as described previously [63] . Raw MS data were processed with the MaxQuant software package , version 1 . 5 . 3 [64] using the built-in Andromeda search engine to search against the human proteome ( UniprotKB , release 2012_01 ) containing forward and reverse sequences concatenated with the DENV polyprotein ( Uniprot ID: P-29990 ) with the individual viral open reading frames manually annotated , and the label-free quantitation algorithm as described previously [65] . Additionally , the intensity-based absolute quantification ( iBAQ ) algorithm and Match Between Runs option were used . In MaxQuant , carbamidomethylation was set as fixed and methionine oxidation and N-acetylation as variable modifications , using an initial mass tolerance of 6 ppm for the precursor ion and 0 . 5 Da for the fragment ions . Search results were filtered with a false discovery rate ( FDR ) of 0 . 01 for peptide and protein identifications . The Perseus software package , version 1 . 5 . 3 . 0 was used to further process the data . Protein tables were filtered to eliminate the identifications from the reverse database and common contaminants . In analyzing MS data , only proteins identified on the basis of at least one peptide and a minimum of three quantitation events in at least one experimental group were considered . IBAQ protein intensity values were normalized against the median intensity of each sample ( using only peptides with recorded intensity values across all samples and biological replicas ) , log-transformed and missing values filled by imputation with random numbers drawn from a normal distribution calculated for each sample . Significant interactors were determined by multiple equal variance t-tests with permutation-based false discovery rate statistics . We performed 250 permutations and the FDR threshold was set at 0 . 001 . The parameter S0 was set at 2 to separate background from specifically enriched interactors . Results were plotted as Volcano plot and heat map using the Perseus software package [64] . Cells were lysed in IP lysis buffer , incubated on ice for 1 h and samples cleared by centrifugation at 21 , 000 x g for 1 h . Protein concentration was measured by Bradford assay , samples were denatured by boiling for 5 min in 98°C in SDS-PAGE buffer ( 120 mM Tris-HCl [pH 6 . 8] , 60 mM SDS , 100 mM DTT , 1 . 75% glycerol , 0 . 1% bromophenol blue ) and 20 to 30 μg of total protein was loaded onto each lane of a gel . Proteins were separated by SDS-PAGE and transferred onto a polyvinylidenfluorid membrane . After blocking of the membrane with 5% milk in PBS-T ( PBS with 0 . 5% Tween ) or 5% BSA in PBS-T , they were incubated with primary and secondary horse radish peroxidase-conjugated antibodies as specified in the antibody section . Signals were developed by using the Western Lightning Plus-ECL reagent ( Perkin Elmer , Waltham , MA , USA ) and visualized with a ChemoCam Imager 3 . 2 ( Intas Science Imaging Instruments GmbH , Göttingen , Germany ) . The LabImage 1D software ( Intas ) was used for quantification of protein-specific signals . Cells grown on glass coverslips were fixed with 2% glutaraldehyde in 50 mM cacodylate buffer [pH 7 , 2] containing 10 mM MgCl2 , 10 mM CaCl2 , 100 mM KCl and 2% sucrose . Cells were either stored at 4°C for up to several days or directly washed 5x with 50 mM cacodylate buffer , incubated with 2% osmium tetroxide for 40 min on ice and 0 . 5% uranyl acetate , dissolved in double distilled water , for 30 min at room temperature or 24 h at 4°C . Samples were washed with double distilled water for 30 min and dehydrated step-wise with 40% to 100% ethanol , embedded in an araldite-Epon mixture ( Araldite 502/Embed 812 kit; Electron Microscopy Sciences ) and left for one to three days at 60°C to allow complete polymerization . After removal of the coverslip , embedded cells were cut into 70-nm thick sections with a Leica Ultracut UCT microtome ( Leica , Wetzlar , Germany ) and a diamond knife and mounted onto a mesh grid . Retrieved sections were further incubated with 3% uranyl acetate in 70% methanol for 5 min , followed by 2% lead citrate in distilled water for 2 min . Samples were analyzed with an EM10 transmission electron microscope ( Carl Zeiss AG , Oberkochen , Germany ) or a Jeol JEM-1400 ( Jeol Ltd . , Tokyo , Japan ) . For quantification of VPs , cells from randomly selected areas of the grid were analyzed . Wildtype samples were always prepared in parallel and only experiments where at least 20% of the wildtype cells were positive for VPs were taken under consideration . For each experiment the number of counted cells is given in the figure legend . For immunofluorescence analysis cells grown and transfected as described above were fixed in 4% paraformaldehyde for 15 min and permeabilized by 15 min incubation with 0 . 2% Triton X-100 in PBS . Cells were stained with primary antibodies as specified in the figure legends , followed by staining with anti-mouse or anti-rabbit secondary antibodies , conjugated with Alexa Fluor 488 or 568 ( ThermoFisher Scientific ) . Coverslips were mounted on glass slides and analyzed with a Nikon Eclipse Ti microscope ( Nikon , Tokyo , Japan ) to assess transfection efficiency , or a Leica SP8 confocal microscope ( Leica ) to analyze subcellular localization of DENV proteins . Statistical analyses were performed using the GraphPad Prism 5 . 0 software package ( LaJolla , CA , USA ) . Two-tailed paired Student’s t-test with Bonferroni correction for multiple samples comparison was used to assess statistical significance . All molecular graphics were prepared with either MOE 2015 or UCSF Chimera software [66] .
|
Dengue virus ( DENV ) is one of the most prevalent mosquito-transmitted human pathogens . The only licensed vaccine has limited efficacy and an antiviral therapy is not available . The multifunctional non-structural protein 1 ( NS1 ) of DENV is secreted from infected cells , counteracts antiviral immune response and contributes to the pathogenesis of DENV infection . In addition , NS1 is essential for the viral replication cycle but the underlying mechanism is unknown . Here we determined the viral interactome of NS1 and identified a novel interaction of NS1 with the NS4A-2K-4B cleavage intermediate , but not with NS4A and NS4B . This interaction is required for RNA replication . Additionally , we identified a domain in NS1 important for efficient secretion of this protein . Finally , we demonstrate that NS1 is required for the biogenesis of the membranous DENV replication organelle . This function does not require RNA replication and is independent from NS1 interaction with NS4A-2K-4B . Our results provide new insights into the role of NS1 in DENV RNA replication and establish a genetic map of residues in NS1 required for the diverse functions of this protein . These results should aid in the design of antiviral strategies targeting NS1 , with the aim to suppress viral replication as well as severe disease manifestations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"transfection",
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"protein",
"interactions",
"pathology",
"and",
"laboratory",
"medicine",
"insertion",
"mutation",
"pathogens",
"microbiology",
"viruses",
"physiological",
"processes",
"mutation",
"rna",
"viruses",
"immunoprecipitation",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"viral",
"replication",
"molecular",
"biology",
"precipitation",
"techniques",
"biochemistry",
"point",
"mutation",
"flaviviruses",
"virology",
"physiology",
"viral",
"pathogens",
"genetics",
"secretion",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2019
|
A novel interaction between dengue virus nonstructural protein 1 and the NS4A-2K-4B precursor is required for viral RNA replication but not for formation of the membranous replication organelle
|
To be able to colonize its host , invading Salmonella enterica serovar Typhimurium must disrupt and severely affect host-microbiome homeostasis . Here we report that S . Typhimurium induces acute infectious colitis by inhibiting peroxisome proliferator-activated receptor gamma ( PPARγ ) expression in intestinal epithelial cells . Interestingly , this PPARγ down-regulation by S . Typhimurium is independent of TLR-4 signaling but triggers a marked elevation of host innate immune response genes , including that encoding the antimicrobial peptide lipocalin-2 ( Lcn2 ) . Accumulation of Lcn2 stabilizes the metalloproteinase MMP-9 via extracellular binding , which further aggravates the colitis . Remarkably , when exposed to S . Typhimurium , Lcn2-null mice exhibited a drastic reduction of the colitis and remained protected even at later stages of infection . Our data suggest a mechanism in which S . Typhimurium hijacks the control of host immune response genes such as those encoding PPARγ and Lcn2 to acquire residence in a host , which by evolution has established a symbiotic relation with its microbiome community to prevent pathogen invasion .
Salmonella enterica serovar Typhimurium is a Gram-negative , facultative intracellular pathogen that causes a wide array of disorders ranging from systemic disease to enterocolitis in multiple hosts [1] . In mice , S . Typhimurium normally causes a disease that resembles systemic typhoid fever . However , compromising the gut microbiome with antibiotics prior to S . Typhimurium infection in mice has been used to mimic salmonellosis in humans , which involves increased S . Typhimurium colonization of the intestine coupled with a marked host-induced inflammatory response leading to colitis [2] . Recent reports indicate that this massive inflammatory response elicited by S . Typhimurium is associated with increased secretion of the interleukins IL-17 and IL-22 [3] , [4] , which are critical components of mucosal immunity to bacterial pathogens in the gut . In particular , the IL-17/IL-22 axis mediates the recruitment of antimicrobial peptides from the intestinal epithelial compartment , including lipocalin-2 ( Lcn2 ) [3]–[7]; these peptides dramatically affect the gut microbiota . Lcn2 ( also known as SIP24 , 24p3 , NGAL , uterocalin , and siderocalin ) was first co-purified and found to be covalently associated with human neutrophil gelatinase ( matrix metalloproteinase ( MMP ) -9 ) [8] , [9] . This association between Lcn2 and MMP-9 has been shown to protect MMP-9 from degradation and to preserve its enzymatic activity [8] , [10] . In addition , Lcn2 functions in mammalian innate immunity by chelating bacterial siderophores , thereby sequestering iron from bacteria and inhibiting their growth [11] , [12] . Intriguingly , S . Typhimurium appears to be resistant to Lcn2 , since its population in the intestinal milieu expands dramatically during inflammation [4] , [13] , [14] . This unique strategy is achieved by genes such as those in the iroN iroBCDE gene cluster , which encodes salmochalin , a siderophore that does not bind Lcn2 [4] , [13] , [15] , [16] , thus conferring a competitive advantage to S . Typhimurium over other microbes during growth in the inflamed gut . Recently , peroxisome proliferator-activated receptor gamma ( PPARγ ) has been shown to be regulated by a number of bacterial pathogens including Helicobacter pylori and Mycobacterium tuberculosis [17]–[19] , greatly impacting disease severity . PPARγ is a member of the nuclear receptor superfamily of ligand-dependent transcription factors and is predominantly expressed in adipose tissue and colonic epithelium [20] , [21] . Expression has also been detected in colonic macrophages and in T and B cells of humans and rodents [20] , [22] . PPARγ has been proclaimed to be a master regulator of inflammation , a role that is achieved in part by antagonizing the activities of the transcription factors AP-1 , STAT , and NFκB [23] , [24] . In vivo studies have demonstrated that PPARγ ligands actively suppress the inflammatory response by attenuating the production of chemokines and cytokines secreted from epithelial cells , macrophages , and T and B lymphocytes [23]–[26] . The role of PPARγ in the etiology and treatment of colitis has been of great interest , because its ligands have long been used to treat type-2 diabetes and are known to decrease the severity of colitis induced in mouse models [27]–[33] . Moreover , PPARγ+/− heterozygous mice exhibit increased susceptibility to experimentally induced colitis , indicating PPARγ's involvement in maintaining gut homeostasis [28] , [33] . Furthermore , the observation that intestinal epithelium-specific ablation of PPARγ aggravates dextran sodium sulfate ( DSS ) -induced colitis demonstrates the strong influence of intestine-derived PPARγ on colitis severity [20] . Studies in human subjects have revealed that colonic epithelial cells from ulcerative colitis patients display drastically reduced expression of PPARγ , suggesting that its presence in gut epithelium may have a protective effect against colonic inflammation in humans [29] . Despite these observations , the role of PPARγ in S . Typhimurium-induced infectious colitis remains unknown . In this study , we explored whether S . Typhimurium regulates host PPARγ levels during infectious colitis and evaluated PPARγ's contributions to the etiology of the disease . Our data reveal that S . Typhimurium inhibits PPARγ expression in the intestinal epithelium , which triggers a massive innate immune response that includes expression of Lcn2 . Selective epithelial ablation of PPARγ dramatically increased Lcn2 expression and its secretion after S . Typhimurium challenge , confirming the importance of epithelium-derived PPARγ in colitis . Furthermore , increased secretion of Lcn2 stabilized MMP-9 via direct protein-protein interaction , which further aggravated the colitis . Finally , we demonstrate that Lcn2-null mice exposed to S . Typhimurium displayed significantly less-severe colitis .
Although PPARγ signaling controls various cellular processes during inflammation and pathogenesis , its regulation during S . Typhimurium-induced colitis remains unexplored . To gain insight into PPARγ's role in infectious colitis , streptomycin-pretreated C57BL/6 mice were infected with S . Typhimurium . An incubation period of 24 h was deliberately selected to evaluate the early phase of PPARγ response , which is crucial for further downstream effector regulation . Strikingly , S . Typhimurium infection resulted in ∼60% down-regulation of PPARγ gene expression in the colon , as detected by real-time PCR ( Fig . 1A ) . Immunoblotting revealed a similar down-regulation of PPARγ expression at the protein level ( Fig . 1B ) in the colon 24 h after infection . As expected , this down-regulation resulted in a marked reduction in the DNA-binding activity of PPARγ in colonic cells in vivo ( Fig . 1C ) , thus reducing PPARγ's tight control over its potential targets . A similar effect was detected in the cecum of infected mice ( Fig . S1A–C ) . Since a previous study reported Toll-like receptor ( TLR ) -4-dependent PPARγ regulation by microbial lipopolysaccharide in macrophages [34] , we asked whether this negative regulation of PPARγ by S . Typhimurium was TLR-dependent . A significant increase in the expression of gene encoding TLR-4 , but not TLR-2 and TLR-5 , was observed in the infected mice , suggesting that TLR4 may be involved in regulating PPARγ ( Fig . S2A–C ) . However , S . Typhimurium infection in TLR-4−/− mice resulted in a similar decline in PPARγ expression ( Fig . 1D and E ) , suggesting that this PPARγ regulation was independent of TLR-4 . However , the expression of TLR-2 and TLR-5 gene in the infected TLR-4−/− mice remained unchanged , negating the possibility of their increased activity in the absence of TLR-4 ( Fig . S2D and E ) . To confirm the specificity of PPARγ deregulation by S . Typhimurium , another potential gut pathogen , Citrobacter rodentium , was used to infect mice . C . rodentium was unable to alter host PPARγ levels , confirming that the process of PPARγ down-regulation by S . Typhimurium was not a general effect ( Fig . 1F ) . Next , we set out to identify the cell types that predominantly respond to S . Typhimurium . A sharp decrease in the gene expression of PPARγ occurred when cultures of colonic epithelial cells ( HT-29 cells ) were infected with S . Typhimurium , suggesting the importance of the intestinal epithelium in this process ( Fig . 1G ) . These observations not only indicate that S . Typhimurium infection directly impacts PPARγ levels in the colonic epithelium , but also suggest that PPARγ's pivotal role in homeostasis within the intestinal tract affects infectious colitis . To better understand the implications of this PPARγ regulation by S . Typhimurium , we bred mice harboring a floxed Pparγ ( PPARγfl/fl ) to mice expressing the Cre transgene under control of the promoter of the villin gene . These mice , in which cre recombinase mediated the targeted disruption of PPARγ in intestinal epithelial cells , were designated PPARγVillinCre+ mice and were used in parallel with littermate control PPARγVillinCre- or wild-type ( C57BL/6 ) mice . Interestingly , S . Typhimurium infection resulted in more severe colitis in the PPARγVillinCre+ mice compared to PPARγVillinCre- or wild-type mice at 24 h ( Fig . 1H and I ) . Shortening and thickening of the cecum and the colon , which are hallmarks of colitis , were much more pronounced in PPARγVillinCre+ mice than in wild-type or PPARγVillinCre- mice ( Fig . 1I ) . Of note , the S . Typhimurium-infected TLR-4−/− mice and the wild-type ( C57BL/6 ) mice infected with C . rodentium exhibited significantly shortened colons , indicating active colitis ( Fig . S2F and G ) . S . Typhimurium cells were recovered in similar numbers from cecum tissue and from the spleens of wild-type , PPARγVillinCre- , and PPARγVillinCre+ mice 24 h after infection ( Fig . 1J and K ) , indicating that infection in the cecum as well as systemic dissemination of S . Typhimurium were comparable between these groups . Histological analysis further revealed that apart from increased infiltration of inflammatory cells , tissue damage was more in the colons of PPARγVillinCre+ mice compared to PPARγVillinCre- mice and to wild-type mice after infection ( Fig . 2 ) . Notably , similar levels of the epithelial cell markers villin 1 , cytokeratin 8 , and cytokeratin 20 were detected in infected and mock-infected colonic samples , confirming that the ratios of epithelial cells in the colonic extracts from these groups were consistent and did not account for the reduced levels of PPARγ ( Fig . S3 ) . These results were consistent with previous reports indicating a protective role of PPARγ in the intestinal epithelium in experimental inflammatory bowel disease [20] . However , the molecular mechanisms underlying these observations remain elusive . We previously demonstrated the direct participation of PPARγ in host-microbe crosstalk and the consequent regulation of innate immune functions [35] , [36] . Moreover , PPARγ has been proposed to regulate inflammation by antagonizing the NFκB and AP-1 pathways [23] , [24] , which in turn may modulate immune responses . To test this hypothesis in infectious colitis , we analyzed the activities of NFκB and AP-1 via electromobility shift assay using nuclear extracts from colonic scrapings . As anticipated , depletion of epithelial PPARγ was coupled with a marked increase in the activities of NFκB and AP-1 in the colon after S . Typhimurium infection compared to littermate control ( PPARγVillinCre- ) mice ( Fig . 3A and B , Fig . S4 ) . A potential challenge was to obtain maximal numbers of epithelial cells in colonic scrapings; epithelial cells contribute the majority of colonic PPARγ , but macrophages and B and T lymphocytes also produce it [20] , [22] . To overcome this problem , we used mucosal scrapings from the colon and tested for the presence of PPARγ transcripts through real-time PCR . Minimal expression and activity of PPARγ was detected in mock-infected and infected PPARγVillinCre+ mice compared to littermate control PPARγVillinCre- and wild-type ( C57BL/6 ) mice ( Fig . S5A and B ) . This finding was further validated in cecum scrapings ( Fig . S1D ) . Thus , it was evident that epithelial cells were the major cell type in our samples; the presence of infiltrating macrophages and T and B cells in the colonic scrapings was possible , but minimally contributed to PPARγ production . Furthermore , no significant differences in the expression or activity of PPARγ were detected between PPARγVillinCre- and wild-type C57BL/6 mice ( Fig . S5 ) , justifying the use of PPARγVillinCre- mice as controls in subsequent experiments . We next investigated the effect of increased NFκB and AP-1 activities on key regulators of inflammation . After infection , the expression levels of TNF-α and IL-6 in the colon were two-fold higher in PPARγVillinCre+ mice compared to PPARγVillinCre- mice ( Fig . 3C and D ) . IL-6 is a key regulator of the innate T helper type 17 ( TH17 ) response , a critical component of mucosal immunity to intestinal pathogens [3] . Consistent with this role , we observed substantial increases in the expression of IL-17 and IL-22 , a typical innate TH17 response signature , in infected PPARγVillinCre+ mice compared to infected PPARγVillinCre- mice , as assessed via real-time PCR ( Fig . 3E and F ) . IL-17 and IL-22 have been linked with intestinal innate epithelial defense mechanisms through the production of antimicrobial peptides [3] , [5]–[7] . The expression of Lcn2 , a principal target of IL-17 and IL-22 [3] , [4] , increased by approximately 7-fold in the colons of S . Typhimurium-infected PPARγVillinCre- mice compared to mock-infected mice , an increase that rose to a striking ∼21-fold change in infected PPARγVillinCre+ mice ( Fig . 4A ) . This increase in Lcn2 expression in infected colons was also noticeable at the protein level ( Fig . 4B and C ) , further confirming a hyper-magnified Lcn2 response in infected PPARγVillinCre+ mice . Lcn2 was similarly elevated in the cecum of these mice ( Fig . S1E ) . Interestingly , TLR-4−/− mice infected with S . Typhimurium also showed a significant increase in Lcn2 expression in the colon compared to mock-infected mice ( Fig . S2H ) . The expression of regenerating islet-derived 3 gamma ( Reg3γ ) , another potent member of the host antimicrobial arsenal and a potential target of IL-17 and IL-22 [3] , [37] , was also tested . The expression of Reg3γ in colon followed a pattern similar to that of Lcn2 in PPARγVillinCre- and PPARγVillinCre+ mice ( Fig . 4D ) , thus establishing a heightened innate immune response to S . Typhimurium in the absence of epithelial PPARγ . Moreover , S . Typhimurium induced a significant increase in the expression and secretion of Lcn2 ( Fig . 4E and F ) in parallel with PPARγ down-regulation ( Fig . 1G ) in colonic epithelial cells , further confirming the importance of epithelial cells in this process . To test whether this Lcn2 up-regulation was a cell-autonomous effect of PPARγ in epithelial cells , we next applied PPARγ small interfering RNA ( siRNA ) to HT-29 cells . A ∼50% reduction in PPARγ levels ( Fig . 4G ) led to a significant increase in Lcn2 expression and secretion in the absence of S . Typhimurium infection ( Fig . 4H and I ) . Moreover , when these PPARγ-siRNA treated HT-29 cells were infected with S . Typhimurium , the expression profile of Lcn2 resembled to that observed in vivo ( Fig . S2I ) . These observations not only confirmed the existence of a direct link between PPARγ and Lcn2 , but also suggested that Lcn2 regulation by PPARγ may occur locally in epithelial cells without the intervention of other cell types . At this juncture , the observed dichotomy in PPARγ's role in infectious colitis was surprising . Although we initially observed that the absence of PPARγ led to increased colitis severity ( Fig . 1H and I; Fig . 2 ) , in sharp contrast we also detected a simultaneous elevated innate immune response that seemed to serve a protective role . Taken together , these results reflect PPARγ's tight control over intestinal homeostasis in the host during S . Typhimurium pathogenesis . We next questioned the rationale behind this heightened Lcn2 expression , given that S . Typhimurium is typically resistant to Lcn2's antimicrobial activity [4] , [13] , [15] , [16] , and speculated that Lcn2 may have a more diverse role in the disease process . Having established that Lcn2 expression was markedly increased in PPARγVillinCre+ mice ( Fig . 4A–C ) , we next sought to dissect its possible role in the increased colonic damage observed in PPARγVillinCre+ mice in the absence of PPARγ ( Fig . 2 ) , suggesting ongoing exaggerated protease action . Interestingly , Lcn2 has been shown to increase the stability of MMP-9 by protecting it from degradation , resulting in an increase in its enzymatic activity independent of transcriptional regulation [8] , [10] , [38] , [39] . To test this hypothesis , we analyzed the activity of gelatin agarose-purified secreted gelatinases by zymography using phosphate-buffered saline ( PBS ) extracts of colonic scrapings of mock- or S . Typhimurium-infected mice . Interestingly , we detected a ∼140 KDa band representative of Lcn2-bound proMMP-9 ( proMMP-9/Lcn2 ) in S . Typhimurium-infected PPARγVillinCre- mice that peaked to an ∼6-fold increase in infected PPARγVillinCre+ mice ( Fig . 5A and B ) . Surprisingly , this increase in the proMMP-9/Lcn2 band in PPARγVillinCre+ mice was associated with a massive increase in the activity of proMMP-9 ( detected by zymography Fig . 5A ) , indicating an increased stability of proMMP-9 . The activity of proMMP-9 increased by ∼6 fold in PPARγVillinCre- mice and ∼15 fold in PPARγVillinCre+ mice after infection , while MMP-9 activity rose from ∼4 fold in PPARγVillinCre- mice to ∼7 fold in PPARγVillinCre+ mice after infection ( Fig . 5A and B ) . In contrast , MMP-2 activity remained almost unchanged between groups . Importantly , no significant difference in MMP-9 gene expression occurred between the PPARγVillinCre- and PPARγVillinCre+ infected groups , as detected by real-time PCR ( Fig . 5C ) , which confirmed that the differences in MMP-9 activity were independent of transcriptional regulation . Furthermore , MMP-2 expression was similar between the infected and mock-infected groups ( Fig . 5D ) . No noticeable difference in the expression of TIMP-1 , the endogenous inhibitor of MMP-9 , was observed between the infected mice groups ( Fig . 5E ) , eliminating the possibility of its involvement in the observed deregulation of MMP-9 activity . To further confirm this phenomenon , we next analyzed the protein levels of MMP-9 and Lcn2 by immunoblotting gelatin-agarose-purified PBS extracts of colonic scrapings under non-reducing conditions . As anticipated , the ∼140 KDa proMMP-9/Lcn2 band , the level of which increased by ∼5 fold in infected PPARγVillinCre- mice , peaked at ∼15-fold in PPARγVillinCre+ infected mice ( Fig . 5F and G ) . This increase led to an ∼15-fold increase in proMMP-9 protein levels in PPARγVillinCre- infected mice and reached a striking increase of ∼40 fold in PPARγVillinCre+ infected mice , while MMP-9 climbed ∼5 fold in PPARγVillinCre+ infected mice ( Fig . 5F and G ) . When the gels were re-probed for Lcn2 , bands at exactly the same location as the ∼140 KDa band , characteristic of proMMP-9 bound to Lcn2 , were detected ( Fig . 5F and G ) . These results indicate that Lcn2 is secreted from the intestinal epithelium , in accordance with previous reports [3] , and more importantly confirm the binding of Lcn2 to secreted proMMP-9 in the intestinal milieu , which is pivotal for subsequent pathological processes . However , since the increase in the levels of the Lcn2/MMP-9 complex in the intestinal milieu correlates with the increased infiltration of neutrophils it is possible that these cells may contribute to the observed elevation of Lcn2/MMP-9 . It is noteworthy that the ∼6-fold increase in Lcn2 secretion in S . Typhimurium-infected PPARγVillinCre+ mice ( Fig . 5F and G ) did not precisely correspond to its ∼15-fold increase at the expression level ( Fig . 4A–C ) , because it represented only the Lcn2 fraction bound to proMMP-9 during gelatin-agarose purification of secreted gelatinases . Reasonably , no low molecular-weight band for Lcn2 alone was detected ( Fig . 5F ) . It may also be noted that the proportion of bound or free Lcn2 or proMMP-9 at any given point would largely depend on the availability , stability , and importantly the stoichiometry of binding between these molecules . Taken together , these results conclusively indicate that lack of epithelial PPARγ substantially elevates Lcn2 expression and its secretion in the intestinal milieu during S . Typhimurium infection , resulting in increased MMP-9 stabilization and activity . To validate the observed involvement of Lcn2 in S . Typhimurium-induced colitis , we next checked for colitis induction in streptomycin-pretreated Lcn2−/− mice mock- or S . Typhimurium-infected for 24 h . As expected from the above observations , marked reductions in the extent and severity of S . Typhimurium-induced colitis were observed in mice devoid of Lcn2 ( Fig . 6A and B ) . Shortening and thickening of the cecum and colon were considerably restricted in the Lcn2−/− mice 24 h after S . Typhimurium challenge ( Fig . 6A and B ) . These results were confirmed by observations of reduced thickening of the mucosa and sub-mucosa , with negligible tissue damage ( Fig . 6C–H ) . However , a modest increase in infiltrating cells and consequent myeloperoxidase ( MPO ) activity was noted ( Fig . 6G and H ) . Moreover , the number of S . Typhimurium in the cecum and spleen of these Lcn2−/− mice was comparable to that of wild-type mice after 24 h ( Fig . 6I and J ) . As expected , Lcn2 expression was not detected in these mice by real-time PCR ( Fig . S6A ) or by immunoblotting ( Fig . S6B ) of the colons of Lcn2−/− mice . Thus , these observations highlight Lcn2's unique role in the induction and severity of infectious colitis . To more precisely track the molecular mechanisms active in Lcn2−/− mice after S . Typhimurium infection , we assessed the secretion of gelatinases in the colon . We observed an ∼3 . 5-fold increase in proMMP-9 activity in Lcn2−/− mice after infection versus mock infection ( Fig . 7A and B ) , compared to a ∼6-fold increase in PPARγVillinCre- infected mice ( Fig . 5A and B ) . No notable differences in the activities of MMP-9 and MMP-2 were detected between groups . The basal secretion of gelatinases was similar in PPARγVillinCre- and Lcn2−/− mock-infected mice ( Fig . S7A ) , justifying the comparison between these groups in this case . We next examined the levels of secreted MMP-9 using gelatin-purified colonic extracts of mock- or S . Typhimurium-infected Lcn2−/− mice . Compared to the ∼15-fold increase in secreted proMMP-9 that was observed in PPARγVillinCre- mice after infection ( Fig . 5F and G ) , a mere ∼3-fold increase in the levels of secreted proMMP-9 was detected in Lcn2−/− mice after infection , suggesting reduced extracellular stability and possible degradation of MMP-9 protein in the absence of Lcn2 ( Fig . 7C and D ) . No significant difference in the colonic expression of MMP-9 was detected between Lcn2−/− and Lcn2+/+ mice after infection ( Fig . 7E ) , confirming that the differences in the protein levels of MMP-9 were due to impaired stability . Moreover , colonic expression of MMP-2 and TIMP-1 was similar in Lcn2+/+ and Lcn2−/− mice after infection ( Fig . S7B and C ) , excluding their involvement in this process . To investigate any possible mechanistic differences between the Lcn2−/− and Lcn2+/+ mice during S . Typhimurium infection , we validated several key regulators involved in the pathophysiology of infectious colitis . Real-time PCR revealed no appreciable differences in the down-regulation of PPARγ by S . Typhimurium between Lcn2−/− and Lcn2+/+ mice ( Fig . S8A ) , confirming that PPARγ regulation by S . Typhimurium was essentially the same in Lcn2−/− mice . Consequently , the expression levels of TNF-α and IL-6 were similar between the infected groups ( Fig . S8B and C ) . Colonic expression of IL-17 and IL-22 also exhibited negligible differences between Lcn2−/− and Lcn2+/+ infected mice ( Fig . S8D and E ) , as did Reg3γ expression ( Fig . S8F ) . Together , these observations suggest that the mechanistic chain of events during S . Typhimurium infection in wild-type and Lcn2−/− mice was fundamentally similar; the absence of Lcn2 exclusively conferred protection to these mice against S . Typhimurium-induced colitis . To determine whether the regulation of PPARγby S . Typhimurium is a transient effect or whether the Lcn2−/− mice are protected at later stages of infection , we infected PPARγVillinCre- , PPARγVillinCre+ , Lcn2−/− , and Lcn2+/+ mice with S . Typhimurium and sacrificed them after 72 h . Long-term S . Typhimurium infection resulted in more severe colitis in PPARγVillinCre+ mice compared to PPARγVillinCre- mice , while Lcn2−/− mice were significantly protected against colitis ( Fig . S9A ) ; colonic shortening displayed a similar profile ( Fig . S9B ) . There was no significant difference in the number of S . Typhimurium recovered from the cecum or spleen of these mice 72 h after infection ( Fig . S9C and D ) . Histological analysis revealed more severe colitis in PPARγVillinCre+ mice than in PPARγVillinCre- mice , with significantly increased neutrophil infiltration and MPO activity ( Fig . S9E-L and S9U-W ) . Although the Lcn2−/− mice exhibited moderate colitis , they were associated with marked reductions in neutrophil infiltration , edema , and MPO activity compared to the rest of the groups 72 h after infection ( Fig . S9M–W ) . Interestingly , PPARγ expression was still significantly reduced after 72 h in the colons of PPARγVillinCre- mice , confirming that PPARγ regulation by S . Typhimurium was not a transient event ( Fig . S9X and Y ) .
Intestinal pathogens employ diverse strategies to modulate the host environment in order to survive in this competitive niche . The approaches that individual pathogens adopt depend largely on the tenure of their residence in the host . For instance , H . pylori up-regulates host PPARγ as part of a feedback mechanism to suppress exaggerated inflammation , ensuring its unperturbed long-term survival in the host [17] , [19] . Similarly , M . tuberculosis induces PPARγ expression in infected individuals and subsequently interacts with host PPARγ by modulating macrophage function for its survival [18] . Here , we unravel a novel mechanism used by S . Typhimurium to down-regulate PPARγ in the intestinal epithelium , initiating acute inflammation via the host immune and protease machinery , thereby transforming the intestine into a more hostile niche where it is best adapted to survive and outgrow its competitors . This TLR-4-independent regulation of PPARγ by S . Typhimurium seems to be characteristic of this pathogen , since C . rodentium infection , did not alter PPARγ levels . PPARγ activation has been shown to ameliorate the severity of inflammatory bowel disease in rodent DSS , trinitrobenzene sulphonic acid , and ischemic colitis models [20] , [27] , [30] , [32] , [33] . Since PPARγ is expressed in epithelial cells as well as in immune cells infiltrating colonic tissue during inflammation , the cell type that mainly contributes to PPARγ production during colitis remains a point of contention . Interestingly , clinical reports indicate that PPARγ expression in colonic epithelium is impaired in ulcerative colitis patients , while its expression in inflammatory cells remains normal [40] . This observation was corroborated by Adachi et al . , who reported that PPARγ expressed in the colonic epithelium has an endogenous role in protection against DSS-induced colitis [20] . These findings , together with our data in Fig . 1G , demonstrating PPARγ regulation by S . Typhimurium in human colonic epithelial cells , prompted us to use epithelial-specific PPARγ-null mice to unravel the chain of events that occur during S . Typhimurium-induced colitis . This strategy allowed us to magnify the subtle molecular changes induced by S . Typhimurium in the host via down-regulation of epithelial PPARγ . S . Typhimurium induced much more severe colitis in these mice ( Fig . 1H and I; Fig . 2 ) , highlighting , for the first time , the importance of intestinal epithelium-derived PPARγ in protection against bacterial pathogenesis . Our data demonstrate that S . Typhimurium-induced depletion of epithelial PPARγ uncouples PPARγ's tight control over the inflammatory transcription factors NFκB and AP-1 , resulting in the release of the pro-inflammatory cytokines TNF-α and IL-6 . This influx of pro-inflammatory cytokines from the intestinal epithelium initiates an acute-phase immune response characterized by elevated expression of IL-17 and IL-22 . These results are consistent with a recent report by Geddes et al . of the induction of the innate TH17 response by S . Typhimurium during the early phases of infection [3] . Collectively , our results imply that these inflammatory signaling circuits are orchestrated by S . Typhimurium during the early phases of infection via the regulation of epithelial PPARγ , which is pivotal for the entire process . The secretion of IL-17 and IL-22 in the inflamed colon , which is initiated by S . Typhimurium , has been shown to facilitate the production of antimicrobials , including Lcn2 , from the intestinal epithelium [3]–[7] . Our observations confirm the contribution of IL-17 and IL-22 to Lcn2 production and secretion , but also suggest that S . Typhimurium-induced increases in NFκB and AP-1 activity in epithelial cells via PPARγ down-regulation may directly influence Lcn2 expression in the same cells independent of IL-17 and IL-22 ( Fig . 7F ) . This hypothesis was further corroborated by the induction of Lcn2 in non-infected human colonic epithelial cells treated with siRNA directed against PPARγ . It seems reasonable to assume that S . Typhimurium may utilize this more direct pathway of Lcn2 regulation , since Lcn2 expression is known to be regulated by NFκB or even AP-1 [41] , [42]; importantly , this entire sequence of events may occur in affected epithelial cells . Understanding the rationale behind the substantial increase in Lcn2 secretion from the intestinal epithelium during S . Typhimurium infection , which typically utilizes salmochalin , a siderophore resistant to Lcn2 action [4] , [13] , [15] , [16] , for iron uptake , was perhaps the biggest challenge in this study . Here , we documented that S . Typhimurium-induced elevated influx of secreted Lcn2 in the intestinal milieu leads to stabilization and a significant increase in MMP-9 activity , through direct extracellular protein-protein binding ( Fig . 7F ) . MMP-9 , a member of a family of zinc-dependent endopeptidases that have broad substrate specificity , has been shown to play a pivotal role in the degradation and remodeling of the extracellular matrix during bacterial pathogenesis [19] , [43] . Moreover , MMP-9-null mice exposed to DSS or to S . Typhimurium were previously significantly protected from colitis [44] , [45] , confirming MMP-9's importance in the etiology of the disease . Although MMPs are secreted by a variety of cell types , such as fibroblasts , epithelial cells , endothelial cells , neutrophils , macrophages , and lymphocytes , MMP-9 is predominantly expressed in epithelial cells and in inflammatory cells during colitis [44] , [46] . Our data suggest that this crosstalk between Lcn2 and MMP-9 in the inflamed gut , which results in increased MMP-9 activity , was crucial for the deleterious impact on the intestinal mucosa observed during infectious colitis . This novel mechanism by which S . Typhimurium exploits the host Lcn2 and MMP-9 synergy to aggravate inflammation and colitis severity is critical , as it calls for the reinterpretation of studies on microbial pathogenesis; other potential pathogens may also employ similar mechanisms . Interestingly , mice lacking Lcn2 were considerably protected against S . Typhimurium-induced colitis even at the later stages of infection , confirming the key role of this secreted protein in S . Typhimurium pathogenesis . We detected no significant differences in the overall mechanism acting in Lcn2-null and wild-type infected mice , with the exception of a decrease in MMP-9 stability and activity in the colon . Our observations of Lcn2-null mice exposed to S . Typhimurium , which indicated increased expression of IL-17 , IL22 , and TNF-α , are consistent with the findings of Raffatellu et al . [4] . We also noted moderate inflammation in the colon during the later stages of infection . However , disease severity may be impacted by differences in experimental setup , including S . Typhimurium strains , bacterial load and phase of growth in the inoculum , and importantly , differences in intestinal microflora between mice . In conclusion , our investigation unveiled a novel pathogenic mechanism utilized by S . Typhimurium to thrive and to induce colitis in its host . This study motivates the development of therapeutic interventions directed against this Lcn2-dependent , MMP-9-driven tissue degradation pathway to combat salmonellosis . However , research aimed toward a better understanding of the pathogenic mechanisms of S . Typhimurium or other pathogens in the gut remains an exciting area for future studies .
All protocols involving animals were approved by the Regional Animal Research Ethical Board , Stockholm , Sweden , following proceedings described in European Union legislation ( Council Directive 86/609/EEC ) . Animal husbandry was in accordance with institutional guidelines at Karolinska Institutet and was approved by the above-mentioned ethical board ( Stockholms norra djurförsöksetiska nämnd , Ref: N 100/10 ) . The naturally streptomycin-resistant wild type strain S . enterica serovar Typhimurium SL1344 [47] , a generous gift from Prof . Mikael Rhen was used in this study . Naturally occurring naldixic acid-resistant Citrobacter rodentium , DBS100 ( ATCC 51459 ) was also used for mouse infection . Prior to inoculation into host mice , strain SL1344 and C . rodentium were grown overnight at 37°C in Luria-Bertani ( LB ) broth , diluted 1∶20 in fresh medium , and sub-cultured for 3–4 h under mild aeration until an optical density of 0 . 4 to 0 . 6 at 600 nm was reached . Bacteria were washed twice in cold PBS and then suspended in cold PBS for mouse inoculation . The cecum and spleen from post-sacrifice mice were collected in 1 mL of sterile PBS . Samples were kept on ice , minced , and homogenized . Serial dilutions of the homogenates were plated on LB agar plates supplemented with 100 µg/mL streptomycin to enumerate S . Typhimurium . Plates were incubated overnight at 37°C , and colonies were counted thereafter . The human epithelial cell line HT-29 ( ATCC-HTB-38 ) was obtained from the American Type Culture Collection . Cells were grown in RPMI 1640 ( Invitrogen ) medium supplemented with 10% heat-inactivated fetal calf serum ( Invitrogen ) . Cells were maintained in a 37°C humidified atmosphere with 5% CO2 . Epithelial morphogenesis was monitored via microscopy; cell densities for each experiment did not exceed 80% to prevent contact inhibition . For co-culture experiments , cells were treated with S . Typhimurium ( 0 . 25×107 cells/well; 10∶1 bacterial cells:eukaryotic cells ) for 6 h . Controls were treated with culture medium only . After 6 h medium was removed , cells were washed , fresh medium with 1% penicillin/streptomycin was added , and the cells were incubated for 18 h , after which the cells were collected and lysed . For siRNA experiments HT-29 cells were plated at a density of 0 . 0625×105 cells/cm2 . Down-regulation of PPARγ transcripts was achieved with SMART Pool siRNA directed against PPARγ ( Thermo Scientific ) . Controls were transfected with non-targeting siRNA ( Thermo Scientific ) at a final siRNA concentration of 40 nM . Transfection was carried out according to the manufacturer's protocol using DharmaFECT 4 ( Thermo Scientific ) reagent at a final concentration of 0 . 3% . Specific pathogen-free C57BL/6 wild-type mice carrying a targeted disruption of the gene encoding PPARγ in intestinal epithelial cells were generated by breeding animals harboring a floxed Pparγ ( PPARγfl/fl ) [48] to mice expressing the Cre transgene under control of the villin promoter; these mice were designated as PPARγVillinCre+ , and their littermate control mice were designated as PPARγVillinCre− . Lcn2-deficient ( Lcn2−/− ) mice [11] , [38] and TLR4−/− mice ( Jackson Laboratory Stock No: 007227 ) aged 8–10 weeks were also used in this study . The Lcn2−/− mice were generated previously [11] and were backcrossed into the C57BL/6 background for at least 10 generations [38] . All experiments were performed under standard controlled conditions and all efforts were made to minimize animal suffering . Groups of mice were pretreated with streptomycin ( 0 . 1 mL of a 200 mg/mL solution in sterile water ) orally 24 h prior to either mock ( PBS ) or S . Typhimurium ( 1×108 colony-forming units/mouse ) inoculation via gavage . At 24 h or 72 h after infection , mice were euthanized and the cecum , spleen , and colon were collected for analysis . For C . rodentium infection , mice were given metronidazole at 750 mg/L for 4 days , which was withdrawn prior to inoculation with 1×109 colony-forming units/mouse via gavage . C . rodentium-infected mice were sacrificed 6 days after infection [49] . Colon lengths were measured using a centimeter scale . The mucosal layer of the colon was carefully scraped and suspended in PBS containing protease inhibitors ( Roche ) , minced , and centrifuged at 6000 g for 15 min . The supernatant was collected for use as PBS extracts for the analysis of secreted proteins , while the pellet was re-extracted in lysis buffer ( 10 mM Tris-HCl [pH 8] , 150 mM NaCl , 1% Triton X-100 , and protease inhibitors ) to obtain Triton X-100 extracts [41] . For partial purification of MMP-9 and MMP-2 , PBS extracts of the respective samples were incubated with gelatin-agarose beads ( Sigma ) at 4°C for 1 h followed by centrifugation at 1500 rpm . The supernatant was collected as flow-through and used as a loading control . The pellet was washed twice with PBS through centrifugation at 1500 rpm and the gelatinases were eluted in Lammeli sample loading buffer . For assays of MMP-9 and MMP-2 activity , gelatin zymography was performed as described previously [43] . Zymographic bands were quantified using LabImage software ( KAPELAN ) . Triton X-100 extracts ( 50 µg/lane ) from colon samples were immunoblotted . Gelatin-agarose-purified PBS extracts were immunoblotted under non-reducing conditions . Cell-culture supernatants were concentrated using a vacuum centrifuge and volumetrically analyzed by immunoblotting . Immunoblots were probed with anti-PPARγ ( Cell Signaling ) , anti-MMP-9 ( Abcam ) , anti-Lcn2 ( Abcam ) , and anti-β-actin ( Santa Cruz Biotechnology ) antibodies . Immunodetection with an appropriate secondary peroxidase-conjugated antibody ( DAKO ) was followed by electrochemiluminescence ( Santa Cruz Biotechnology ) . Quantification of protein bands was performed with the LabImage software . Fold changes were calculated using densitometry values for bands representing proteins of interest , normalized to densitometry values for β-actin bands of respective samples . Representative blots from at least two independent experiments are shown . Total RNA was extracted with an RNeasy Mini Kit ( Qiagen ) , and cDNA was synthesized with SuperScript II ( Invitrogen ) , both procedures according to the manufacturers' protocols . We measured gene expressions with SYBR Green ( Applied Biosystems ) -based quantitative reverse transcription PCR . Primers were designed and tested according to Applied Biosystems recommendations ( Table S1 ) . Sample setups always included at least five biological replicates and experimental triplicates . The changes in mRNA expression of respective samples compared to control were expressed as ΔΔCt = ΔCtcontrol−ΔCtrespective samples ( ΔCt = Ct value for the gene of interest - Ct value for β-actin of the respective sample ) . Relative expressions in the genes in respective samples were calculated as 2ΔΔCt [19] . Nuclear extracts from colonic scrapings were prepared according the nuclear extraction protocol of Schreiber [50] . DNA binding was assayed with 10 µg of nuclear extract in binding buffer ( 25 mM HEPES [pH 7 . 9] , 70 mM KCl , 10% glycerol , 5 mM dithiothreitol , and 1 µg polydIdC ( Amersham ) ) in the presence of 50 , 000 cpm of a radiolabeled oligonucleotide probe . The probe for PPARγ ( sequences 5′-TCTCTCTGGGTGAAATGTGC-3′ and 5′-AGAGGCACATTTCACCCAGAGAGA-3′ ) has high PPARγ-specificity and moderate affinity to ensure weak binding to other PPARs [35] . Probes for NFκB ( sequences 5′-GATCCAGAGGGGACTTTCCGAG-3′ and 5′- TCGACTCGGAAAGTCCCCTCTG-3′ ) and for AP-1 ( sequences 5′-CTGATGACTCAGAG-3′ and 5′-CTCTGAGTCATCAG-3′ ) were used . polydIdC and probe were added to extracts and incubated for 30 min before gel electrophoresis . Bands were quantified with LabImage . The colons of mock- or S . Typhimurium-infected mice from the respective groups were sectioned for histological studies . Distal part of the colon was used for histological analysis in all cases . Tissue samples were fixed in 10% formalin and embedded in paraffin . Sections ( 5 µm ) were cut with a microtome , stained with hematoxylin and eosin [43] , and observed under a Zeiss microscope . Images were captured using Axiovision , LE 64 software ( Carl Zeiss Microscopy ) at original magnification 5×10 and 20×10 and processed in Adobe Photoshop CS6 ( Adobe Systems incorporated ) . Pathology scoring for neutrophil infiltration and edema was rated from 0 to 5 according to severity , under blinded conditions , by an experienced pathologist ( RMB ) . Whole-cell extracts from colonic scrapings were assayed for MPO activity using the Myeloperoxidase Activity Assay Kit ( Invitrogen ) following the manufacturer's protocol . Densitometry data were fitted using SigmaPlot 2001 ( SPSS ) or GraphPad Prism 5 ( GraphPad Software ) . Data are presented as the mean ± standard error of the mean . Between-group comparisons were carried out using either Student's t-test or Student-Newman-Keuls test ( ANOVA ) .
|
Enteric pathogens like S . Typhimurium convert the host intestine into an inflamed environment in which they are well adapted to thrive . However , the precise strategy that this pathogen employs to achieve such favorable conditions for its survival remains unclear . Here , we uncovered a novel mechanism whereby S . Typhimurium inhibits the expression of the transcription factor PPARγ in the host intestine , surprisingly without TLR-4 involvement; this inhibition worsened the severity of the host's colitis . Subsequent detailed analysis revealed that colitis severity was coupled with elevated levels of antimicrobials like Lcn2 , which stabilized the pro-inflammatory endopeptidase MMP-9 in the intestinal milieu . Combination of this escalated antimicrobial action together with enhanced protease activity disrupted the intestinal homeostasis , promoting an inflamed environment suitable for S . Typhimurium . Interestingly , using Lcn2 mutant mice we show that lack of Lcn2 effectively reduced tissue damage and the degree of inflammation , thus supporting a pivotal role of Lcn2 and MMP-9 in infectious colitis . Our data suggests a model whereby the pathogenesis of S . Typhimurium involves manipulation of the host innate immune and protease system , here illustrated by PPARγ , Lcn2 and MMP-9 , to establish colonization and infection within the host .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"gastroenterology",
"and",
"hepatology",
"medicine",
"biology",
"microbiology"
] |
2014
|
Absence of Intestinal PPARγ Aggravates Acute Infectious Colitis in Mice through a Lipocalin-2–Dependent Pathway
|
The chromosomes of multicellular animals are organized into a series of topologically independent looped domains . This domain organization is critical for the proper utilization and propagation of the genetic information encoded by the chromosome . A special set of architectural elements , called boundaries or insulators , are responsible both for subdividing the chromatin into discrete domains and for determining the topological organization of these domains . Central to the architectural functions of insulators are homologous and heterologous insulator:insulator pairing interactions . The former ( pairing between copies of the same insulator ) dictates the process of homolog alignment and pairing in trans , while the latter ( pairing between different insulators ) defines the topology of looped domains in cis . To elucidate the principles governing these architectural functions , we use two insulators , Homie and Nhomie , that flank the Drosophila even skipped locus . We show that homologous insulator interactions in trans , between Homie on one homolog and Homie on the other , or between Nhomie on one homolog and Nhomie on the other , mediate transvection . Critically , these homologous insulator:insulator interactions are orientation-dependent . Consistent with a role in the alignment and pairing of homologs , self-pairing in trans is head-to-head . Head-to-head self-interactions in cis have been reported for other fly insulators , suggesting that this is a general principle of self-pairing . Homie and Nhomie not only pair with themselves , but with each other . Heterologous Homie-Nhomie interactions occur in cis , and we show that they serve to delimit a looped chromosomal domain that contains the even skipped transcription unit and its associated enhancers . The topology of this loop is defined by the heterologous pairing properties of Homie and Nhomie . Instead of being head-to-head , which would generate a circular loop , Homie-Nhomie pairing is head-to-tail . Head-to-tail pairing in cis generates a stem-loop , a configuration much like that observed in classical lampbrush chromosomes . These pairing principles provide a mechanistic underpinning for the observed topologies within and between chromosomes .
The highly regular and reproducible physical organization of chromosomes in multicellular eukaryotes was recognized a century ago in cytological studies on the lampbrush chromosomes that are found in oocytes arrested at the diplotene phase of meiosis I [1–3] . At this stage , homologous chromosomes are paired . The two homologs display a similar and reproducible architecture . It consists of a series of loops emanating from the main axis , that are arranged in pairs , one from each homolog . In between the loops are regions of more compacted chromatin [2] . A similar physical organization is evident in insect polytene chromosomes [4] . As with lampbrush chromosomes , the paired homologs are aligned in precise register . However , instead of one copy of each homolog , there are hundreds . While loops are not readily visible , each polytene segment has a unique pattern of bands and interbands that depends upon the underlying DNA sequence and chromosome structure . Subsequent studies have shown that the key features of chromosome architecture evident in lampbrush and polytene chromosomes are also found in diploid somatic cells [5–13] . One of these is the subdivision of the chromatin fiber into a series of loop domains . There are now many lines of evidence indicating that looping is a characteristic architectural feature . Biochemical evidence comes from chromosome conformation capture ( 3C ) experiments , which show that distant sites come into contact with each other in a consistent pattern of topologically associating domains ( TADs ) . While the first studies in mammals suggested that TADs have an average length of 1 Mb [14–16] , subsequent experiments showed that the average is only about 180 kb [17] . In flies , TADs are smaller , between 10–100 kb [18 , 19] . Neighboring TADs are separated from each other by boundaries that constrain both physical and regulatory interactions . In mammals and also in flies , these boundaries typically correspond to sequences bound by insulator proteins like CTCF [17] . That TAD boundaries correspond to insulators is consistent with the known properties of these elements . Insulators subdivide the chromosome into functionally autonomous regulatory domains . When interposed between enhancers or silencers and target promoters , insulators block regulatory interactions . They also have an architectural function in that they can bring distant chromosomal sequences together , and in the proper configuration can promote rather than restrict regulatory interactions [20 , 21] . Moreover , insulators are known to mediate contacts between distant sequences ( loop formation ) , and these physical contacts depend upon specific interactions between proteins bound to the insulators [22 , 23] . The notion that insulators are responsible for subdividing eukaryotic chromosomes into a series of looped domains raises questions about the rules governing loop formation in cis . One of these is the basis for partner choice . Is choice based simply on proximity , or is there an intrinsic partner preference ? A second concerns the topology of the loop formed by interacting partners in cis . Do the partners interact to form a stem-loop-like structure , or does the interaction generate a circular loop ( “circle-loop” ) ? The answer to this question will depend upon whether there is an orientation dependence to the interactions between two heterologous insulators . In flies , homologs are typically paired in somatic cells , not just in cells that are polyploid [24] . This means that the loop domains in each homolog must be aligned in precise register along their entire length . A plausible hypothesis is that both alignment and homolog pairing are mediated by insulator interactions in trans . If this is case , there are similar questions about the rules that govern trans interactions . Is there a partner preference in the interactions that mediate homolog pairing ? Is there an orientation dependence , and if so , what is the topological outcome of the looped domains generated by insulator interactions in paired chromosomes in cis and in trans ? In the studies reported here , we have used insulators from the even skipped ( eve ) locus to address the questions posed above about the architecture of eukaryotic chromosomes . The eve domain spans 16 kb and is bordered upstream by the Nhomie ( Neighbor of Homie , this study ) insulator and downstream by Homie ( Homing insulator at eve ) [25 , 26] . eve encodes a homeodomain transcription factor that is required initially for segmentation , and subsequently in the development of the CNS , muscles , and anal plate [27 , 28] . It has a complex set of enhancers that activate expression at different stages and tissues [25 , 29–31] , and a Polycomb response element ( PRE ) that silences the gene in cells where it isn’t needed [32] . In early embryos , the stripe enhancers upstream ( 3+7 , 2 , late stripes ) and downstream ( 4+6 , 1 , and 5 ) of the eve gene activate transcription in a pair-rule pattern . Later in development , around the time that germband retraction commences , mesodermal ( Me ) and neuronal ( CNS ) enhancers turn on eve expression in a subset of cells in each of these tissues . These late enhancers continue to function once germband retraction is complete , while another enhancer ( APR ) induces transcription in the presumptive anal plate . Located just upstream of eve is CG12134 , while the TER94 gene is downstream . Unlike eve , both of these genes are ubiquitously expressed throughout much of embryogenesis .
The Homie insulator has two striking properties [26] . First it directs homing of otherwise randomly inserting transgenes to a ~5 Mb region centered on the eve locus . Second , when the homed transgene carries a reporter , it is expressed in an eve-like pattern , the completeness of which diminishes with distance . Early stripe and later CNS expression are limited to 200 kb from eve , mesodermal expression has an intermediate distance dependence , while anal plate ring ( APR ) expression is seen at distances of several Mb . We showed previously that reporter expression at a site within the hebe gene 142 kb upstream of eve requires Homie [26] . Since other fly insulators mediate long-distance regulatory interactions by direct physical contact [22 , 33] , we used high-resolution chromosome conformation capture ( H3C ) [34] to map contacts between transgenes at -142 kb and eve ( see below ) . The transgenes have an eve-promoter-lacZ ( lacZ ) reporter and Homie . One of them is inserted into the chromosome so that Homie is oriented in the same direction ( →; Fig 1A , transgene #1 ) as the endogenous Homie in the eve locus , while the other transgene is inserted in the opposite orientation ( ←; Fig 1A , transgene #2 ) . In the control transgene , Homie was replaced by DNA ( Fig 1A , transgene #3 ) . Fig 1A shows that the reporters in both Homie transgenes are regulated by the eve enhancers in a pattern which recapitulates that of endogenous eve . Thus , the orientation of the entire Homie:lacZ transgene in the chromosome doesn’t affect long-distance regulation . On the other hand , because of a hebe CNS enhancer located upstream of -142 kb , the expression pattern is not identical . In the transgene that is oriented so that Homie is closer to the eve locus than the reporter ( Fig 1A , 2nd column: transgene #1 ) , lacZ is regulated by both the hebe CNS enhancer ( arrow in bottom panel ) and the eve enhancers ( all 4 panels ) . When the transgene is inserted in the opposite orientation so that the lacZ reporter is closer to the eve locus ( Fig 1A , 3rd column: transgene #2 ) , Homie blocks the hebe enhancer , and only the eve pattern is seen ( all 4 panels ) . Finally , as expected , the reporter in the DNA control transgene ( Fig 1A , right column: transgene #3 ) is not regulated by the eve enhancers ( all 4 panels ) , but is regulated by the hebe enhancer ( arrow in bottom panel ) . In this case , the reporter is separated from the hebe enhancer by DNA , not Homie . These results show that Homie induces a long-range interaction between a reporter transgene located many kilobases away and endogenous eve enhancers , and that this interaction is not sensitive to the orientation of the transgene in the chromosome . ( However , this experiment does not test the orientation dependence of the reporter relative to the insulator , as this does not change between these two cases . This is tested below . ) Furthermore , the long-range looping interactions between the transgene and the eve locus do not change the local enhancer blocking activity of the Homie insulator . Since insulator bypass assays show that fly insulators pair with themselves [35–37] , we expected that Homie:Homie pairing is responsible for long-distance regulation . However , as the transgene Homie might also interact with other eve elements , we used 11 primers spanning the locus ( Fig 1B , arrows numbered 0–10 ) for H3C . Fig 1C shows the 3C results for the experimental and for the control DNA transgene , while in the inset we controlled for “non-specific” interactions using data from the DNA transgene as the reference . Whereas there is little interaction between the control transgene and the eve locus ( Fig 1C green line ) , the experimental transgene shows significant interactions with two elements in the locus ( magenta line ) . One is endogenous Homie . The other is located at the 5’ boundary of the eve Polycomb domain [38 , 39] , and , from genome-wide chromatin immunoprecipitation studies [40] , is bound in vivo by many insulator proteins . Based on these and findings below , we call this eve 5’ insulator Nhomie . The experiments in Fig 1 demonstrate that reporter activation by the enhancers in the eve locus is independent of the orientation of the Homie-lacZ transgene in the chromosome . However , this doesn’t mean that reporter activation is independent of the relative orientation within the transgene of Homie and the reporter . To explore this possibility , we generated a transgene with two divergently transcribed reporters , lacZ and GFP ( both are driven by the same eve basal promoter , see Materials and Methods ) . We then inserted Homie in both orientations between the two reporters . Fig 2A shows that in the endogenous eve locus , Homie is located downstream of the eve transcription unit in what we have designated as the “forward” 5’→3’ orientation ( → ) . In transgene #4 ( Fig 2A ) , using this same 5’→3’ convention for the relative orientation of the Homie insulator , the lacZ reporter would be located 5’ with respect to Homie . Thus , in this transgene the relationship between the reporter and Homie is just like the endogenous eve locus where the eve gene is located 5’ to Homie . The eve-GFP reporter is in turn located 3’ to the Homie insulator in the same relative position as the TER94 gene is with respect to the endogenous Homie . In transgene #5 ( Fig 2A ) , the 5’→3’ orientation of Homie is flipped , so that GFP is now located 5’ relative to the Homie insulator , while lacZ is 3’ . Each transgene was then inserted at -142 kb so that GFP is on the same side of Homie as the hebe enhancer , while lacZ is separated from the hebe enhancer by Homie ( see diagrams in Fig 3A and 3B ) . The two reporters in transgene #4 differ dramatically in their pattern ( s ) of expression ( Fig 3A ) . In the case of the lacZ reporter , the eve enhancers activate expression in stripes in the early embryo , as well as in the CNS , mesoderm , and anal plate during mid-embryogenesis ( green arrows ) . The lacZ reporter is not , however , activated by the hebe enhancer , as it is insulated by Homie . A quite different result is observed for the GFP reporter . First , unlike lacZ , it is not subject to regulation by the eve enhancers . Second , it is subject to regulation by the hebe enhancer ( Fig 3A , black arrow ) . In transgene #5 , the target for regulatory interactions with the eve locus is reversed ( Fig 3B ) . Here , GFP is regulated by the eve enhancers ( green arrows ) , while lacZ is not . And , since the orientation of the transgene in the chromosome remains the same , the hebe enhancer still activates GFP ( Fig 3B , black arrow ) , while Homie blocks it from regulating lacZ . These findings indicate that pairing interactions between the transgene Homie and the eve locus are orientation-specific . With respect to the endogenous Homie insulator ( see below for Nhomie ) , the pairing interactions could be head-to-head or head-to-tail ( Fig 3 diagrams ) . In the simplest topological model , head-to-head interactions predict that the lacZ reporter will be activated by eve enhancers when the 5’→3’ orientation of Homie in the transgene places this reporter 5’ of Homie , just as the eve enhancers in the eve locus are 5’ of the endogenous Homie ( Fig 3A , transgene #4 , “stem-loop” topology ) . The GFP reporter will be activated when the orientation of the transgene Homie is reversed ( Fig 3B , transgene #5 , “circle-loop” topology ) . The opposite pattern of activation is expected if Homie interactions are head-to-tail ( Fig 3A and 3B , bottom diagrams ) . In each case , the topology of one of the variants is a stem-loop , while the topology of the other variant is a circle-loop . As can be seen from the expression patterns in Fig 3 , it is head-to-head pairing between the transgene and endogenous Homie that fits the pattern of activation . Why are the regulatory interactions in Fig 3 orientation-dependent , while those in Fig 1 are not ? The difference lies in how we altered the orientation of the insulator in the two experiments . In the experiments in Fig 3 , the 5’→3’ orientation of the Homie insulator in the transgene with respect to the two reporter genes was reversed . In contrast , in the experiments in Fig 1 , the relative 5’→3’ orientation the Homie insulator with respect to the reporter was maintained ( the reporter is 5’ with respect to Homie ) , while the orientation of the entire transgene was flipped . In the experiment in Fig 1 ( illustrated in Fig 2B ) head-to-head pairing between the Homie insulators in the transgene and the eve locus generates either a circle-loop ( top , transgene #1 ) or a stem-loop ( bottom , transgene #2 ) . However , in both cases , the eve enhancers are brought into close proximity to the lacZ reporter . Note that as in Fig 3 , the structure of the loops predicted for head-to-tail pairing of the Homie insulators in the Fig 1 experiments would place the lacZ reporter and the enhancers in the eve locus on opposite sides of the paired insulators , which would not be conducive for productive regulatory interactions ( not illustrated ) . Our 3C experiments identified a second element , Nhomie , in the eve locus that interacts physically with Homie at -142 kb ( Fig 1C ) . We wondered whether Nhomie could also promote long distance regulatory interactions and function as an insulator . To test for these activities , we combined the Nhomie insulator with the lacZ reporter ( Fig 2A ) . The Nhomie:lacZ transgene was inserted at -142 kb so that Nhomie is located between the lacZ reporter and the hebe enhancer ( Fig 4A diagram , transgene #6 ) . Since the relative orientation of Homie and the reporter was critical for productive regulatory interactions , we tested Nhomie in both orientations relative to lacZ . Using the same convention as was used for Homie , the 5’→3’ orientation of Nhomie in the endogenous locus places the eve enhancers and the eve gene 3’ of Nhomie . In transgene #6 ( Fig 4A ) , the 5’→3’ orientation of Nhomie places lacZ is the same position relative to Nhomie as are the eve enhancers and eve gene in the endogenous locus: the reporter is located 3’ relative to Nhomie . In transgene #7 ( Fig 4B ) , Nhomie is in the reverse orientation with respect to lacZ . In this case , the 5’→3’ orientation of Nhomie places the reporter 5’ with respect to the insulator . Our experiments show that Nhomie shares many properties with Homie . Like Homie , it functions as an insulator and blocks the hebe enhancer from activating the reporter ( Fig 4 ) . It is also able to mediate long-distance regulation of the reporter by eve enhancers ( Fig 4A ) . Moreover , as for Homie , these regulatory interactions depend upon the orientation of Nhomie relative to lacZ . However , the orientation of Nhomie with respect to the reporter that engenders robust activation is the opposite that of Homie . For Homie , the reporter is activated when it is located 5’ with respect to the orientation of the insulator , just like the eve gene is 5’ of the endogenous Homie . By contrast , for Nhomie , the reporter is activated when the orientation of the insulator places it 3’ relative to Nhomie; again , just as the eve gene is located 3’ relative to the endogenous Nhomie . Our 3C experiments show that Homie at -142 kb physically interacts with Homie at the 3’ end of the eve locus ( Fig 1C ) . It is clear from bypass experiments that self-interactions like that observed for Homie are not unusual , but instead are a characteristic property of fly insulators [35–37 , 41 , 42] . However , these transgene assays artificially juxtapose homologous partners in cis , as we have done here . In the endogenous setting , homologous partners are only present on the other homolog , and it is in this context that homologous interactions would be biologically relevant . Given that most fly insulators self-interact head-to-head , a plausible idea is that insulators are the elements responsible both for locally aligning homologs in precise register and for maintaining their stable association . The classical evidence for homolog pairing in Drosophila is transvection [43–45] . Transvection is a regulatory interaction that occurs in trans rather than in cis , and requires local pairing of homologs . Typically two mutant alleles complement because the regulatory elements on one homolog activate the gene on the other homolog . Complementation is lost when pairing of the two alleles is disrupted by chromosomal rearrangements [46] . While a special combination of mutations is generally required to detect transvection , trans-regulatory interactions are clearly important for achieving appropriate levels of gene activity in wild-type flies [47] . The hypothesis that homologous insulator:insulator interactions are responsible for the pairing of homologs in register makes two predictions . First , placing homologous insulators in trans should promote transvection . Second , if the homologous interactions of the test insulator are orientation-dependent , transvection is expected to be greater when both copies are oriented in the same direction than when they are oriented in opposite directions . This is expected because self-pairing interactions are likely to be head-to-head rather than head-to-tail . There are two reasons behind this expectation . One is that the self-interactions detected in insulator bypass experiments are typically head-to-head , not head-to-tail [41 , 48 , 49] . The other is that head-to-tail self-interactions between endogenous insulators on each homolog would likely interfere with homolog alignment as well as transvection . To test these predictions , we generated two transgenes , one containing the eve APR and mesoderm ( Me ) enhancers [25] , and the second containing the lacZ reporter . The transgenes were inserted into a site far away from endogenous eve ( on a different chromosome arm , at cytological location 23C4 , where we do not see interactions with endogenous enhancers [26] ) , oriented so that both the enhancers and reporter are on the centromere-distal side of their respective transgene ( Fig 5A ) . In the first experiment , the enhancer transgene had λ DNA , while the reporter had either DNA or Homie . Since there are insulator-like elements near the 23C4 attP site ( one ~50 bp distal to the attP site , another ~8 kb proximal ) [40] , we expected to see some transvection [50 , 51] when either:lacZ or Homie:lacZ is trans to the:enhancer transgene . Fig 5A ( top two panels ) shows that the APR enhancer weakly activates lacZ ( green arrows ) , while there is virtually no Me-driven expression ( red arrows ) . As predicted , the presence of a forward-oriented Homie in the enhancer transgene substantially augments transvection ( Fig 5A , 3rd panel ) . Not only is APR expression much stronger ( green arrow ) , but Me-driven expression is also clearly observed ( red arrow ) . To confirm that stable pairing is head-to-head , we reversed Homie in the enhancer transgene ( maintaining the overall transgene orientation in the chromosome ) . In this configuration , head-to-head pairing would introduce an S-shaped double loop . As illustrated in Fig 5A ( “twisted pairing” ) , this would place the reporter on the opposite side of the paired insulators from the transgenic enhancers . This configuration would not be expected to increase enhancer-reporter interactions . Consistent with this prediction , reporter expression is about the same ( Fig 5A , bottom panel ) as in the negative controls carrying DNA ( Fig 5A , top two panels ) . Alternatively , the need to form such a double loop might make this pairing interaction less stable than for the other orientation , when the head-to-head pairing reinforces the normal pairing of the homologs ( “tightly paired” in the diagram ) . In fact , evidence below is more consistent with such “twisted pairing” interactions forming only transiently , or not at all ( hence the red “X” in the diagram for “twisted pairing” ) . This is in line with the expectation , stated above , that head-to-head self-interactions between endogenous insulators mediate homolog alignment and pairing , while head-to-tail self-interactions are incompatible with smooth alignment and tight pairing . To further explore the relationship between pairing direction and transvection , we generated dual reporters with divergently transcribed GFP and lacZ that have either DNA or Homie inserted between the reporters ( Fig 5B ) . When the DNA:dual reporter is trans to the Homie-enhancer transgene , the APR enhancer weakly stimulates lacZ and GFP in the APR , while neither reporter is activated by the Me enhancer . The addition of Homie to the reporter ( in the same orientation in the chromosome as that of Homie in the enhancer transgene ) substantially enhances APR lacZ transcription , and turns on lacZ in the mesoderm . By contrast , there is only a slight increase in APR GFP expression , while mesoderm expression is detectable , but only weakly . The differences in transvection for the two reporters are consistent with the topology generated by head-to-head , not head-to-tail pairing ( Fig 5B , “tightly paired” ) . We also combined the dual reporter with an enhancer transgene in which the entire transgene containing Homie and the enhancers are flipped ( Fig 5B , bottom panel ) . Head-to-head pairing of Homie would generate an S-shaped double loop ( as diagrammed in Fig 5B , “twisted pairing” ) . In this case , there is little or no enhancement of transvection for either reporter , suggesting that the introduction of such a double loop between the paired homologs either is unstable or does not form ( indicated by the red “X” in the diagram for “twisted pairing” ) . We note that there are some subtle differences in the expression patterns for transgene combinations in which transvection is not significantly enhanced . This includes all the cases where our topology diagrams are labeled as “loosely paired” or “twisted pairing” . These differences may be due to a combination of several factors , such as differences in the size of the transgenes , weak or unstable interactions with insulators near the site of transgene insertion , or the shielding of transgenic reporters from position effects that weakly upregulate or downregulate reporter activity . We next tested whether Nhomie self-interactions in trans also induce transvection . Nhomie was oriented in the single reporter transgene so that the lacZ reporter ( diagrammed in Fig 6A ) is 3’ with respect to Nhomie . It was then combined in trans with an enhancer transgene that had or Nhomie ( in the same 5’→3’ orientation in the chromosome ) so that the two enhancers are 3’ of Nhomie ( Fig 6A ) . In the:Nhomie combination , the APR enhancer drives only weak expression , and activation by the Me enhancer is not seen . As would be predicted if head-to-head pairing aligns the enhancers and the reporter , lacZ expression is substantially elevated in the Nhomie:Nhomie combination . This conclusion is confirmed by the dual reporter assay . As shown in the lower half of Fig 6B , head-to-head pairing of Nhomie in the enhancer and dual reporter transgenes would juxtapose the Me and APR enhancers with the lacZ reporter , while the GFP reporter would be separated from the enhancers by the paired Nhomie insulators . In this configuration , the Me and APR should preferentially drive lacZ expression , not GFP expression , and this is what is observed . While Homie-Homie or Nhomie-Nhomie self-interactions normally occur at the endogenous eve locus only in trans , this is not the case for Nhomie-Homie interactions . Heterologous interactions between neighboring insulators in cis are thought to be responsible for subdividing chromosomes into a series of topologically independent domains , and are expected to occur all along the chromosome . Like self-interactions , heterologous interactions are known to be specific [37 , 42 , 52 , 53] , and consequently are likely also orientation-dependent . For heterologous insulators interacting in cis , we define their endogenous directionalities to be the same . That is , the arrows that represent them point in the same “forward” direction along the chromosome ( as in Figs 1–4 for endogenous Nhomie and Homie ) . Using this convention , at the endogenous eve locus , head-to-tail interactions between Nhomie and Homie would generate a stem-loop , while head-to-head interactions would generate a looped circle or “circle-loop . ” To test whether these two insulators can interact with each other independently of the eve locus , and ( if so ) determine their orientation dependence , we combined a Nhomie-lacZ reporter with two different Homie-enhancer transgenes . In the one in which the enhancers are 5’ of Homie ( Fig 6C , top panel ) , head-to-tail pairing with Nhomie should align the enhancer and reporter , and favor transvection . When the enhancer transgene has Homie in the reverse orientation ( Fig 6C , bottom panel ) , enhancer-reporter alignment would be favored by head-to-head pairing . Fig 6C shows that Nhomie and Homie can pair with each other in a foreign context ( top panel ) , and that transvection is favored by head-to-tail pairing ( top panel vs . bottom panel ) . These findings parallel those for self-pairing ( Figs 5 and 6A and 6B ) , except that heterologous pairing is head-to-tail rather than head-to-head . To confirm these results , we combined the dual lacZ , GFP reporter containing Homie with an enhancer transgene containing Nhomie . As illustrated in Fig 6D , head-to-tail pairing of Nhomie and Homie would juxtapose the enhancers with lacZ , while head-to-head pairing would juxtapose the enhancers with GFP . Consistent with head-to-tail pairing , lacZ transvection is stimulated , while GFP is not ( compare 6D with the control in the upper half of 6B ) . The insulator interactions in the transvection assay are local and likely facilitated by homolog pairing . To confirm that the eve insulators can interact specifically with themselves and with each other over large chromosomal distances , we took advantage of attP 25C1 , located 2 Mb distal to 23C4 . A Homie:lacZ transgene was inserted at 25C1 . It was combined with an enhancer transgene at 23C4 containing either DNA or Homie ( Fig 7A ) . No interaction between the transgenes is evident with the DNA control or when is replaced by the su ( Hw ) insulator . On the other hand , when both the reporter and the enhancer have a Homie insulator , the APR enhancer is able to activate lacZ expression ( Fig 7A , upper left panel ) . This result is consistent with previous studies which showed that APR was the only enhancer in the endogenous eve locus that could act over distances >1 Mb with Homie-carrying transgenes [26] . As would be expected from the orientation dependence of insulator self-pairing , when Homie is inverted within the enhancer transgene ( Fig 7A , upper right panel ) , expression is not seen , confirming that Homie-Homie pairing is head-to-head . We also tested whether Nhomie can mediate distant regulatory interactions either with itself or with Homie . In the two transgenes used to test Nhomie self-interactions , the enhancers or reporter , respectively , are each located 3’ relative to the adjacent insulator . These Nhomie transgenes were inserted ( separately ) at 23C4 and 25C1 , then crossed into the same animals . Fig 7B shows that Nhomie:Nhomie interactions can mediate long-distance activation of lacZ by the APR enhancer ( Fig 7B , lower right ) . Nhomie also pairs with Homie , enabling the APR enhancer in the Nhomie transgene at 23C4 to activate a Homie-lacZ reporter at 25C1 ( Fig 7B , upper left ) . As illustrated in Fig 7 , these interactions are all consistent with the orientation dependence seen in the other assays , namely head-to-head self-pairing and head-to-tail heterologous pairing . The experiments described above indicate that Nhomie and Homie must be able to physically pair with each other , and do so in a head-to-tail orientation . In the endogenous locus , head-to-tail pairing would generate a stem-loop containing the eve transcription unit and its associated enhancers and Polycomb silencers , linked together at the base by the Nhomie and Homie insulators . 3C experiments with Homie as the anchor confirm that Nhomie and Homie contact each other in the eve locus ( Fig 1D ) .
Homolog pairing requires mechanisms for aligning homologs in precise register , and maintaining their stable association . While many schemes are imaginable , the simplest utilizes elements distributed along each homolog that have self-interaction specificity . Such a mechanism would be consistent with the persistence of local pairing and transvection in chromosomal rearrangements [44 , 56–60] . It would also fit with studies on the pairing process [56 , 61 , 62] . Self-association of pairing elements would locally align sequences in register , and ultimately link homologs together along their entire length . In this mechanism , self-association must be specific and also directional , namely head-to-head . This avoids the introduction of unresolvable loops and maximizes pairing for transvection . In Drosophila , the homing of P-element transgenes , in which normally random insertion becomes targeted , suggested the ability of genomic elements to self-interact . Such a homing activity was found in the engrailed locus for a region that includes two PREs [63–65] , and later studies showed that some insulators [26 , 66 , 67] and a promoter region [68] also possess homing activity . The self-interaction implied by homing suggests that these elements may facilitate homolog pairing . However , in contrast to PREs and promoters , insulators have consistently been found to engage in specific self-interactions ( see below ) . Thus , among the known elements in the fly genome , insulators are the best candidates to align homologs in register and maintain pairing [20 , 21] . Moreover , genome-wide chromatin immunoprecipitation experiments ( ChIPs ) show that insulators are distributed at appropriate intervals along each chromosome [18 , 19] . A role in homolog pairing was first suggested by the discovery that the su ( Hw ) and Mcp insulators each can mediate regulatory interactions between transgenes inserted at distant sites [69 , 70] . The Fab-7 insulator can also mediate long-range regulatory effects [71] . Further evidence that self-association is characteristic of fly insulators came from insulator bypass experiments [35 , 36] . These experiments showed that bypass is observed when an insulator is paired with itself , while heterologous combinations are less effective or don’t give bypass [37 , 41 , 42 , 48 , 72 , 73] . Moreover , self-pairing is , with few exceptions , head-to-head . That insulators mediate homolog pairing through specific self-interactions is further supported by our studies . Using a classical transvection assay , we found that Homie-Homie and Nhomie-Nhomie combinations stimulate trans-regulatory interactions between enhancers on one homolog and a reporter on the other ( Figs 5 , 6A and 6B ) . Moreover , the parameters that favor transvection dovetail with those expected for a pairing mechanism based on insulator self-interactions in trans . First , the two insulators must be in the same orientation . When they are in opposite orientations , transvection is not enhanced ( or enhancement is much weaker , Fig 5 ) . Second , the enhancers and reporter must be located on the same side ( centromere proximal or distal ) of the insulators ( Figs 5 , 6A and 6B ) . In addition to transvection , Homie and Nhomie also engage in highly specific and directional distant regulatory interactions ( Fig 7 ) . While there is compelling evidence that insulator self-interactions are responsible for homolog pairing , many issues remained unresolved . Perhaps the most important is the nature of the code used for self-recognition and orientation . The best hint comes from bypass experiments using multimerized binding sites for Su ( Hw ) , dCTCF , or Zw5 . Homologous multimer combinations give bypass , while heterologous combinations do not . However , bypass is observed for composite multimers when they are inserted in opposite orientations ( e . g . , Su ( Hw ) dCTCF ↔ dCTCF Su ( Hw ) ) , but not the same orientation ( e . g . , Su ( Hw ) dCTCF →→ Su ( Hw ) dCTCF ) [53] . These findings argue that the identity and order of proteins bound to the insulator determine its self-association properties . The first direct evidence that insulators generate loops came from 3C experiments on the mouse β-globin and the fly 87A7 heat shock loci [23 , 74] . These studies suggested that physical interactions between adjacent insulators in cis could subdivide chromosomes into looped domains . Subsequent work has confirmed this conclusion [17] . However , while these experiments demonstrate that cis insulator interactions generate loops , they provided no information about the topology of these loops , or how they are arranged . Cis interactions could , a priori , be either head-to-head like self-association in trans , or head-to-tail . The consequences are quite different . Head-to-head interactions generate a circle-loop , while head-to-tail interactions generate a stem-loop ( Fig 8A and 8D , respectively ) . If heterologous insulators interact with only one specific partner , the circle-loop or the stem-loop will be linked to neighboring circles or stem-loops by loops without anchors . These unanchored loops would correspond to the main axis of the chromosome , and the circle-loops or stem-loops would then protrude from the main axis in a random orientation and at distances determined by the length and compaction of the unanchored loops . On the other hand , if insulators in a chromosomal segment are able to interact with both of their neighbors , then the main axis of the chromosome in this region would be defined by the insulators . Quite different structures are predicted for head-to-head and head-to-tail interactions ( Fig 8B and 8E ) . Head-to-head would give a series of variably sized circle-loops linked together at their base by an array of interacting insulators . The base would correspond to the main axis of the chromosome , and each circle-loop would extend from one side of the main axis to the other . If the direction of coiling were always the same , this would give a structure resembling a helix anchored to a rod ( Fig 8B ) . If the direction of coiling were random , the structure would be more complicated and variable , since neighboring circle-loops could extend out from the main axis in either the same or the opposite direction ( not illustrated ) . The loop-axis relationship would be more regular for head-to-tail insulator pairing in cis . Adjacent stem-loops would extend out from the main axis in opposite directions much like the lampbrush chromosomes formed when haploid sperm heads are injected into amphibian oocytes ( Fig 8E ) [75] . This stem-loop organization would also fit with the radial loop model proposed by Laemmli and others for the first level of folding of metaphase chromosomes [7 , 11] . Since our experiments show that Homie-Nhomie association is head-to-tail , the topology of the eve locus in vivo is a stem-loop , not a circle-loop . This finding raises a number of questions . Perhaps the most important is whether head-to-tail interactions are the rule rather than the exception . While the orientation dependence of homologous interactions has been extensively investigated , there have been no systematic studies on interactions between neighboring insulators . However , there are reasons to think that cis interactions are more likely head-to-tail than head-to-head . One is homolog pairing . As mentioned above , the circle-loops formed by head-to-head interactions can coil in either direction , either left-handed or right-handed . If coiling were random , then about half of the circle-loops on each homolog would be coiled in opposite directions . In this case , head-to-head pairing of homologous insulators in each homolog would generate a structure in which the circle-loops would point in opposite directions ( Fig 8C , left circles ) . This topology would not be compatible with transvection . Coiling of the circle-loops in the same direction on both homologs would permit interdigitation of one circle-loop inside the other ( Fig 8C , right circles ) ; however , the chromatin fiber from the inside circle-loop would need to cross in on one side and out on the other . If the main axis of the chromosome in the paired region is defined by a series of interacting insulators in cis , then generating a topology permissive for transvection ( not illustrated ) would require coiling of successive homologous circle-loops on each homolog in the same direction , one inside the other ( Fig 8C , right circles ) . These topological issues aren’t encountered when heterologous insulator interactions in cis are head-to-tail . Head-to-head pairing of homologous insulators in trans would bring regulatory elements and genes in the two homologous stem-loops into close proximity . Alignment of the two homologs is straightforward whether or not the main axis of the chromosome is defined by a series of interacting insulators ( Fig 8F illustrates one of these cases ) . Alternating loops extending upwards and downwards from the main axis of the chromosome would be directly aligned when homologous insulators pair head-to-head in trans . While the requirements for aligning and pairing homologs would appear to favor stem-loops between heterologous insulators in cis in flies , homolog pairing does not occur in vertebrates except in specialized cell types [76] . This could mean that circle-loops formed by cis interactions between heterologous insulators are permissible in vertebrate chromosomes . However , even in organisms in which homolog pairing doesn’t occur in somatic cells , it seems possible that cis-pairing interactions more commonly generate stem-loops than circle-loops . First , following DNA replication and before mitosis ( during the S and G2 phases of the cell cycle ) , sister chromatids are aligned . Maintaining this alignment may facilitate epigenetic mechanisms that template chromatin structures from one cellular generation to the next , such as the copying of histone modifications onto both daughter chromosomes . The simpler topology of stem-loops could facilitate this sister chromatid pairing , as well as their separation during mitosis . Second , recent studies on the relationship between loop domains and CTCF insulators showed that in more than 90% of the cases , the CTCF binding sites on opposite ends of a loop are in opposite orientation [17] . Thus , assuming that the orientation of pairing is such that the CTCF sites are aligned in parallel to form the loop , pairing between CTCF insulators at the ends of the loop would generate stem-loops rather than circle-loops . If insulators form the main axis of the chromosome , there is an additional explanation for such a bias . As shown in Fig 8B , head-to-head pairing in cis could generate a series of circular loops that extend out from the same side of the main axis . This configuration would be favorable for crosstalk between regulatory elements and genes in adjacent loops . By contrast , head-to-tail pairing , where adjacent stem-loops extend out in opposite directions ( Fig 8E ) , would disfavor crosstalk , helping to explain how insulators block enhancer-promoter communication between adjacent loops .
See S1 Fig for the Homie and Nhomie regions used , control DNA , and tag sequences . Reporters contain the eve basal promoter , -275 to +106bp from eve +1 ( TSS ) , either the lacZ or GFP coding region , and the eve 3'-UTR , +1300 to +1525 bp . Enhancers are: eve APR , +3 . 0 to +4 . 1 kb; eve Me , +5 . 7 to +6 . 6 kb , each cloned in plasmid attB∆2 [32] for transgenesis [77] using ΦC31 [78] . Target sites were: -142 kb from eve [26]; 23C4 ( 2L;3029226 ) , generated by us; and 25C1 [77] . Sequence coordinates are Flybase version dm6 [79] . Two genomic fragments used in this study that span the insulator protein binding region we call Nhomie , based on genome-wide studies [40] , were found to have indistinguishable function in our assays ( S2 Fig ) . The corresponding sequences are given in S1 Fig . RNA in situ hybridization and anti-β-galactosidase staining were as described [25] . In all cases , conclusions drawn were based on comparisons between control and experimental collections of embryos that were stained in parallel . H3C analysis was performed as described [34] , with the following modifications . Embryos ( 200 μl aged 0-6h at ~23°C ) were cross-linked in either 2% or 3% formaldehyde for either 15 or 30 min ( each gave similar results , and were included in the data presented ) , digested with 100U each of EcoRI ( Roche ) and MfeI ( NEB ) at 37°C overnight . About half of the material was ligated ( Takara , 3500U ) for >4 hr . at ~23°C , and incubated at 65°C overnight to reverse cross-links . Following RNase A ( Roche , 40μg/sample ) and proteinase K ( Roche , 220μg/sample ) digestions , purified DNA ( 20ng/reaction ) was subjected to real-time PCR analysis using SYBR Green Master Mix ( Roche ) . All transgenes inserted at -142 kb used for 3C analysis had the same tag sequence , which was used as the anchor primer ( Fig 1C ) , in combination with each of a series of accompanying primers from within the eve locus . To identify Homie-interacting regions within endogenous eve , an endogenous Homie fragment-specific primer was used as anchor ( Fig 1D ) , along with the same series of accompanying primers . These sequences are given in S1 Fig . PCR quantification was done as described [34] , with the following set-up . The fragments in the eve locus created by EcoRI and MfeI digestion were cloned into anchor fragment-carrying plasmids , and served as standards for the expected ligation products . These plasmids were linearized and mixed with equimolar amounts of digested genomic DNA . Details of the various controls , such as the choice of primers and enzymes , were appropriate for each specific experiment [34] . Additional details are given in the legend to Fig 1 .
|
The chromosomes of multicellular animals are organized into a series of topologically independent looped domains . This domain organization is critical for the proper utilization and propagation of the genetic information encoded by the chromosome . A special set of architectural elements , called boundaries or insulators , are responsible for both subdividing the chromatin fiber into discrete domains , and determining the topological organization of these domains . Central to the architectural functions of insulators are heterologous and homologous insulator:insulator pairing interactions . In Drosophila , the former defines the topology of individual looped domains in cis , while the latter dictates the process of homolog alignment and pairing in trans . Here we use two insulators from the even skipped locus to elucidate the principles governing these two architectural functions . These principles align with several longstanding observations , and resolve a number of conundrums regarding chromosome topology and function .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"homologous",
"chromosomes",
"chromosome",
"structure",
"and",
"function",
"developmental",
"biology",
"insulators",
"materials",
"science",
"epigenetics",
"embryos",
"chromatin",
"embryology",
"chromosome",
"biology",
"gene",
"expression",
"materials",
"by",
"attribute",
"chromosome",
"pairs",
"genetic",
"loci",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"chromosomes"
] |
2016
|
Determinants of Chromosome Architecture: Insulator Pairing in cis and in trans
|
We reviewed studies that analyzed cysticercosis ( CC ) , neurocysticercosis ( NCC ) and epilepsy across Latin America , Asia and Sub-Saharan Africa , to estimate the odds ratio and etiologic fraction of epilepsy due to CC in tropical regions . We conducted a systematic review of the literature on cysticercosis and epilepsy in the tropics , collecting data from case-control and cross-sectional studies . Exposure criteria for CC included one or more of the following: serum ELISA or EITB positivity , presence of subcutaneous cysts ( both not verified and unverified by histology ) , histology consistent with calcified cysts , and brain CT scan consistent with NCC . A common odds-ratio was then estimated using meta-analysis . 37 studies from 23 countries were included ( n = 24 , 646 subjects , 14 , 934 with epilepsy and 9 , 712 without epilepsy ) . Of these , 29 were case-control ( 14 matched ) . The association between CC and epilepsy was significant in 19 scientific articles . Odds ratios ranged from 0 . 2 to 25 . 4 ( a posteriori power 4 . 5–100% ) and the common odds ratio was 2 . 7 ( 95% CI 2 . 1–3 . 6 , p <0 . 001 ) . Three subgroup analyses performed gave odds ratios as: 2 . 2 ( EITB-based studies ) , 3 . 2 ( CT-based studies ) , 1 . 9 ( neurologist-confirmed epilepsy; door-to-door survey and at least one matched control per case ) . Etiologic fraction was estimated to be 63% in the exposed group among the population . Despite differences in findings , this meta-analysis suggests that cysticercosis is a significant contributor to late-onset epilepsy in tropical regions around the world , and its impact may vary depending on transmission intensity .
Cysticercosis ( CC ) is a parasitic infection caused by the larva stage ( cysticercus ) of the tapeworm Taenia solium . It has been a major public health problem since historical times [1] , and remains so , particularly in the developing world ( low-and middle-income countries; LMIC ) , due to inadequate hygiene , rudimentary pig management and slaughter , and poor waste water management [2] . Developed regions such as Europe and North America are considered to be virtually free of endemic transmission , although there remains a substantial disease burden in these regions due to migration [3] . Neurocysticercosis ( NCC ) is considered a common helminthic infection of the central nervous system ( CNS ) across Latin America , Sub-Saharan Africa and Asia [4–7] , and a common cause of late-onset epilepsy in LMIC [4 , 5 , 8] . For instance , a study in Burundi showed a strong link between CC and epilepsy , with an etiologic fraction of 50% ( 95% CI: 42–57 ) and an odds ratio of 3 . 8 ( 95% CI: 2 . 5 to 5 . 1 ) [6] . It was estimated in a recent meta-analysis that people infected with CC in Sub-Saharan Africa ( SSA ) are at 3 . 4–3 . 8 fold greater risk of having epilepsy [7] . It is noted that despite the importance of these diseases at an individual and population level , there are still discrepancies in the literature about their precise impact [4] . Moreover , earlier reviews focused on specific regions alone [7] . We conducted a review of studies that analyzed CC , NCC , and epilepsy across Latin America , Asia and Africa , to estimate the probability and etiologic fraction of epilepsy due to CC in tropical regions .
Systematic searches were conducted for articles in English and French using the following databases: Medline , Scopus , Science Direct , Ingenta Connect , Refdoc ( formerly Article Science ) . We also searched for articles and theses in the bibliographic database of the Institut d’Epidemiologie et de Neurologie Tropicale http://www . unilim . fr/ient/ . Keywords used were ( cysticercosis OR Taenia solium OR neurocysticercosis ) AND epilepsy . Logical operators ( AND , OR , NOT ) were used . Bibliographies of published reviews and meta-analyses were also searched . We conducted three subgroup analyses by taking into account those studies that used specific diagnostic tools for CC or NCC and epilepsy . The first group comprised studies that used EITB to determine CC exposure . The second analysis included studies that used brain CT scan to assess NCC exposure . The third analysis involved studies that had used standardized diagnostic methods to confirm epilepsy in population-based studies , such as neurological surveys applied in a door-to-door fashion with evaluation by well-trained general practitioners and /or neurological evaluation to confirm cases , and including at least one matched control per case . Finally , we performed an analysis by continent . For each of the selected studies , the odds ratio ( OR ) and its 95% confidence interval was determined using Epi-Info 6 ( Centers for Disease Control and Prevention , Atlanta , USA ) . A meta-analysis was used to estimate the risk of developing epilepsy when exposed to CC , applying a random-effects model using Stata software , version 10 . 1 ( Stata-Corp , College Station , TX , USA ) to account for the variance of each included study [9] . Odds ratios ( OR ) and 95% confidence intervals ( 95% CIs ) were determined . Homogeneity was assessed by I squared tests . Subgroup analyses were also conducted for studies ascertaining CC by EITB assays , CT scan and those studies that followed certain requirements for determining epilepsy ( as mentioned under Methods , above ) . Because epilepsy has multiple causes and associated factors , we calculated the etiologic fraction ( EF ) i . e . the proportion of cases “attributable” to cysticercosis , by comparing the prevalence among exposed and the unexposed . The EF provided an unadjusted estimate of the proportion of cases of epilepsy that could be prevented if exposure to CC were eliminated . The etiologic fraction was based on the pooled estimate of risk , rather than single risk estimates for individual studies , by using the following formula: proportion exposed ( common OR-1 ) /proportion exposed ( common OR-1 ) +1 .
Twenty studies defined epilepsy , of which 18/20 ( 90% ) followed at least one definition recommended by the International League against Epilepsy ( ILAE , 1981 , 1989 , 1993 , 2006 ) . One study each used definitions proposed by the World Health Organization and that recommended for LMICs [10] . As noted in Table 1 , there was great variability in the tools used for assessing exposure to CC , ranging from physical examination of subcutaneous nodules to Computed Tomography of the brain ( CT ) , MRI images , cyst histology , and bioassays in serum or cerebrospinal fluid ( CSF ) . A total of 21/37 ( 56 . 75% ) studies determined exposure to CC by detecting antibodies or antigens in serum using ELISA or ELISA-Ag . Seven studies used EITB to confirm or refute the results of ELISA and 12 studies used only EITB to determine CC exposure . One study used a hemagglutination test with sheep red blood cells sensitized to cysticercus antigens to determine exposure to CC . NCC exposure was determined by measuring antibodies in the CSF , but only 5/37 studies ( 13 . 51% ) did so , by using ELISA ( n = 3 ) or EITB ( n = 2 ) . CT was used in 14 studies , including 13 to assess NCC , and one [11] focused on the soft parts of the thigh . As shown in Table 2 , the association between CC and epilepsy was statistically significant in 19 studies , leaving 18 with a non-significant association . The association was in fact nearly significant for two studies [12 , 13] . Odds ratios ranged from 0 . 2 to 25 . 4 and the a posteriori statistical power ranged from 4 . 5% to 100 . 0% . A meta-analysis of 37 studies based on the determination of exposure through detection of antibodies by ELISA or EITB , antigen detection by ELISA , or CT findings , is shown in Fig 3 . The common odds ratio was estimated to be 2 . 7 ( 95% CI 2 . 1–3 . 6 ) , p<0 . 001 . Heterogeneity was substantial with a I squared at 78% ( p<0 . 0001 ) . Three subgroup analyses were also performed as detailed in the methods section above . The first was based on studies that used EITB ( n = 19 ) , Fig 4A . The common odds ratio obtained was 2 . 2 ( 95% CI 1 . 6–3 . 0 ) , p<0 . 001 . Another subgroup analysis ( Fig 4B ) was based on studies that used brain computer tomography ( n = 8 ) . This gave a common odds-ratio of 3 . 2 , ( 95% CI 2 . 5–4 . 1 , p<0 . 001 ) . The third subgroup analysis , Fig 4C , was based on the methods used to confirm epilepsy ( n = 13 ) , and gave a common odds ratio of 1 . 9 ( 95% CI 1 . 2–3 . 0 ) , p<0 . 001 . We also performed an analysis by continent , showing that the effect was quite similar around the world ( Fig 3 ) . The etiologic fraction was estimated to be 63 . 0% ( 95% CI: 61 . 4–64 . 5 ) in the exposed group among the population . In other words , 63% of epilepsies were reportedly due to CC .
We propose that future field interventions should meet basic requirements to be more useful:
Cysticercosis is an active helminthic infection common in tropical regions . Many questions are still unanswered and there are still many limitations in epidemiological base-studies . Based on the current data , NCC is significantly associated with symptomatic epilepsy in low and middle-income countries . However , the strength of this association certainly varies depending on the transmission intensity ( rural areas , poor sanitation , lack of potable water , etc ) . More meta-analyses that are meaningful require good quality studies in tropical regions following certain basic methodological requirements listed above . Finally , epilepsy attributable to CC is preventable . There is a need to focus our efforts on research , control and prevention of CC to avoid increased costly neurological morbidity of this zoonotic disease .
|
Cysticercosis is a helminthic infection of the central nervous system ( CNS ) and the leading cause of late onset epilepsy in low-and middle-income countries . This neurological disease is a public health problem in Sub-Saharan Africa , Asia and Latin America , affecting impoverished rural and peri-urban populations where sanitation is inadequate . Diagnostic criteria for NCC vary according to regional availability and access to imaging and serological tests . This systematic review highlights the lack of appropriate methodology in most observational studies , with few studies including control groups , a basic epidemiological criteria needed to demonstrate an association . EITB for cysticercosis was widely used to measure exposure , and brain CT scan was a non-invasive alternative used to identify cysts . Neurocysticercosis ( NCC ) is a preventable neurological condition in the tropics despite resource limitations in LMIC regions . Well-designed studies are needed to provide quality evidence to support control interventions and surveillance systems for this important zoonotic disease .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Conclusions"
] |
[
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"diagnostic",
"radiology",
"tropical",
"diseases",
"geographical",
"locations",
"parasitic",
"diseases",
"neuroscience",
"mathematics",
"statistics",
"(mathematics)",
"neglected",
"tropical",
"diseases",
"immunologic",
"techniques",
"neuroimaging",
"africa",
"research",
"and",
"analysis",
"methods",
"epilepsy",
"imaging",
"techniques",
"mathematical",
"and",
"statistical",
"techniques",
"immunoassays",
"tomography",
"computed",
"axial",
"tomography",
"people",
"and",
"places",
"helminth",
"infections",
"radiology",
"and",
"imaging",
"diagnostic",
"medicine",
"asia",
"neurology",
"cysticercosis",
"meta-analysis",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"statistical",
"methods"
] |
2017
|
Systematic review and meta-analysis estimating association of cysticercosis and neurocysticercosis with epilepsy
|
Halobacterium salinarum is a bioenergetically flexible , halophilic microorganism that can generate energy by respiration , photosynthesis , and the fermentation of arginine . In a previous study , using a genome-scale metabolic model , we have shown that the archaeon unexpectedly degrades essential amino acids under aerobic conditions , a behavior that can lead to the termination of growth earlier than necessary . Here , we further integratively investigate energy generation , nutrient utilization , and biomass production using an extended methodology that accounts for dynamically changing transport patterns , including those that arise from interactions among the supplied metabolites . Moreover , we widen the scope of our analysis to include phototrophic conditions to explore the interplay between different bioenergetic modes . Surprisingly , we found that cells also degrade essential amino acids even during phototropy , when energy should already be abundant . We also found that under both conditions considerable amounts of nutrients that were taken up were neither incorporated into the biomass nor used as respiratory substrates , implying the considerable production and accumulation of several metabolites in the medium . Some of these are likely the products of forms of overflow metabolism . In addition , our results also show that arginine fermentation , contrary to what is typically assumed , occurs simultaneously with respiration and photosynthesis and can contribute energy in levels that are comparable to the primary bioenergetic modes , if not more . These findings portray a picture that the organism takes an approach toward growth that favors the here and now , even at the cost of longer-term concerns . We believe that the seemingly “greedy” behavior exhibited actually consists of adaptations by the organism to its natural environments , where nutrients are not only irregularly available but may altogether be absent for extended periods that may span several years . Such a setting probably predisposed the cells to grow as much as possible when the conditions become favorable .
Halobacterium salinarum is a halophilic archaeon that thrives in extremely saline environments with salt concentrations reaching 4 M or higher . The organism is perhaps most well known for its retinal-protein bacteriorhodopsin ( BR ) , which is a light-driven proton pump . BR is the only known nonchlorophyll structure that allows photosynthesis [1] . It is currently being developed for applications in optical security [2] , optical data storage [3] , and holography [4] . Accordingly , H . salinarum's photosynthetic capabilities are its most well-studied aspects . For example , the 3D structure of BR has been resolved , and its complete catalytic cycle elucidated at the molecular level ( reviewed in [5] ) . However , despite the focus on BR , photosynthesis is not the only means by which H . salinarum can generate energy . Respiration [6] , [7] as well as the fermentation of arginine [8] , [9] are other mechanisms utilized by the organism . This bioenergetic flexibility makes the archaeon a good model system for investigating the interplay between different energy production modes . H . salinarum is also one of the few reported organisms that can use potassium gradients for long term energy storage in a battery-like manner [10] . The metabolic network of an organism can be reconstructed from genomic , biochemical , and physiological data [11] , [12] , [13] . This network consists of the known and hypothesized reactions that take place within the organism , and is considered to be on a genome-scale when most or all of the genes with known metabolic function are included [14] . We , in a previous study , have reconstructed and proposed such a network for Halobacterium salinarum [15] . In addition to the immediate information gained from metabolic reconstructions , these networks can be analyzed to gain insights on emergent system properties through the use of appropriate computational methods . In this respect , the constraints-based framework has emerged as an important and convenient tool for modeling such systems because it does not require the detailed information typically required by full kinetic models . Rather , constraints-based models require only generally available physicochemical information such as stoichiometry , reversibility , energy balance , and , when available , reaction velocities [16] , [17] , [18] . One of the methods available under the constraints-based framework is Flux Balance Analysis ( FBA ) . Essentially , FBA uses linear optimization to find a flux distribution that maximizes a particular objective function , e . g . , growth rate or ATP production [19] , [20] . It has been shown that such optimality principles , within limits and under defined conditions , can describe the operation of metabolic networks , including the prediction of internal fluxes [21] , [22] . Extensions to FBA include hybrid models that introduce some degree of dynamics through the integration of time-variant input rates to the static model [15] , [19] , [20] . Our aim in this study is two-fold . First , we set out to investigate the interplay between energy generation , nutrient utilization and biomass production under different bioenergetic modes . Second , we also analyzed the relationships between the different energy producing mechanisms of respiration , photosynthesis and fermentation themselves , which are typically examined individually . To achieve these , we used a genome-scale metabolic network that connects the different aspects . Our results include several findings that are contrary to assumptions which are typically made; particularly with respect to the utilization of nutrients , and how the bioenergetic modes operate . From a more methodological perspective , we also sought out to extend the existing framework for hybrid genome-scale metabolic models to handle biological systems where nutrient utilization and growth rates vary with time . Such changes in nutrient consumption , for example , can be the result of the differences between growth phases , or can arise from the interactions between the supplied metabolites . We demonstrate that the extended methodology not only accounts for such dynamics , but in several instances actually led to the identification of the underlying causes .
Most of the reactions in the genome-scale metabolic network we use in this study are taken directly from previous reconstructions [15] , [23] . After updating some pathways , the final network is now composed of 664 reactions ( 567 internal and 97 transport ) —covering 478 genes—and 545 metabolites . One of the most significant pathways that was modified is the respiratory chain . The proposed oxidative phosphorylation pathway is shown in Figure 1 . H . salinarum has analogs of all five complexes associated with oxidative phosporylation found in mitochondria and E . coli ( complexes I through V ) . However , there are significant differences from its better-studied counterparts . For example , the subunits of complex I that comprise the NADH acceptor module ( nuoEFG ) could not be assigned , similar to what has been observed for Sulfolobus solfataricus [24] . Indeed , it has been experimentally excluded that NADH is oxidized by a type I dehydrogenase in H . salinarum . Rather , NADH is oxidized by a non-homologous type II NADH dehydrogenase that also reduces quinones but is incapable of proton translocation [6] . Nevertheless , the conservation of eleven complex I subunits with high levels of sequence similarity make it likely that the complex is functional and translocates protons . Moreover , the lack of the NADH-specific acceptor module and the experimental evidences that NADH is not oxidized by complex I make it likely that the complex I analog actually accepts electrons from another donor molecule , which we left unspecified in Figure 1 . In addition , the H . salinarum pathway is also different in that menaquinone , rather than ubiquinone , is the likely mobile carrier used to shuttle electrons from the complex I analog to the complex III analog . A similar proposal has been made for the closely related organism Natronomonas pharaonis [25] , [26] . Two other important differences of the proposed pathway from the systems in mitochondria and E . coli are the composition of the complex III analog and the mobile carrier which carries electrons to the terminal oxidase ( complex IV analog ) . The cytochrome c1 subunit ( petC ) of E . coli complex III , responsible for transfering electrons to the mobile carrier of the organism , cytochrome c , does not seem to have a homolog in H . salinarum . In fact , genes which code for cytochrome c could also not be assigned . Accordingly , we believe that H . salinarum likely uses a different carrier . We propose this to be halocyanin , which is a blue copper protein originally isolated from N . pharaonis [27] . A similar function has been proposed for the molecule in N . pharaonis , based on its localization as a membrane protein and its midpoint potential that is consistent with a mobile carrier's [25] . This proposed function is further supported by the fact that the halocyanin gene , hcpB , is fused with the cbaD subunit of complex IV in H . salinarum [28] . Very little is known regarding the stoichiometry of the proton translocating processes in the respiratory chain of H . salinarum . Fortunately , while information on the individual components are unavailable , data on the aggregate process of respiration exists . This is very important for modeling because it determines the overall energy production capability of the organism . O2 pulse experiments indicated an ATP to O2 ratio of 1∶1 . Measurements of initial proton uptake during phosphorylation demonstrated a ratio of 10∶1 between H+ and ATP [29] . These values are consistent with the experimentally determined photosynthetic stoichiometries of 22 photons per ATP and 2 photons per H+ , given that light inhibits respiration with an observed stoichiometry of 24 photons per O2 molecule [7] , [29] . We fixed the stoichiometry of the oxidative phosphorylation pathway in the model according to these values . Consumption and production rates were modeled using differential equations of the form of either Equation ( 2 ) or Equation ( 3 ) ( cf . Methods ) , with the rational that the production or consumption of a metabolite depends on the availability of the metabolite , the population size , and the current growth rate; note that growth-rate is time-varying in batch cultures . The piecewise extension of Equation ( 2 ) , Equation ( 3 ) , was necessary because several metabolites exhibited distinct modes over the growth period . For example , alanine , at some point , switched from production to consumption . We obtained parameters for each nutrient using both equation forms . To minimize the possibility of overfitting , we used the piecewise form in the final model only if it resulted in at least a reduction of 10% in residual error , which is a measure of the ( dis- ) agreement between model and data ( see methods ) . In such a case , the boundary parameter of the equation , which is optimized and automatically obtained , indicates a time near where the qualitative change in the uptake pattern occurs . Table 1 lists the best ( lowest ) residual error values we obtained using equations ( 2 ) and ( 3 ) for each nutrient ( see methods ) . From the table , it is clear that the piecewise form hardly makes any difference for a number of metabolites , including aspartate , isoleucine , and leucine . For these three amino acids , residual error values only decreased by 0 . 8% , 0 . 2% , and 1% , respectively . Given that the same three amino acids , after arginine , are also the ones with the highest uptake rates , it would seem that they are the preferred or primary metabolites of H . salinarum during aerobic growth , at least among the nutrients supplied . Unlike most of the other provided metabolites , the uptake patterns of these amino acids remained constant , and their consumption rates high , up until their depletion . The experimental nutrient utilization data and the corresponding model simulations are shown in Figure 2 . In contrast to aspartate , leucine and isoleucine , the piecewise uptake equation form is clearly superior over the basic definition ( Equation 2 ) for some metabolites . Most notable of these are ornithine , alanine and proline , for which residual errors dropped by 98 . 6% , 92 . 4% and 54 . 5% , respectively . Each of these amino acids , at some point , switches from production to consumption . For ornithine and proline , it is interesting that the points at which their respective utilization patterns change are close to each other , and that both occur near the time when arginine is depleted . This makes sense as arginine is consumed rapidly by the cells , and both ornithine and proline are downstream of its catabolic route . Flux-balance simulation at a point before arginine depletion shows the situation depicted in Figure 3 . Most of the arginine , which at the beginning exhibits the highest uptake rate , is deaminated to citrulline via the action of arginine deiminase ( EC 3 . 5 . 3 . 6; OE5208R ) . Citrulline is then converted to ornithine and carbamoyl-phosphate by ornithine carbamoyltransferase ( EC 2 . 1 . 3 . 3; OE5205R ) . Ornithine is mainly transported outside through an arginine-ornithine antiporter ( ≈95% ) , a process that accounts for most of the arginine taken up . On the other hand , carbamoyl-phosphate , via the action of carbamate kinase ( EC 2 . 7 . 2 . 2; OE5206R ) , is degraded to NH3 and CO2 in a reaction that produces ATP from ADP . In summary , most of the supplied arginine is rapidly converted to ornithine through a process that produces ATP . As will be discussed later , this fermentation process accounts for most of the energy in the cells at the early stages of growth . The uptake rate of valine accelerates at , near the point where isoleucine is depleted . Similarly , the uptake rate of methionine accelerates at , near the time when leucine is depleted . While it is hard to conclude with certainty , given the resolution and quality of the current data , that the acceleration of valine uptake actually precedes that of methionine , it is very likely that the increased consumption rates for both are the result of cells compensating for the depletion of leucine and isoleucine . Note that the two depleted branched-chain amino acids seem to be preferred metabolites of Halobacterium salinarum during aerobic growth . Accordingly , cells had to switch their metabolism to utilize more of the substitute sources in order to sustain growth . Alternative explanations are provided as supplementary information ( Text S1 ) . The other supplied metabolites with utilization patterns that we found to exhibit distinct modes are alanine , serine , threonine and glycine . The first switches from gradual production to rapid consumption , and the latter three demonstrate uptake rates that significantly accelerate . In each of these cases , the critical point seems to be near , which is also near the time when proline , methionine and valine are depleted . Although we believe that the increased consumption rates are also compensatory measures for the depleted metabolites , it is again difficult to conclude this with certainty , given the current data . Moreover , if true , it is also difficult to determine whether which nutrients serves to balance for what . Considering the significant qualitative interpretations that the parameters carry , we performed further steps aimed at analyzing the quality of the obtained values . We provide these results as supplementary information ( Text S1 ) . Nutrients consumed by cells can have various fates: ( 1 ) they can be incorporated into the biomass with at most very minimal changes , such as in the case of amino acids incorporated as protein residues or free metabolites; ( 2 ) they can be converted to other biomass components , typically after partial degradation; ( 3 ) they can be oxidatively degraded to CO2 for the production of energy through respiration; and ( 4 ) they can be secreted after conversion to another metabolite , for example following partial degradation . Given that the rates at which most of the supplied amino acids were consumed far exceeded the requirements for biomass incorporation as described in the first fate mentioned above , then at least one of the other three possible fates must also be true for the relevant amino acids . The results for individual metabolites are summarized in Figure 4 , and the comparison between total ( global ) carbon uptake and total carbon incorporation ( fates 1 and 2 ) is provided as supplementary information ( Figure S1 ) . Consistent with our previous observations [15] , leucine , isoleucine , valine and methionine are among the nutrients which are mostly degraded , whether for energy or as carbon skeleton donor . This is remarkable because the four are , along with lysine , essential amino acids for H . salinarum . The heavy rates at which they are degraded can cause the organism to prematurely terminate its own growth . One would typically assume that it would be better for the cells to use essential nutrients sparingly , especially considering the fact that there are other non-essential sources of energy and carbon available . However , leucine and isoleucine , as discussed earlier , are even among the preferred metabolites of H . salinarum , i . e . , nutrients for which the highest uptake rates were observed . All nutrients that are taken up by cells are either incorporated into the biomass ( fates 1 and 2 mentioned above ) or converted to metabolic by-products ( fates 3 and 4 ) . By approximating the biomass composition using measurements , calculations and assumptions , and then correlating these information with the observed population sizes , analysis and prediction of the latter set of fates becomes possible . We took most of the biomass composition values from our previous work [15] , which includes the contribution of individual amino acids , nucleotides , lipids , sugars , and several other metabolites . However , we repeated the determination of the amino acid content , which accounts for over 60% of the total organic mass , using measurements at various optical densities . The results are summarized in supplementary Figure S2 . While the relationships between the amino acids and the optical density are not strictly linear , average errors from linear fits are below 12 . 5% in all cases . Most of the deviations are situated in the lower optical density range . With respect to the previous set of measurements [15] , the most significant difference of the current one is that the total amino acid content is about 30% higher at approximately 503 µg/OD·ml . Possible factors that could have contributed to this discrepancy include differences in the conditions used , including media composition , and the fact we more rigorously adapted the current cultures . The discrepancy is well within the range of inaccuracy that may be introduced by morphological changes in the cells to the relationship between biomass and optical density . Accordingly , we also adjusted the biomass requirement ( composition ) of the other constituents . CO2 is an expected metabolic by-product of cells under aerobic conditions . Indeed , if it is assumed that all carbon atoms that are taken up and do not appear in the biomass are completely degraded through respiration , then CO2 should account for the bulk of the by-product pool . We calculated the ( theoretical ) amount of oxygen that will be needed under such a scenario at various points during growth , using flux balance analysis where we set energy production as the objective function and assumed an unlimited oxygen supply . Note that stoichiometric relationships between the oxidation of carbon substrates and the consumption of oxygen are implicitly defined by the reactions ( pathways ) in the metabolic network . However , comparison of these calculated results with actual oxygen consumption measurements indicated that respiratory-linked degradation is not the fate of most carbon atoms that are consumed but do not get incorporated into the biomass . In fact , the data shows that cells use only about 20% of the oxygen that they would otherwise need to completely oxidize all the material . Thus , the formation and accumulation of other by-products in significant quantities ( in the mM range ) , probably including those that result from forms of overflow metabolism , likely occur in the preparations . In addition , the described situation already seems to have been the case even before the oxygen supply became limiting at . The computations are summarized in Figure S3 ( supplemental information ) . Further flux analyses show that the most efficient by-products with respect to energy production are common intermediates such as acetate and succinate . We observed a maximal respiratory rate of approximately 1 µmol O2/OD·ml·hr during growth . At about , the amount of dissolved oxygen in the medium was reduced to 0% . Given that prior to this point , large amounts of nutrients that are consumed but neither incorporated into the biomass nor subjected to oxidative phosphorylation could already be observed , the respiratory process itself is likely the bottleneck prior to . Past this point however , cells could only respire in at most the rate at which oxygen dissolved into the medium . Note that a 0% oxygen saturation level does not mean that oxygen is no longer available to the cells because flasks were kept open . Given that growth continued well past the point when the oxygen supply started to become limiting , the respiratory rate dropped steadily from then on . Under aerobic conditions , cells can derive energy ( ATP ) through respiration and substrate level phosphorylation . We calculated the maximum ( theoretical ) energy that the system can produce as well as the fraction of this that can be attributed to respiration at various points , using the observed nutrient utilization . The values are reflected by the red and blue curves , respectively , in Figure S4 ( supplemental information ) . The system maximum energy was calculated as ( 1 ) where is the current growth rate , and is the growth-related energy that is implicitly taken into account by the metabolic network in synthesizing the compounds included in the biomass [30] . In order to find flux distributions that maximize energy production , we introduced an ATP hydrolysis reaction ( ATP+H2O = ADP+P ) into the network , and performed flux balance analysis with it defined as the objective function . in Equation ( 1 ) refers to the flux through this reaction . Throughout growth , the theoretical maximum energy that the system can produce ( red curve in Figure S4 ) is significantly higher than the energy produced by respiration ( blue curve ) . Although it is possible that the actual energy production of the system is closer to the respiratory curve , it is clear that prior to the depletion of arginine ( ) , the earlier described fermentation process for this amino acid accounts for most of the energy produced by the system . Evidently , the energy generated by this process can be comparable to the two “primary” modes ( respiration and photosynthesis ) , contrary to what is typically assumed . This is consistent with the fact that H . salinarum can be grown with neither oxygen nor light just as readily , by supplying large amounts of arginine . After the amino acid is depleted , respiration accounts for most of the produced energy . The additional energy that can be produced through the further non-respiratory-related degradation of other ( non-arginine ) nutrients is significantly smaller . This being the case , it is remarkable that non-respiratory-related processes account for approximately 80% of the nutrients that are consumed but not incorporated into the biomass . Again , we should note that the pertinent metabolites include essential ones . We computed fluxes that are consistent with the observed consumption and production rates during the exponential phase ( specifically ) and which additionally maximize energy production . These are illustrated in Figure 5 . In it , compounds which are taken up from the medium are represented by yellow ellipses , and compounds which are accumulated in the medium by red ellipses . Fluxes through reactions are presented using arrows , where thickness is used to indicate strength . Considering that it is uncertain to what extent energy production is actually the objective of cells in vivo , we used flux variability analysis [31] to complement the results by computing the minimum and maximum possible fluxes through each reaction after removing all optimality assumptions . Fluxes through reactions for which the values are either both positive or both negative are drawn in black . Intuitively , these are reactions that are constrained enough by the network stucture , pseudo-stability and the observed exchange fluxes , to have non-zero activity regardless of any optimality assumptions . That is , the network structure and the measured data are already enough to guarantee the existence of the fluxes and their indicated directions . During the exponential phase ) , we found 378 ( 67% ) of the 567 internal reactions of the metabolic network to have qualitatively invariable fluxes . Of these , 246 carry non-zero fluxes , and 132 are blocked under the conditions . The fact that measurements of a relatively few input and output fluxes already qualitatively determines the fluxes through most of the reactions is related to the bow tie structure of the metabolic network [15] , [32] . Catabolic pathways , which together form one fan of the bow tie , fan-in into the knot of central metabolism to supply the common intermediates . These pathways are highly linear and convergent in structure . Accordingly , nutrients which are taken up and not directly incorporated into the biomass will have to go through the associated catabolic pathway with no variability until the level of the common intermediates is reached . Similarly , biosynthetic pathways , which form the other fan of the bow tie , fan-out and branch from the knot of central metabolism , and are also highly linear in structure . Thus , most biosynthetic routes of cellular constituents will also be invariable once past the level of the metabolic core . Phototrophic cultures were grown under conditions similar to the aerobic preparations , except that oxygen was absent and they were illuminated with white light . We were able to observe final optical densities well above 1 . 5 OD , comparable to the population levels achieved by their counterparts ( see Figure S5 ) . The most apparent difference is with respect to having slower growth rates . While aerobic preparations exhibited exponential phase doubling times of only roughly 11 hours , phototrophic cultures demonstrated times close to 50 hours . The reason for this disparity is unclear . Representative aerobic and phototrophic growth curves are provided as supplementary information ( Figure S5 ) . In both cases , the preparation depicted is the result of three rounds of adaptation by iterative reinoculation into the respective condition . We modeled the consumption and production of nutrients during phototrophy using the same procedure we employed for the aerobic case . Accordingly , the parameter , when applicable , again reflects the point at which qualitative change in the uptake pattern of the respective nutrient occurs . The experimental data and the corresponding model simulations are provided as supplementary information ( Figure S6 ) . Analogous to the aerobic case , arginine was rapidly depleted near the start of growth , again with most of it being converted to ornithine and secreted by cells into the medium . This is interesting because arginine fermentation genes ( arcABC ) have been found to be repressed during phototrophy , presumably to repress the secondary energy source [33] . However , the rapid depletion of arginine and the equally rapid accumulation of ornithine is a clear indication that the fermentation process was in fact active despite the phototrophic condition of the cells . Considering that we used inoculants taken from cultures already growing under phototrophy , it is unlikely that the observations are only due to the time ( lag ) that is required for regulation to take effect . Next to arginine , aspartate exhibited the highest uptake rate , and is the second supplied amino acid that was depleted at . Up to that point , very little , if at all , glutamate uptake could be observed . However , consumption of the amino acid either starts or accelerates appreciably after the exogenous aspartate supply is exhausted . Similarly , methionine uptake also seems to accelerate at this point . Some 200 hours later ( ) , phenylalanine is depleted , and threonine and serine are nearly exhausted . Also at about the same time , alanine switches from production to consumption , and the uptake of glycine accelerates . The three amino acids glycine , serine and threonine are connected to each other through their degradative pathways . Presumably , when glycine is no longer produced through the degradation of excess amounts of either serine or threonine , then cells have to take up more of it . This is consistent with the observation that prior to this point , glycine uptake is significantly lower than the rate at which the amino acid is incorporated into the biomass whether as protein residues or free metabolites ( Figure S7 ) , not even considering the fact that glycine , via tetrahydrofolate , is also used as methyl donor in various biosynthetic pathways . Unlike in the aerobic case , the depletion of phenylalanine coincided with a number of discernable qualitative changes in the uptake patterns of the nutrients . In addition to the ones stated above , consumption of the amino acid serine seems to have stopped at about the same time , even though growth , albeit already in the late exponential phase , could still be observed . The reason for this is unclear . Serine was already nearly exhausted at about 0 . 2 mM by then . Moreover , it is also at this point that alanine switched from production to consumption . Serine and alanine are metabolically related to each other through their degradative pathways , as the two can enter central metabolism through pyruvate . In the aerobic case , not only were we unable to observe an arrest in serine uptake , but the switch of alanine from production to consumption actually conincided with the acceleration of serine consumption . While neither serine nor alanine share the catabolic route of phenylalanine , it is possible that phenylalanine played a more important role in the phototrophic case; that is , the observed behaviors may have been indirect effects of the amino acid's depletion . In this respect , we should note that the early biosynthetic steps of aromatic amino acids are shared with the quinones , which play critical roles in respiration . Analogous to the aerobic case and somewhat unexpectedly , the rates at which most of the supplied amino acids were taken up from the medium far exceeded the rates at which they were incorporated into the biomass with minimal modifications ( fate 1 ) . These findings are summarized in Figure S7 . Again , the implication is that the consumed quantities of the pertinent amino acids are considerably catabolized by the cells . Certainly , significant portions of these are used as building blocks ( carbon skeleton donors ) for the production of other biomass constituents ( fate 2 ) , such as nucleotides and lipid molecules . However , even the substrate requirements for the synthesis of these unsupplied biomass components are not enough to explain the differences . This can readily be seen in Figure S8 , which shows the total carbon uptake against the total carbon incorporation under the conditions . Accordingly , similar to the aerobic case , other metabolites ( by-products ) are also likely produced and accumulated in considerable quantities . This is somewhat more unexpected during phototrophy than in respiration because the necessary oxidative breakdown of nutrients in the latter is not present . Under phototrophic conditions , one would typically expect material consumption to be only near the quantity sufficient for biomass production because energy can already be derived from light , rather than from the supplied metabolites . The energy yield of the non-respiratory degradation of these materials should be very small compared to the energy produced through photosynthesis , especially considering that arginine is the only fermentative substrate of H . salinarum . Nevertheless , in retrospect , this behavior is consistent with the emerging picture that the organism takes an approach toward growth that is focused on the here and now , even at the cost of longer-term concerns . Indeed , the essential amino acids leucine , lysine , isoleucine , methionine and valine are again among the amino acids that are taken up in amounts far more than required for biomass formation . When phototrophic cultures reached 1 . 5 OD , only about 14% of the supplied carbon that had been consumed could be attributed to the amino acid content of the biomass ( directly measured; see Figure S8 ) . This means that the total carbon incoporation rate was only about 20% ( accounting for the other biomass components: eg . , nucleotides , lipids , etc… ) . In comparison , when aerobic cultures reached comparable levels of population ( also in the late exponential phase ) , we could attribute about 13% of the consumed carbon to the amino acid content of the biomass ( directly measured; see Figure S1 ) , which implies a total carbon incorporation rate of approximately 19% . This means that about 80% and 81% of the consumed carbon were not incorporated in the phototrophic and aerobic cases , respectively . Correcting the aerobic value to account for the respiratory-related oxidative degradation of nutrients—using a 1∶1 CO2 to O2 ratio which was calculated using the stoichiometry of the parts of the metabolic network that are involved in the oxidation of the relevant amino acids—shows that approximately 66% of the carbon was neither incorporated nor respired . Considering the fact that the growth rates observed for phototrophy were nearly five-fold slower , it is remarkable that the ratios of the nonincorporated , nonrespired carbon under both conditions are as close to each other as they are . Indeed , aerobic cultures already reached 1 . 5 OD after only slightly more than 80 hours , while , in stark contrast , phototrophic cultures took about 450 hours to reach similar levels . It would therefore seem that a large part of the nutrients that were taken up but were neither incorporated into the biomass nor used as respiratory substrates has more to do with growth ( biomass production ) than with any form of maintenance . For this reason , it is likely that some as of yet uncharacterized , non-maintenance growth processes are at least partially responsible for the low carbon incorporation rates . Several unexpected findings were made during the course of this study . These include: ( 1 ) that essential amino acids are degraded not only during respiration but even under phototrophic ( anaerobic ) conditions , where energy should already be abundant; ( 2 ) that fermentation of arginine , which is often considered a secondary , alternative energy source , occurs simultaneously with either respiration or photosynthesis; ( 3 ) that considerable amounts of metabolites are produced and accumulate in the medium under both conditions , as a result of nutrients that are consumed but not incorporated into the biomass nor used as respiratory substrates; and ( 4 ) in connection with the previous point , that the total carbon incorporation rate is extremely low even under phototrophic ( anaerobic ) conditions ( approximately 20% ) , when one would typically expect nutrient consumption to be in quantities that are only sufficient for biomass production since energy is already derived from light . All of these findings are consistent with the seemingly “greedy” behavior demonstrated by Halobacterium salinarum that we noted in our previous study [15] , which we believe actually consists of adaptations to its natural environment , where nutrient availability is not only irregular but can also be absent for extended periods of time . In the salt lakes and solar salterns where Halobacterium salinarum may be found , life is characterized by blooms that may not occur for years after a previous episode . In the Dead Sea for example , no growth of the unicellular green alga , Dunaliella parva , which is responsible for all of the primary productivity , is possible when the salt concentration of the water column is invariably high . Blooms of Dunaliella in the Dead Sea occur only after significant dilution of the upper levels by the influx of freshwater . In turn , this allows the archaeal community to bloom at the expense of the organic material produced by the alga [34] . The fact that the conditions under which these blooms are possible may not be realized for years after a previous episode [35] , [36] may have inclined Halobacterium salinarum to grow as much as possible when the conditions become favorable . In connection to this , as for the times when their environments are not conducive to growth , the capacity of the halobacteria to survive in adverse conditions for extended periods has been well established [37] , [38] . This capacity allows them to survive until the next bloom . In this respect , we should note that a considerable number of the reported viable ancient cells that have been recovered are at least moderately halophilic ( halotolerant ) [39] , [40] , [41] . One of the predictions of our model is that non-CO2 by-products/metabolites accumulate in the medium in significant amounts under both aerobic and photorophic conditions , likely including partially degraded forms of the supplied nutrients . Unfortunately , attempts to identify the secreted molecules are complicated by the ( necessary ) extremely high salt concentrations ( 4 M ) in the growth media . We intend to focus on the identification of these metabolites and on the elucidation of the reasons/dynamics behind their accumulation in a succeeding study . Specific sets of nutrients dominate material uptake at various points during growth . For example , at the early stages , highest uptake rates were observed for aspartate and leucine , and , to a lesser extent , glutamate and serine . Note that arginine uptake was the highest , but most of its carbon skeleton was secreted as ornithine . With this information in hand , we would like to perform flux analysis using labeled substrates [22] in order to obtain quantitative constraints for at least a fraction of the internal fluxes . This should allow for a more detailed view of the metabolic strategy used by the cells .
For the consumption or production of each nutrient , we used a simple model consisting of three terms: ( 2 ) where is the uptake rate of nutrient , is the population size at time , is the current growth rate defined as , and , and are optimizable parameters . The rational of the construction is that the production or consumption of a metabolite depends on the availability of the metabolite , the population size and the current growth rate . Preliminary inspection of the data revealed that the transport patterns of some of the metabolites change during growth . For example , ornithine , at some point , switches from being produced and accumulated in the medium to being steadily consumed by cells . Accordingly , we used a simple extension of Equation ( 2 ) that allows for two modes: ( 3 ) where , , , , , and are optimizable parameters . Similar to the basic form , the consumption or production of each metabolite still depends on the availability of the metabolite , the population size and the current growth rate . However , the new parameter now separates time into two intervals , and the distinct set of parameters for each of these allows the utilization pattern to qualitatively change in moving from the first to the second . For each nutrient , parameters for both Equations ( 2 ) and ( 3 ) were estimated from experimental data using the fminsearch function of MATLAB . Systematically defined sets of initial parameter values were used in solving the inverse problems ( parameter estimation ) . To minimize the possibility of overfitting , the more sophisticated form of Equation ( 3 ) was used in the final model , in favor over the simpler equation form , only if it resulted in significant improvement , that is , the residual error , a measure of the ( dis- ) agreement between the model and data , is substantially lower . In this respect , it is reasuring that although Equation ( 3 ) is clearly better for some metabolites , such as alanine and ornithine , it hardly makes any improvement for others , such as leucine and isoleucine , which seem to be staples of the cells ( see results ) . In cases where Equation ( 3 ) is used , the optimized parameter , intuitively , corresponds to a point near where the uptake pattern changes qualitatively . Note that the borders ( ) need not be equal for all Xi's . Because the choice between the two equation forms can be done automatically , the computational system we employ not only accounts for changes in metabolite modes , but can actually lead to the detection and identification of these modes , and , subsequently , to the recognition of the biological processes behind them . A metabolic network can be conveniently represented as a stoichiometric matrix S , where each row corresponds to a metabolite and each column to a reaction . The entries of S are the stoichiometric coefficients that define the relationships between the reactions and compounds . A positive value for indicates that compound is produced in the left to right direction of reaction , while a negative value indicates that it is consumed . For a particular time interval during growth with length , a set of fluxes which are consistent with the observed nutrient depletion/accumulation and biomass formation can be obtained by solving the linear program ( 4 ) where is a vector of fluxes defining the flux through each reaction , is the objective function , is the set of reversible internal reactions , is the set of irreversible internal reactions , is the set of exchange fluxes associated with ubiquitous metabolites , and is the set of exchange fluxes that correspond to experimentally measured nutrients . The set of ubiquitous compounds include CO2 , H2O , Na+ , Cl− and H+ . For each reaction , the value is the appropriate evaluation of the corresponding uptake equation ( 2 or 3 ) for the interval . In this work , all of the supplied carbon and energy sources , except for citrate , were included in . In addition , the set also includes ornithine , which is initially accumulated in the medium , and oxygen . Biomass production is treated in a similar manner by fixing the flux of the growth ( pseudo- ) reaction to observed rates . To account for the possibility that cells produce and accumulate in the medium certain metabolites , for example in the case of overflow metabolism , the set of one-way exchange reactions , , was included . It includes central metabolites such as acetate , pyruvate , malate and GAP , nucleotides , and some sugars such as glucose . Unless otherwise specified , energy production was used as the objective function . Strain Halobacterium salinarum R1 ( DSM 671 ) cells were grown in chemically-defined medium , with composition defined in Table S1 . The medium composition that we used is the simplest known for the halophile that allows population levels comparable to those reached when using complex media . Preparatory cultures were grown in 100 ml flasks containing 35 ml of the medium to a cell density of ≈1 OD , from which 1 ml inoculants were taken to start the next culture . This was done repeatedly to adapt cells to their growth conditions . All cultures were prepared in flasks which had side arms to measure turbidity ( cell density ) via a Klett photometer , and were carried out in duplicates . Cell suspensions were shaken at 105 rpm at 40°C in the dark . At specific points , samples were taken from the cultures so that 14–18 samples were collected over the growth period , and these were stored at 4°C . To separate the cells from the medium , the samples were centrifuged for five minutes at 15 , 000 rpm , using a SS34 rotor . Pellets were resuspended in 500 l basal salt ( medium without the amino acids ) and spun down as before . Amino acid analysis was performed on both the pellets and the original supernatants , using an Amino Acid Analyzer ( Biotronik LC3000 ) . Oxygen saturation in the medium was monitored using the “Fibox 3-trace v3 , fiber-optic oxygen meter” from Precision Sensing GmbH ( Regensburg , Germany ) . Calculation of the actual oxygen consumption rates is provided as supplementary information ( Text S2 ) . Similar protocols were used for phototrophically grown ( anaerobic ) cells , except that flasks were first flushed with nitrogen to remove oxygen , and were closed using air-tight septa . To maintain the anaerobic condition of cultures while taking samples , syringes with a long needle , inserted through the septa , were used .
|
Living cells can produce usable energy through various means . For example , animals derive energy , through respiration , from nutrients that they consume , and plants from light using photosynthesis . The particular microorganism that we study , Halobacterium salinarum , is a model organism for the archaeal domain of life . It is bioenergetically flexible in that it can perform both respiration and photosynthesis and in addition can also derive energy using fermentation . Accordingly , it is a good model system for investigating the interplay between different energy generating mechanisms . In this study , we investigate these relationships as well as how energy production is linked to the other processes involved in growth , including the consumption of nutrients and the production of cellular material . Because Halobacterium salinarum thrives in salt-saturated solutions , such as those that may be found in salt lakes and solar salterns , our study yields insight on how these cellular processes operate in environments that are lethal to most life on Earth .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"cell",
"biology/microbial",
"physiology",
"and",
"metabolism",
"biochemistry/bioinformatics",
"computational",
"biology/metabolic",
"networks",
"computational",
"biology/systems",
"biology"
] |
2009
|
Systems Analysis of Bioenergetics and Growth of the Extreme Halophile
Halobacterium salinarum
|
One of network epidemiology's central assumptions is that the contact structure over which infectious diseases propagate can be represented as a static network . However , contacts are highly dynamic , changing at many time scales . In this paper , we investigate conceptually simple methods to construct static graphs for network epidemiology from temporal contact data . We evaluate these methods on empirical and synthetic model data . For almost all our cases , the network representation that captures most relevant information is a so-called exponential-threshold network . In these , each contact contributes with a weight decreasing exponentially with time , and there is an edge between a pair of vertices if the weight between them exceeds a threshold . Networks of aggregated contacts over an optimally chosen time window perform almost as good as the exponential-threshold networks . On the other hand , networks of accumulated contacts over the entire sampling time , and networks of concurrent partnerships , perform worse . We discuss these observations in the context of the temporal and topological structure of the data sets .
In the 1980's and 90's , mathematical epidemiology of infectious diseases made great progress . During these years , researchers went from models where every individual meets everyone else with equal probability , to a framework of networks where people are considered as connected if one can infect the other . This new body of theories , network epidemiology [1]–[3] , has altered our understanding of disease spreading profoundly . For example , it has changed the concept of epidemic thresholds , outbreak diversity and the role of social networks in intervening infectious disease outbreaks . Furthermore , research in network epidemiology has produced many new techniques to analyze contact data [3] , [4] , model disease spreading [3] , [5] , [6] , discovering influential spreaders [7]–[9] , detecting outbreaks [10]–[12] , etc . Network epidemiology rests on coarse simplifications , perhaps the biggest being that that one usually does not explicitly model the dynamic aspects of contact patterns [13] . If we consider two individuals , and assume one of them is infective ( infected by a pathogen and able to spread it further ) , the probability of contagion between the two is in practice not constant in time . The changes in the chance of contagion happens at different time scales—from the order of decades , as people are born and die , to the order of minutes , as they come in and out of range for pathogen transmission . The situation becomes even more complicated if we consider an emerging disease outbreak . First , if we want to apply network epidemiology to control the spreading , we have to predict the future contacts , not just map out the past [9] . Second , the mechanisms behind how and when people make contacts may change from the fact that the people are aware of the epidemics [14] , [15] . However , the theory and methods to handle full contact patterns—including both temporal and topological information—is not at all as developed as static network epidemiology . There are for example works presenting mathematical representations for temporal networks and computational studies of the effects of temporal and topological structures ( see the survey papers Refs . [16]–[18] and references therein ) , but the picture they present is far from as complete as the static network epidemiology . For this reason is static networks still an important mathematical modeling framework for epidemiology . To be able to study contact patterns by static methods , one needs to eliminate the temporal dimension . This can be done in many ways , and this paper aims at finding the optimal way . Consider a sequence of contacts—triples ( i , j , t ) carrying the information about when ( assuming a discrete time t ) pairs of individuals ( i and j ) have been in contact . A good static network representation of such a contact sequence should give the same predictions about the disease dynamics as the contact sequence itself . The predictions we focus on in this paper are related to how influential an individual is in the disease spreading . Assuming a disease is introduced in a population by individual i , we compare the expected outbreak size ∑i in a simulation on the contact sequence with a static-network predictors of importance: i's degree ki . The better the performance of the static network predictors , the better is the network representation . The reason we focus on predicting the importance of individuals rather than e . g . the epidemic threshold is that it is less dependent on the parameters of the disease-spreading model . If , for a given set of parameter values , one can predict the total outbreak size but not the important disease spreaders , that is more likely a coincidence than if one can predict the important disease spreaders but not the outbreak size . This is important since , as our main focus is to scan different types of network representations , we will have to limit ourselves to a few parameter values of the disease spreading simulations . In this paper , we will use both empirical and artificially generated temporal-network data sets . We investigate three classes of network representations to find which one that can predict ∑i from ki or ci the best . Furthermore , we investigate how the performance depends on the temporal and topological structures of the data .
We will compare three conceptually simple methods of reducing a contact sequence to a static network ( illustrated in Fig . 1 ) . In the first method , time-slice networks [19] , [20] , an edge means that two vertices have been in contact within an interval [tstart , tstop] . The second representation , ongoing networks [21] ( short for “networks of ongoing contacts” ) , adds edges between pairs of vertices with contacts both before and after the interval [tstart , tstop] . This is thus a network of edges , or relationships that are concurrently active over the time window . This method takes its name from literature of sexually transmitted infections where it is believed that the level of concurrent partnerships is a key-factor to understand how contact patterns influence epidemics [22] , [23] . The third method is exponential-threshold networks . In these , every contact contributes with a weight to the pairs of vertices—decaying exponentially e−t/τ with the time t of the contact since the first contact of the data and τ is a parameter value setting the relative weight of newer and older contacts . Then this network of weights is reduced to a simple graph by including edges for all weighted edges above a certain threshold Ω . All three methods have two control parameter each—the first two methods having the endpoints of an interval as parameters , the last one has a decay parameter τ and a threshold Ω . Note that other common network representations are limit cases of one of these methods . E . g . , a network of accumulated contacts is the same as a time-slice network over the entire sampling time . An aggregated threshold network ( where an edge represent all pairs of vertices with at least Ω contacts ) is the τ→∞ limit of the exponential-threshold networks . A more detailed description of the network representations can be found in the Methods section . As mentioned , we evaluate the network representations by comparing the importance ( ∑i ) of individuals for the disease spreading in a temporal network and the assessed importance ( ki or ci ) from the derived static network . ∑i comes from a Susceptible–Infective–Removed ( SIR ) simulation ( for details , see the Methods section ) . The stronger the correlation between ∑i , on one hand , and ki or ci on the other , the better is the network representation . As it turns out ∑i and ki , or ci , typically have strongly nonlinear relationships , which makes the Pearson correlation coefficient less related to the actual predictability of the data ( see Supporting Information , Fig . S1 ) . For this reason , and that Kendall's tau is prohibitively slow to compute in our case , we focus on the Spearman rank correlation . ( We test the Pearson and Kendall coefficient for some of the data and find that all three coefficients rank the methods in the same order and are optimized for about the same parameter values . ) As a start , we will analyze empirical contact sequences of the type outlined above ( lists of potentially contagious contacts—who has been in contact with whom at what time ) . These empirical data sets are more or less related to disease spreading; but they all serve as examples of different temporal-network structures . The data sets fall into three categories—online communication , face-to-face and sexual encounters . The latter two categories are of course more interesting for the spread of infectious diseases ( while the former perhaps could be interesting for the spread of e-mail viruses ) . Of online communication data , we study two e-mail networks—from Refs . [24] ( E-mail 1 ) and [25] ( E-mail 2 ) . In these data sets , a sent e-mail represents a contact . Even though an e-mail is naturally directed , to analyze all the data in the same way , we treat it as undirected . The two e-mail data sets are sampled from a group of e-mail accounts . One difference between them is that the data of Ref . [24] includes contacts to external e-mail accounts while the data of Ref . [25] only records e-mails between the sampled accounts . One method is probably not better than the other . To avoid these boundary effects , one can study communication within a closed community . We do this with data from an Internet dating community first presented in Ref . [26] ( Dating ) . The face-to-face data sets are gathered by radio-frequency identification sensors worn by the participants of a conference [27] ( Conference ) and visitors of a gallery [28] ( Gallery ) . In these data sets , a contact is recorded , at 20 seconds intervals , if two individuals are within range ( 1–1 . 5 m ) . Finally , we use a data set of sexual encounters gathered from a web forum where sex-buyers evaluate escorts [29] ( Prostitution ) . We list some basic statistics of the data sets in Table 1 . Turning to the main results of this section , we display the performance of the network representations in Table 2 . The most discernable result is that the exponential-threshold networks have the highest score for all data sets and importance measures except one case ( the Conference data ) . Indeed , the Spearman ρ-values are all relatively high , meaning that important spreaders are highly predictable from just the contact patterns ( although not possible to forecast , as this is a post hoc analysis ) . This suggests that the exponential-threshold representation is a good general way of constructing networks for network epidemiology ( which we will argue for more below ) . Another observation is that the aggregate networks , the most common static network representation of temporal network data , perform very poorly ( ranging from 51%–91% of the maximal correlation value ) . The ongoing networks perform very differently for different data sets—sometimes ( E-mail 2 ) they are close to the best , sometimes remarkably bad . We note that the ongoing network representation is typically optimized for tstart = tstop ( the special case studied in Ref . [21] ) . I . e . , longer concurrent partnerships ( the set of contacts between a pair of vertices ) does not predict disease spreaders better that the mere fact that they are concurrent . The occasional poor performance of the ongoing networks is a bit surprising in the light of the reported importance of concurrent partnerships for disease spreading in sexual networks [22] , [23] . An explanation could be that these studies concern population averages rather than the relative importance of individuals . The time-slice networks are performing consistently well—in one case better , and in the other cases close to the exponential-threshold networks ( on average ρ≈0 . 09 lower ) . They have most relevant information if the time interval begins early . Indeed , the optimizing starting time is almost always the same as the beginning of the epidemics . This means they are also in practice , like the exponential-threshold networks , weighing the interactions with a weight decreasing with time ( only that this weight function is discontinuous ) . The relative duration of the optimal time slice varies considerably ( from 10% to 77% of the entire sampling time ) . Ref . [20] points out that time-slice networks of phone communications are most complex for intermediate time windows; perhaps our optimal time-slice networks coincide with this region . We now take a deeper look at the regions of optimal parameter values for the three classes of network representations . If one wants a quick analysis without the optimization procedure of this paper , then how can one set the parameters ? Are there rules of thumb ? We use the Prostitution data as an example in Fig . 2 . The other data sets behave qualitatively similar ( with one exception , mentioned below ) . The window of the optimal time-slice networks starts , with few exceptions , at the same time as the first contact . In other words , the initial contacts of the seed and its surroundings are so important that the other early contacts ( between vertices that are out of reach of the infection at that stage , and thus with a potentially negative contribution to the correlation coefficient ) do not matter . However , the end of the interval is harder to guess . Presumably , this value should be of the order of the peak of the outbreak . After the peak , the contacts should have less influence on the disease evolution and thus on ∑i . Since simulated disease should spread fast in dense data sets like E-mail 2 and Conference ( with about 36 and 184 contacts per individual on average , respectively ) , it is natural that these data sets show low tstop-values ( relative both to the sampling time and the mean interevent time ) . Nevertheless , we still do not know how to estimate this value without running disease simulations . The good news is that the network representation is rather insensitive to the choice of tstop . The ongoing networks typically are maximized at tstart≈tstop for some intermediate value smaller than the duration . Also here , it is hard to give an estimate of this parameter value , more than it happens within the optimal time window of the time-slice data . The last method , the exponential-threshold networks , is frequently optimized along a curve τ∼eΩ/Ω′ , where Ω′ is a constant . This is because larger decay factors give larger weights and thus larger thresholds . The Conference data , however , is optimized for values close to the lower limit of decay exponent ( which is linearly increasing with the threshold value ) . Our conclusion is that no matter which one of our three representations one use , one typically needs to optimize one , but not two , of its parameters—tstop for time-slice networks , tstart ( = tstop ) for ongoing networks and Ω′ for the exponential-threshold networks . The fact that different methods works better for different data set and that the important disease spreaders are harder to predict in some data than others , of course , comes from differences in the temporal network structure . In Table 1 , we list values of some structural measures . We see many similarities between the data sets , perhaps because they are all social networks ( in the sense that the vertices represent individuals ) . All data sets have skewed and broad degree distributions ( not shown ) and they all have bursty contact patterns between along the edges . We will take these observations as guidance when we test our network representations on synthetic data below . Next , we turn to studying the network structure of the optimized networks of the three types of network representations . The results are shown in Tables 3–5 . We include numbers for the accumulated networks for comparison ( Table 6 ) . First , a little side remark—we note that these accumulated networks differ much in structure . The Gallery data have longer distances ( which is natural because the visitor at the gallery are connected to visitors around the time , so it becomes stretched out in time ) . The Conference data , we note , has a peculiarly high average degree . The average sizes of the largest connected component are typically larger in the empirical network than in the null models ( again Gallery being an exception ) . The average distances , d , in the networks are sometimes smaller and sometimes larger than the null-model networks . For networks embedded in space , like the Prostitution data , one can expect the empirical network to have larger d-values than in the null model . This is indeed the case as can be seen in Tables 3–6 . Compared to the network of accumulated contacts , the time-slice and exponential-threshold networks have fewer ( non-zero degree ) vertices and edges . However , the difference is never larger than 50% . Furthermore , these networks have a size of the largest connected component being close to unity . This is good if we would like to use the derived networks for other types of network epidemiological studies . If the networks would be disconnected , an epidemic simulation could miss possible system-wide outbreaks . Many of these conclusions do not hold for the ongoing networks . These networks are often much smaller—e . g . in the Dating data , the numbers of vertices and edges are 761 and 548 respectively , in contrast to the network of accumulated contacts with 22 , 287 vertices and 78 , 608 edges . ( At the same time , it is not the case that more edges necessarily are better , as we know from the performance of the accumulated networks . ) This sheds a new light on the poor performance of the ongoing network representation in the empirical data sets—there is a too low fraction of agents with concurrent partnerships for these to be efficient . However , for the E-mail 2 data , the performance is actually even better in the ongoing than the time-slice networks . This is also the data set with the largest fraction of concurrent relationships . Therefore , in our case , even if concurrent partnerships increases the importance in disease spreading , they are less significant than accumulated serial contacts ( as captured by the time-slice networks ) . In sum , both the time-slice and exponential-threshold models do not change the structure of the network ( compared to the networks of accumulated edges ) in any systematic way , but the ongoing networks do . Now we will explore effects of the temporal-network structure and the stability of the above observations in a model network . It would be quite impossible to scan all facets of temporal-network structure . Rather , we will focus on the effect of overlapping relationships on the performance of the representations . Can it be the case that they are outperforming the time-slice and exponential-threshold networks for some temporal-networks with a high degree of overlapping relationships ? We set up the simulation so as to mimic as much of the observed structure as possible , while simultaneously controlling the average fraction of concurrent partnerships . The latter is achieved through a parameter , μ ∈ ( 0 , 1] , where larger values mean more relationships that are concurrent . An outline of the construction algorithm is shown in Fig . 3; for more details about the simulation , see the Methods section . In Fig . 4 , we plot the performance ( same as before—the maximum of the Spearman rank correlations between ∑i in SIR simulations and the degree of the respective static network ) as a function of μ . As expected , the ongoing networks works better for larger values of μ , but they are never able to catch up with the time-slice and exponential-threshold networks . The difference between the latter two representations is—just like for the empirical networks—small , but with an edge to the exponential-threshold networks . For the largest value of μ , the time-slice networks perform slightly better ( but the values are within one standard deviation from one another ) . When μ = 1 , in the limit of many contacts per edge , the ongoing and time-sliced networks will be the same ( simply equaling the network of aggregated contacts ) . The difference , seen in Fig . 3 , is because we have on average just ten contacts per edge . To explore the difference in topology a bit further , we plot the number of vertices of degree larger than zero and average degree in Fig . 5A and B . As expected , when μ is large , these two quantities are quite similar for all network representations . For a lower fraction of concurrent partnerships , however , both the size and the average degree are considerably smaller for the ongoing networks . Like the empirical networks , it seems that the ongoing network representation is too restrictive in its edge definition . Another phenomenon observed in the empirical data that is also reproduced by the synthetic data , is that the networks have larger sizes of the largest connected components than be expected from a randomized null-model ( see Fig . 5C and D ) . This means that the optimized networks have a bias for being more connected . Probably , this reflects that the performance measure relates the local network structure to the outbreak size . I . e . , by constructing network whose local properties ( degree ) encode a global dynamic property ( outbreak size ) of the original data , one also affects a global topological property ( size of the largest connected component ) of the constructed network . The optimizing parameter values are presented in the Supporting Information , Figs . S2 , S3 , S4 .
We have explored how to encode as much information from a temporal network and a known start time of an infection into static graphs so that a predictor of disease-spreading importance—degree—is as accurate as possible . The main conclusions are that , on one hand , exponential-threshold networks generally perform best; on the other hand , time-slice networks often perform almost as good . Our general recommendation is thus to use exponential-threshold networks if possible . However , the simplicity in constructing and optimizing a time-slice network makes it a feasible alternative . To straightforwardly use a network of accumulated contacts is not a good idea—for some data sets , the performance is less than 60% of the maximum . In addition , the ongoing networks—recording contacts that are active simultaneously—perform rather poor . The performance is better when there are relatively many concurrent edges , or partnerships , ( i . e . when these networks are rather dense ) , but never as good as the other two methods . It is well established that the overall level of concurrent partnerships increases the frequency of population-wide outbreaks [22] , [23] , but it seems like , at least in our data sets , the non-concurrent contacts are necessary for determining the importance of individuals in the spreading process . How much do our results generalize beyond our current analysis ? There are of course many other ways to evaluate the performance of network representations . Instead of the performance measure that we consider ( the ability of a vertex’ degree to predict its rank in a list of estimated sizes of outbreaks originating at that particular vertex ) , one can imagine other measures . Different types of centrality measures [30]–[32] are candidates for such measures , but these are often global quantities . In practical applications , it is hard to assess quantities other than local—cf . it is easy to check one's degree in an online social network as Facebook , but much harder to know one's value of a global property . Moreover , in many empirical networks , centrality measures ( including degree ) are strongly correlated in empirical data [33] , so we expect our conclusions to remain if we change the static-network importance estimators . In the Supporting Information Table S1 , we show that our conclusions are the same if we use a global predictor of importance—coreness [7] . Furthermore , the importance measure for the dynamic simulation can be chosen differently . It measures expected outbreak size if the outbreak starts at the focal vertex . In general , another factor affecting the importance is the chance to acquire the infection . Ideally , an importance measure should weigh together both these aspects . In most types of data , we expect these aspects to be strongly correlated and we settle for the mentioned expected outbreak size . One can also think of other prediction tasks for the comparison than finding influential spreaders—for example , predicting epidemic threshold , peak-time of the epidemics , prevalence as a function of time or the final outbreak size . Such studies would require us to study a specific disease-spreading model for the static network . This added complication is the main reason that we avoid such a direction . However , we also believe ( as mentioned above ) , that predicting influential spreaders is a comparatively easy task . If one cannot say who would be an influential spreader , but still get the epidemic threshold right , the latter seems rather like luck . ( Investigating this hypothesis rigorously would be an interesting future direction . ) How much do our conclusions depend on the disease simulation model and its parameter values ? The per-contact transmission probability probably does not affect the ranking of the vertices ( even if the expected outbreak sizes can vary non-linearly ) . The duration of the infective state , however , could change the ranking . In Ref . [34] , the authors find a threshold-like response to changing δ in the Prostitution data , albeit softer than in the other parameter—the transmission probability . If the duration is longer , then we anticipate contacts over a longer time-span to matter . The network representations should of course be adapted to such a change , in the sense that their optimal parameter values would change . It is hard to see why this would change the ranking of the representations , and a preliminary study ( investigating the Prostitution data for other δ-values ) shows it does not . Other studies [9] , [35] also find that qualitative results , like the ranking of influential individuals , are robust to the choice of compartmental model and parameter values . Maybe the most serious reason to be cautious about generalizing our results is that we have investigated only a limited set of temporal-network structures . Indeed one can imagine numerous types of correlations between temporal structure and network position—correlations between edges connected to the same vertex , between vertices connected by an edge , etc . A promising sign , however , is that the empirical data sets span a rather large range of static network structure ( both in terms of the network of accumulated contacts and the optimized networks ) . In the end , it is probably impossible to scan all temporal-network structures . Rather , we hope for higher quality empirical data . This would also allow us to better tailor the network representations to specific pathogens . The ρmax-values—between 0 . 68 ( for the synthetic data ) to 0 . 93 ( for the E-mail 2 data ) —are , on one hand , rather high ( we could predict important spreaders with a quite high confidence ) . On the other hand , since many goals of network epidemiology ( predicting thresholds , etc . , as listed above ) are more dependent on the details of the contact structure and thus more difficult , we can appreciate the value of having the full , temporal contact patterns . The conclusion from this is to , as long as possible , avoid reducing contact data to static networks [6] , [9] , [35]–[40] . An interesting question for the future is why some data sets give higher performance values . With the degree sequences of the accumulated networks ρmax is bounded above by about 0 . 95–0 . 98 ( 1 is unattainable because of the degeneracy of degrees ) . The discrepancy comes from the network-construction methods being too blunt to capture the relevant temporal-network structure . On the other side , it may be too much to ask from the method to rank the bulk of peripheral vertices accurately—the difference between them will probably be smaller than the errors in the raw data set . Another open future direction is to design other network representations , perhaps putting different weight depending on burstiness [41] or other temporal traits . It could also be interesting to generalize the study to general spreading phenomena in social networks . This paper finds results also generalize to datasets not directly relevant for infectious disease spreading ( the e-mail data ) , perhaps reflecting strong general patterns in human activity [40] , [41] .
We consider a set C of L contacts among N vertices . T is the total sampling time . We count time ( usually denoted t ) from the data set's first contact . E is the set of vertex pairs with at least one contact . In the context of concurrency , we also call edges “partnerships” to conform to the terminology of the theory of sexually transmitted infections . By construction of our data sets , all the vertices will be part of at least one contact . We denote the number of elements in E by M . When we discuss the constructed networks , we use N and M to represent the number of vertices and edges , respectively , in that particular network . We simulate disease spreading by a version of the SIR model defined as follows . Start the simulation from a situation where all vertices are susceptible . The outbreak is then initiated from a seed i at the time of i's first contact . Then , at every contact involving one infective and one susceptible , we make the susceptible infective with a probability λ . An individual stays infective for a duration δ , whereupon it becomes removed . ( This is different from the differential equation formulations of the SIR model that assumes that infective vertices become removed at a fixed rate—i . e . with an exponentially distributed duration—which is neither realistic [42] nor parsimonious in an individual-based simulation like ours . ) We go through the contacts in time order . If more than one contact occurs at a time unit , we sample them in random order . For every vertex as seed , we run the simulation between 1 , 000 and 10 , 000 times . Ideally , we should scan the entire ( λ , δ ) parameter space , but this would be computationally too demanding . Rather , we will try to simulate the disease spreading where it is easy to separate the more from the less important individuals . This happens at intermediate λ- and δ-values . ( For an infinite system , it would be around the epidemic threshold , but for the finite systems that we consider , thresholds are ill defined , so we avoid that terminology . ) As a simple principle , we chose δ as one fifth of the sampling time and λ such that the average outbreak size becomes one fifth of the size with λ = 1 and δ = T/5 . The actual values that we use can be found in Table 1 . We limit ourselves to simple graphs ( unweighted and undirected graphs that have no multiple edges or self-edges ) and require that their construction should be conceptually simple . The simplest type of such representations is the time-slice network—an edge in these is any pair of vertices ( i , j ) that have one or more contacts ( i , j , t ) with tstart≤t≤tstop [19] , [20] . If tstart and tstop are the beginning and end of the data set , then we speak of an aggregated network ( which probably is the most common representation when running disease simulations on empirical network data [3] ) . The second network representation that we consider is ongoing networks . Here an edge represents a pair of vertices ( i , j ) that have at least two contacts ( i , j , t ) and ( i , j , t′ ) where t<tstart≤tstop<t′ . Ref . [21] studied these in the special case tstart = tstop . The last type of network representation that we test is exponential-threshold networks . In these , each pair of vertices is assigned a weight ( 1 ) and if this weight exceeds a threshold Ω , then ( i , j ) is considered an edge . The motivation for this type of network is that contacts that are further from the introduction of the disease ( which in our case happens early in the sampling period ) should be less important . The time-slice networks can also include edges by a decaying function of time , only that the function is discontinuous . The exponential weight is a smoother way to account for this decrease of importance ( so that many later contacts can equal a few recent contacts ) . In our tables discussing the structure of the data sets and derived networks , we use a number of quantities that we will define here . To quantify the tendency of contacts to be temporally separated by broadly distributed intervals , we use the burstiness measure of Ref . [41] . For all pairs of vertices i and j with two or more contacts , we collect the times between contacts to one long series of interevent times . Then the burstiness B of the data set is the coefficient of variation of this series . Another important quantity is the relative size of the largest connected component ( i . e . a subgraph where every vertex is reachable by following a sequence of adjacent edges ) . We measure it as a fraction S of the total number of vertices in the graph . Note that this , when applied to the optimized networks , is not necessarily the same as the total number of individuals in the original data set . Finally , we measure the distances in the largest connected component , d—the number of edges in the shortest path between two vertices in the largest connected component , averaged over all its pairs of vertices . We compare the static network measures by the corresponding values from a randomized null model with the same set of degrees but otherwise no structure . An instance of this model is generated by: sequentially going through all edges ( i , j ) , pick a random new edge ( i′ , j′ ) , replace these two edges by ( i , j′ ) and ( i′ , j ) , or ( with the same probability ) by ( i , i′ ) and ( j , j′ ) . If the replacement step would introduce a multiple or self-edge another edge , then a new ( i′ , j′ ) is chosen randomly . We estimate importance of a vertex i in the SIR simulation as the average outbreak size ∑i over 1 , 000 ( Gallery , Conference and Prostitution ) or 10 , 000 independent simulation runs if the disease is introduced by the vertex at its first contact . To estimate the importance of a vertex in the disease spreading from the static networks , we degree ki—the number of neighbors of a vertex . This is a useful quantity for its simplicity . It is local , meaning that every individual should be able to estimate its own value ( in practice this could of course be difficult , depending of the mode of transmission of the pathogen ) . Degree , as a measure of influence , is also intuitive—meeting more individuals should increase both the chance of getting a disease and the number of others one can spread the infection to . In the Supporting information , we also present results for another measure of influence—the so-called coreness , roughly measuring how large the well-connected neighborhood of a vertex is [7] . The method to generate synthetic contact sequences is outlined in Fig . 3 . Here we describe the process in greater detail . We start by constructing a ( static ) simple graph , N = 5 , 000 , by the configuration model [30] . This means that we assign one discrete random number for every vertex i from some probability distribution . These numbers represent “stubs” or “half-edges” desired to be a part of an edge . Then we choose stubs of random pairs of vertices i and j and attach them to form an edge ( i , j ) provided that no such edge already existed and i≠j . This adding of edges continues until there is no stub that is not a part of an edge . One caveat , however , is that a complete matching may not be possible ( if , for example , one vertex has two stubs left and the others all zero ) . To handle this , if the matching is unsuccessful for 104 consecutive times , we give up and delete the remaining stubs . In this paper , we use a truncated power-law distribution to mimic the skewed , broad degree distributions of the empirical networks . To be specific , we draw the random numbers from a distribution ( 2 ) where , in our work , kmin = 1 , kmax = N−1 and γ = 2 . 2 . This gives , on average , M = 10 , 595 ( 5 ) ( the last number being the standard error in order of the last decimal ) . After the network topology is generated , we proceed to assign times of contacts to the edges . We assume a contact over an edge can only take place during an activity interval of duration , μT . We recognize that the activity intervals would more accurately be modeled as skewedly distributed . However , if we choose the intervals as broadly distributed as e . g . the degrees , then we cannot control the overlap of contacts over such wide a range . μ ∈ ( 0 , 1] is then a control parameter for the overlap with larger values meaning a higher amount of overlap . ( Note that μ increases with the average fraction concurrent relationships , but to avoid confusion by the concurrency measure of Ref . [23] , we do not call μ concurrency . ) The starting times of the intervals are chosen with uniform probability in the interval [0 , ( 1 – μ ) T] . We proceed by generate a time series with , once again , a truncated power-law shape . We use the equation ( 3 ) where Δmin = 1 , Δmax = 104 and β = 2 . We generate L = 10M such contacts . This times series is then split over the active intervals . When that is finished , the temporal network is done . Note that this procedure does not induce any particular correlations between topology and temporal structure . Ref . [43] uses a similar method that differs in that it does not assign active intervals ( and thus does not have the control parameter μ ) .
|
To understand how diseases spread in a population , it is important to study the network of people in contact . Many methods to model epidemic outbreaks make the assumption that one can treat this network as static . In reality , we know that contact patterns between people change in time , and old contacts are soon irrelevant—it does not matter that we know Marie Antoinette's lovers to understand the HIV epidemic . This paper investigates methods for constructing networks of people that are as relevant as possible for disease spreading . The most promising method we call exponential-threshold network works by letting contacts contribute less , the further from the beginning of an outbreak they take place . We investigate the methods both on artificial models of the contact patterns and empirical data . Except searching for the optimal network representation , we also investigate how the structure of the original data set affects the performance of the representations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"infectious",
"diseases",
"social",
"and",
"behavioral",
"sciences",
"infectious",
"disease",
"epidemiology",
"epidemiology",
"social",
"networks",
"sociology",
"population",
"biology",
"biology",
"infectious",
"disease",
"modeling"
] |
2013
|
Epidemiologically Optimal Static Networks from Temporal Network Data
|
Studies of social microbes often focus on one fitness component ( reproductive success within the social complex ) , with little information about or attention to other stages of the life cycle or the ecological context . This can lead to paradoxical results . The life cycle of the social amoeba Dictyostelium discoideum includes a multicellular stage in which not necessarily clonal amoebae aggregate upon starvation to form a possibly chimeric ( genetically heterogeneous ) fruiting body made of dead stalk cells and spores . The lab-measured reproductive skew in the spores of chimeras indicates strong social antagonism that should result in low genotypic diversity , which is inconsistent with observations from nature . Two studies have suggested that this inconsistency stems from the one-dimensional assessment of fitness ( spore production ) and that the solution lies in tradeoffs between multiple life-history traits , e . g . : spore size versus viability; and spore-formation ( via aggregation ) versus staying vegetative ( as non-aggregated cells ) . We develop an ecologically-grounded , socially-neutral model ( i . e . no social interactions between genotypes ) for the life cycle of social amoebae in which we theoretically explore multiple non-social life-history traits , tradeoffs and tradeoff-implementing mechanisms . We find that spore production comes at the expense of time to complete aggregation , and , depending on the experimental setup , spore size and viability . Furthermore , experimental results regarding apparent social interactions within chimeric mixes can be qualitatively recapitulated under this neutral hypothesis , without needing to invoke social interactions . This allows for simple potential resolutions to the previously paradoxical results . We conclude that the complexities of life histories , including social behavior and multicellularity , can only be understood in the appropriate multidimensional ecological context , when considering all stages of the life cycle .
The cellular slime mold Dictyostelium discoideum is one of the most studied examples of cooperation and altruism in microbes . Upon starvation , solitary amoebae aggregate with neighbors to form a multicellular fruiting body made of stalk and spores . The spores are resistant and will germinate upon encountering favorable conditions while the stalk cells die during stalk development [1–4] . In the process of aggregation these amoebae do not exclude non-kin; consequently , chimeras ( multicellular fruiting bodies consisting of at least two genotypes ) have been observed both in the lab and in nature [5–8] . These chimeras are functional: the multiple genotypes participate both in stalk formation and in spore production ( although not necessarily in equal measures , a phenomenon known as reproductive skew [6] ) . Studies to date have found significant reproductive skew in D . discoideum chimeras [9 , 10] and in a variety of other cellular slime molds [11] . This was interpreted as a sign of social conflict and strains that were over-represented in the spores were seen as socially dominant [9 , 10] . Furthermore , the apparent dominance of some strains over all others ( e . g . , in D . discoideum , linear dominance hierarchies were recorded ) raised into question the persistence of the latter in the wild . Thus , the experimentally uncovered dominance relationships seemed to point towards a decrease in species-wide genetic diversity that was inconsistent with the immense diversity and coexistence observed among strains in nature [5 , 8 , 9 , 11] . Recent studies [12 , 13] have suggested that both the impression of social conflict and the inconsistency between experimental predictions and natural observations arise due to the one-dimensional assessment of D . discoideum fitness , which is equated to spore contribution , when in fact life-history tradeoffs between non-social traits can lead to multiple fitness components . The first proposed tradeoff concerns the response to starvation and lies in the decision between staying vegetative and becoming a spore [12] . In D . discoideum not all cells aggregate to become multicellular; experiments have shown that these cells are viable [12] and theoretical approaches suggested that they could be part of a bet-hedging strategy in uncertain environments [12 , 14]: non-aggregators ( also called loner cells in [12] ) have a high chance of death if the starvation period is long , but , if food does return to the environment , they have a head start against spores that need time to germinate . Given this tradeoff , genotypes that are overrepresented in spores might not be social cheats but rather they could simply be those that have been selected to leave fewer non-aggregators behind [12] . While this tradeoff does not yield coexistence in a well-mixed setting , it can do so in the presence of temporal heterogeneity ( e . g . , seasonality [15] ) , or of spatial heterogeneity and spore dispersal [12] . Subsequently , a second tradeoff was determined empirically , between spore number and viability: genotypes that were overrepresented in spores also made smaller and less viable spores when grown clonally , suggesting that the previously-determined social dominance might be only apparent [13] . This tradeoff was hypothesized to allow coexistence of multiple strains [13] . While [12–14] convincingly argue that D . discoideum fitness has multiple components that , when ignored , can create the illusion of social interactions , there is no immediate link between the two different tradeoffs they propose , nor is there a theoretical framework in which to assess ( a ) how many such tradeoffs are likely to occur and what are the relationships between them , ( b ) by what mechanisms they play out and ( c ) whether they are sufficient to resolve the inconsistency between the lab-based dominance hierarchies and the experimental observations . To this end , we set out to provide such an eco-evolutionary framework in which we study selection on non-social life-history traits in D . discoideum , explore the possible tradeoffs that can arise , and make testable predictions for future empirical work . While the framework is not all-encompassing , it provides a theoretical starting point on which additional ecological and life-history knowledge can be built . We recapitulate existing experimental results without needing to invoke social interactions , we make testable predictions , and we propose improved experimental designs and measures of D . discoideum chimeric interactions that capture the variety of tradeoffs . More broadly , these findings affect our understanding of D . discoideum social behavior and multicellularity and they resonate with recent studies emphasizing the challenges associated with assessing cooperation and free-riding in microbes in the absence of the ecological context under which the trait deemed cooperative has evolved and is maintained [12–14 , 16–20] . We showcase some of the complex outcomes ( e . g . bet-hedging , coexistence ) that ecological context and selection on non-social traits can produce in the absence of social interactions . Many of these outcomes have been theoretically shown before when studied independently ( i . e . focus on one trait or one environmental characteristic , e . g . see [21–23] ) but here we identify powerful synergistic interactions that cannot be predicted from one-dimensional analyses . Finally , these results also contribute towards clarifying the misinterpretations that can arise from sociobiological investigations into microbe behavior that are not grounded in an ecological understanding .
Starting from experimental observations , we build a model of cellular slime molds incorporating hypothesized life-history traits and tradeoffs and their functional consequences [24] . The model we employ is a generalization of [12] to study more broadly life-history tradeoffs in D . discoideum in response to environmental stressors . Since we are interested in nonsocial traits , we study a well-mixed population in which we assume that there are no social interactions between genotypes and in which we do not model the spatial aggregation process . This model therefore ignores interactions between genotypes that may arise during the developmental process . Although D . discoideum and other cellular slime molds are likely to be found in spatially structured environments where movement in the vegetative state is limited , a first well-mixed approach is necessary to tease apart the effects of spatial structure from effects arising from simple life history tradeoffs in a well-mixed setting . In this socially neutral , well-mixed context , aggregation occurs randomly ( with anyone in the population ) and within chimeric aggregates there are no interactions . In a given environment , only one genotype emerges as a winner if given long enough time . For computational optimization , the winning genotype was determined as the most abundant after t = 108 hours in the simulations , when a few genotypes are still present in the system . Longer realizations , however , show that the rest of the genotypes are outcompeted and only one strategy survives , consistent with [12 , 14] . Also consistent with previous results [12 , 14] , deterministic environments always select for pure strategies: environments where food recovers faster select for all non-aggregators ( α = 0 ) and environments where food recovers slower select for all spores ( α = 1 ) ( Fig 2A ) . The switch between these two regions ( all-non-aggregators versus all-spores ) takes place at starvation times close to the maximum lifespan of a non-aggregator in the absence of food . In both regions , the lower the starvation time , the lower the growth rate c . In the first region ( selection for all-non-aggregators ) , due to the very short starvation periods , the whole population increases over time so that each new growth period starts with a higher initial cell density . Since we assume that the food pulse is always the same size , then an increasing population finishes the food faster and thus reduces the length of the growing periods . The shorter the growing periods , the less advantage for the fast reproducing strains . Hence , for very short starvation periods , fast reproducing strains have an initial growth advantage but , over many growth-starvation cycles , they are outcompeted by slow reproducing ones as growth benefits diminish and are outweighed by the higher survival costs incurred during the starvation phase ( S1 Fig , panels A , B ) . As the length of the starvation periods increases , faster reproducing strains start to win ( Fig 2A , S1 Fig , panels A , C ) . In the second region ( selection for all-spores ) , we find the same increasing trend in c as the length of the starvation period increases ( Fig 2A ) . This is due to the fact that few spores die during short starvation times and therefore , the initial population sizes responding to a new food pulse are large and able to consume the food quickly . This results in short growth periods; consequently , strains that reproduce faster do not get enough divisions during a growth cycle to overcome the cost of having a lower spore viability ( S1 Fig , panels A , D ) . Stochastic environments also select for only one surviving genotype , which was identified in simulations as in the deterministic case above . Consistent with previous results [12 , 14] as well as extensive theoretical work on bet-hedging ( e . g . [22 , 31] ) , this winning strategy is pure at the extremes–all-non-aggregators for very fast and all-spores for very slow environments–but mixed otherwise ( Fig 2B ) . This holds regardless of whether we use a discrete or a continuous spore-forming mechanism . In this case the growth rates follow similar trends to the deterministic case for fast and slow environments but follow the opposite trend for intermediate environments ( Fig 2B ) . Henceforth , we will call intermediate environments precisely those for which the reproduction rate of the winning genotype decreases with increasing length of starvation; environments to the left of this region will be called fast and environments to the right of this region will be called slow . The intuition behind the increasing trend in fast and slow environments is the same as for the deterministic case ( S2 Fig , panels A , B , D ) . For intermediate environments , the decreasing trend in division rate of the winning genotype occurs when the starvation periods are long enough to be costly for the survival of non-aggregators but they are still short so that spores incur little mortality leading to short growth periods during which fast reproducing cells do not derive enough benefit to offset the lower viability of their spores ( S2 Fig , panel C ) . For realistic values of the spore death rate ( not too large ) , these qualitative trends are robust with respect to parameter choice both in the stochastic and in the deterministic case ( S3 Fig and S4 Fig ) . The results are independent of the initial conditions ( S5 Fig , panel A ) but they are affected by the saturation behavior assumed for the Michaelis-Menten dynamics ( S5 Fig , panel B ) . In what follows we will focus on stochastic environments since they are more likely to capture the realities of microbial lives [32] . When we refer to fast-environment , intermediate-environment or slow-environment genotypes , we mean the winning genotypes from fast , intermediate , respectively slow environments . To determine the correlations that emerge between reproduction , survival and viability as a function of the traits and tradeoffs included in our model , we used the winning genotypes from each stochastic environment ( Fig 2B ) and we allowed initial clonal populations of identical size to complete one growth phase on identical resources and undergo the sporulation process triggered by starvation upon resource depletion . Thus , unlike in our evolutionary setup above where competing genotypes underwent many successive growth-starvation cycles , for the purposes of exploring correlations between non-social traits we replicated laboratory setups of one growth period followed by the subsequent starvation to allow for comparison with existing data . Since the population partitioning between aggregators and non-aggregators followed by the spore:stalk cell differentiation takes place in our model instantaneously upon resource depletion , we do not need to consider the duration of the starvation phase; all quantities of interest are evaluated at the onset of starvation . We determined the total population size , number of aggregators and number of non-aggregators for each of these genotypes and related them to each other , to the division time c , and to the investment in aggregators versus non-aggregators , α ( Fig 3 ) . At the end of a growth period started from identical and monoclonal initial conditions the population size is constant with respect to c ( Fig 3A ) . This follows from the assumption that resources are used proportional to the division rate and that there is no ( or negligible ) cell mortality during the growth phase: i . e . , strains that divide faster ( higher c ) use more resources per division and therefore reach their carrying capacity sooner , while strains that divide slower ( lower c ) use fewer resources per division and reach their carrying capacity later . However , since all strains start with identical numbers on identical resources , when the resources are exhausted they all reach the same carrying capacity . Thus , when strains grow clonally for only one growth period , higher division rate does not confer a benefit in terms of increased number of cells; however , it continues to incur a cost during starvation in terms of cell survival . This is an essential difference from the evolutionary setup where mixes of several strains compete for resources during many growth-starvation cycles . As a consequence of the hypothesized tradeoff between aggregator and non-aggregator production , the number of aggregators is equal to αP and the number of non-aggregators is equal to ( 1-α ) P , where P is the population size at the end of the growth phase , which , as discussed above , is independent of α and constant with respect to c . Thus , the number of aggregators always increases with α ( Fig 3B ) and it has the same behavior with respect to c as does α: i . e . the production of aggregators increases with the division rate c for slow-environment genotypes , decreases with c for intermediate-environment ones and is constant with respect to c for fast-environment genotypes ( Fig 3C ) . Aggregators and non-aggregators are anticorrelated , so that a high investment in aggregators comes at the cost of a low investment in non-aggregators ( Fig 3D ) . Non-aggregators always decrease with α ( Fig 3E ) and with respect to c they have the inverse behavior of α: i . e . non-aggregator production decreases with c for fast-environment genotypes and increases for intermediate ones ( Fig 3F ) . Slow-environment genotypes do not produce non-aggregators . It is important to note that these results are sensitive to the experimental/simulation conditions . All results above hold if monocultures are allowed to undergo a full growth-starvation cycle . If , as has been the case in experimental studies [13] , monocultures are grown exponentially in abundant resources and then washed and abruptly starved , we find an overall positive correlation between number of spores and division rate ( S6 Fig ) . This yields a negative correlation between number of spores and cell and spore size that agrees with experimental findings reported in [13] . In this neutral context where individuals do not interact with each other except indirectly via their competition for resources ( possible inter-strain interactions during the aggregation and the development of the fruiting body are neglected ) , chimeric success is only apparent and it is simply a measure of which genotype makes more spores in a mix . So far chimeric success has been measured at the end of one starvation event ( i . e . no mixed growth period ) [10 , 11 , 13]; however , genotypes that starve together in nature must have shared at least one growth period ( i . e . the one prior to the starvation event ) . Therefore , in the context of our new understanding of the interplay between life history traits we propose a new measure of chimeric success that emphasizes the importance of both growth and starvation . Since the natural history of slime molds is insufficient to infer how many such growth-starvation cycles two genotypes are likely to share , we restrict our measure of chimeric success to only one growth-starvation cycle . Thus , chimeric success is not a measure of fitness , but rather a measure of spore production in a mixed context . Finally , although chimeras made of several genotypes occur in nature occasionally , experimental work has exclusively focused on pairwise mixes and therefore for ease of comparison we will limit our analysis to this case . The difference between our proposed measure of chimeric success and the existing one consists in the setup: for the existing measure , which we will denote CSS , one starts with a 50:50 mix of two starving genotypes and allows them to aggregate and form spores; for the new measure , which we will denote CSGSX0 , one starts an initial population of size X0 composed of a 50:50 mix of two genotypes on an amount of food , R0 , and allows the cells first to grow , then starve naturally , and subsequently aggregate and form spores . In both setups , in keeping with previous work [9 , 10 , 13] , the chimeric success of a genotype in a pairwise mix is given by its fraction of the total spores; the overall chimeric success of a genotype is the average over all such pairwise mixes . For mathematical definitions and details see Methods . Although the two setups are different , the measure we propose reduces to the existing one in the limit of high cell relative to resource density , when food is insufficient to support growth and therefore cells starve instantaneously . The chimeric success is measured upon completion of spore formation . Thus , for the discrete spore formation mechanism chimeric success is measured at the onset of starvation , when the aggregate and the spores are formed instantaneously; for the continuous spore formation mechanism , the number of spores has to be measured after a standardized starvation time , at which all genotypes will have completed their aggregation . We therefore measure the quantities of interest at time Tsur = 200 hours of starvation , which is the maximum amount of time that a starving solitary cell can live . A more extensive discussion of the continuous mechanism is shown in S1 Text . As already outlined above , since each environment only selects for one genotype , we assume that , at equilibrium , genotypes that can end up in mixes are these winning genotypes , perhaps as a result of dispersal between different environments . Therefore , for the pairwise comparisons we used the winning genotypes from each environment obtained in Fig 2B and started with 50:50 mixes of low ( Fig 4A and Fig 5 top row ) , intermediate ( Fig 4B and Fig 5 middle row ) and high ( Fig 4C and Fig 5 bottom row ) initial cell densities relative to resource magnitude . As before , in what follows , when we refer to fast-environment , intermediate-environment or slow-environment genotypes , we mean the winning genotypes from these respective environments . We want to investigate two aspects . First , we want to explore relative chimeric success between pairs of genotypes ( Fig 4 ) . Second , we want to determine how overall ( average ) chimeric success correlates with non-social life history traits: division time c and implicitly cell survival and spore viability , spore production , aggregators to non-aggregators ratio and number of non-aggregators ( Fig 5 ) . Because in mixes that undergo a growth period the production of spores and non-aggregators depends on the mixing partner , we compare the chimeric success of a genotype not to its production of spores and non-aggregators in monoculture , as determined in the Correlations between non-social life-history traits section , but to its average production of aggregators and non-aggregators , where the average is taken over all pairwise mixes . As explained above for clonal growth , in one growth-starvation cycle the spore-forming mechanism does not affect the growth period . Unlike in the case of monocultures , however , where the population size is constant with respect to c at the end of growth , in mixes , a higher division rate c gives the benefit of faster food consumption and increased population size . Although higher c still incurs the cost of higher mortality during starvation , if the growth period is sufficiently long , its potential benefits can outweigh the costs . Since chimeric success is determined upon starvation-induced fruiting body formation following one growth period , what matters is the length of the growth period . This depends on the initial cell density relative to the amount of available food such that the lower the initial cell density relative to the food pulse , the longer the growth period . Both in the pairwise and in the average scenarios we find that , at low cell-to-resource initial densities , chimeric success is mainly determined by growth and therefore its behavior is dominated by the division rate rather than the aggregator to non-aggregator ratio . As the initial cell density relative to resource magnitude is increased , the two forces–division rate and aggregator to non-aggregator ratio–start to equilibrate and eventually , at high cell-to-resource density where there is only residual growth , the starvation period dominates and the chimeric success is mostly determined by the aggregator to non-aggregator ratio . For pairwise comparisons , regardless of the initial density we find that our strains are organized in a linear hierarchy of chimeric dominance , consistent with lab results for D . discoideum [9 , 10] ( Fig 4 ) . However , the ranking of the strains in the hierarchy , given by the number of pairs ( red squares in Fig 4 ) in which the focal strain produces more spores than its partner , strongly depends on initial cell-to-resource densities . For high initial cell densities , the hierarchy ordering is dominated by the spore investment and we find that the higher the α , the higher the chimeric success of the genotype ( Fig 4C ) . As the initial cell density decreases and the division rate starts to play a stronger role , intermediate-environment genotypes for which spore investment is anticorrelated with the division rate separate into two symmetric groups that interchange their places in the hierarchy ( Fig 4A and 4B ) : the group with the higher α and lower c gets displaced by the group with the lower α and higher c . The slow-environment genotypes continue to top the hierarchy and the fast-environment genotypes continue to be at the bottom of the hierarchy . Moving on to the overall ( average ) chimeric success , we find that at low initial densities it is positively correlated with the division rate ( Fig 5A ) ; it is positively correlated with the average number of spores for fast- and slow-environment genotypes but it is negatively correlated with it for intermediate-environment genotypes for which the negative effect of decreasing c is stronger than the positive effect of increasing α ( Fig 5B and 5C ) . Finally , the behavior of the chimeric success with respect to the number of non-aggregators is the opposite of the behavior with respect to spores ( Fig 5D ) . As the initial cell density relative to resource magnitude is increased , the division rate and aggregator to non-aggregator ratio start to play more equal roles ( Fig 5E–5H ) and eventually , at high cell density , the latter dominates . Thus , at the high density extreme where growth is only residual and starvation dominates , chimeric success correlates positively with the aggregator to non-aggregator ratio ( Fig 5J ) , and with the number of spores ( Fig 5K ) , and it correlates negatively with the average number of non-aggregators ( Fig 5L ) . Chimeric success still correlates positively with division rate for fast-environment genotypes but it correlates negatively with division rate for intermediate-environment genotypes where now the effect of α dominates the effect of c ( Fig 5I ) . For slow-environment genotypes where the spore investment α , is fixed ( maximal ) , the correlation with division rate is slightly positive , since a higher division rate implies a higher production of spores .
We built a theoretical framework , focused on D . discoideum but easily generalizable to other sporulating microbes , to explore the effects of selection on non-social life history traits in variable environments . Within this framework , we were able to qualitatively recapitulate existing results attributed to apparent social behavior in D . discoideum in a model that assumes no social interactions and does not consider any other interaction that may occur during the development of the fruiting body . This highlights the importance of an extended understanding of ecology and life history . In the absence of ecological knowledge–which environment each genotype evolved in , and in the absence of life history knowledge–which life history traits selection acts on and what relationships exist between them , one can incorrectly interpret differences in spore investment as chimeric success and attribute the latter to complex interactions ( e . g . social cheating ) . In the neutral context , previously paradoxical findings do not even arise . Specifically , we assume that two genotypic traits are under selection–division rate and aggregator to non-aggregator ratio ( the latter yielding an immediate anticorrelation between spore number and non-aggregator number ) . We find a negative correlation between spore number and aggregation time ( see S1 Text for details ) . Furthermore , depending on the simulation setup ( encapsulating different experimental setups ) , additional correlations can be identified between spore production and cell and spore size , survival and viability , such that spore production generally comes at a cost of spore viability , consistent with [13] . In this multi-trait context , measures that attempted to quantify success upon starvation–such as spore number and chimeric success , were revealed to be ill-defined , as we showed starvation and growth to be inextricably linked . We therefore proposed a new and more general measure of chimeric success that accounts for both growth and starvation and found that genotypes are organized in a linear hierarchy based on their pairwise chimeric success , consistent with experimental results [9 , 10] . Furthermore , the overall chimeric success of a genotype , measured as an average over pairwise comparisons , generally increases with the division rate ( i . e . decreases with cell/spore size ) consistent with experimental results [13] , it increases with the average number of spores produced and with the aggregator to non-aggregator ratio , and it decreases with aggregation time ( see S1 Text for details ) . Finally , we found that the relationship between chimeric success and investment in non-aggregators depends on the experimental conditions . Our results are robust with respect to the choice of mechanism underlying the non-social traits ( e . g . stochastic switching versus phenotypic variation ) . Despite the multiple life history traits and associated tradeoffs , however , we found that each environment , whether deterministic or stochastic , selects for only one winning genotype . Coexistence can be achieved in the current model if we incorporate spatial heterogeneity and weak-to-moderate dispersal between different environments ( consistent with [12] ) , or temporal heterogeneity [15] . In the absence of any spatial or temporal heterogeneity , multiple genotype coexistence is possible only when strains balance their tradeoffs so that they have identical fitness , as suggested in [13] . Such balancing however reduces the scenarios where coexistence may occur and yields them ungeneric . This is in agreement with classic works in community ecology that show that in the absence of frequency-dependent mechanisms tradeoffs alone do not generally result in coexistence [33] . Several future directions arise . First , our results constitute general testable predictions ( Table 1 ) and emphasize the importance of standardized measures and experimental protocols , ideally chosen to most closely approximate natural conditions . Future work needs to build on existing work in life-history theory that highlights that the reproductive output of just one life-stage is not necessarily an appropriate fitness measure [34–36] , to propose a new measure of D . discoideum fitness , that accounts for the way natural selection acts on the life histories of individuals and their ancestors , in variable environments . Recent work has proposed a very elegant framework in bacteria that could be adapted for other microbes [37] . Second , although we were able to qualitatively capture existing results in a well-mixed , neutral framework , this dismisses neither the importance of spatial structure , which is likely to influence the dynamics at least quantitatively , nor the possibility of social interactions in general . While the linear hierarchies in D . discoideum are well aligned with the neutral hypothesis , other slime mold species have exhibited short dominance loops [11] , possibly indicative of small subsets of interacting genotypes . Even in D . discoideum , there exists at least one extreme lab mutant that has been shown to interact in chimeras , a social parasite that is unable to form its own stalk but uses other genotypes’ stalks to support its spores [38] . Although such a mutant has not been found in nature , it has been shown to be very destructive in lab experiments [38] and therefore its effects are worth re-investigating using this new multi-trait understanding of fitness . One hypothesis is that the non-aggregators can act as buffers against the destructive effects of a parasite that can only exploit the social aspects of the behavior . This is consistent with existing results from the theory of cooperation and sociality showing that individuals that do not participate in the social dynamics are of primary importance for the evolution of social behavior [39 , 40] . Third , the life cycle of D . discoideum may involve many other tradeoffs to be considered in the future , such as allocating more cells in the spore body at the expense of reducing the dispersal ability by creating a shorter stalk . Fourth , for simplicity , we have encapsulated here all the ecological variability in the starvation times; however , environmental heterogeneity may also come into play in other ways , e . g . in the way resources reappear in the system ( progressively instead of instantaneously , or seasonally instead of throughout the year [15] ) or in the amount of nutrients available at the beginning of each growth cycle , which here we have taken to be constant as a first approach . This could introduce additional tradeoffs arising from strains showing different feeding behavior [41] . Finally , our model does not include mutation , meiotic recombination or horizontal gene transfer , all of which can play an important role in producing diversity that could lead to coexistence , at least temporarily . Although mutation rates are very low in D . discoideum [42] , meiotic recombination appears to occur at a sufficiently high rate that it could potentially influence population composition [43] . Our results show broadly that in fluctuating environments multicellularity and sociality are just part of a set of risk-management strategies . Although our work was motivated by D . discoideum in particular and by microbes in general , these results can be extended to other species where similar ecologically-induced tradeoffs between aggregation time , size and social behavior have been identified [44 , 45] .
To determine the winning genotype in each environment , we performed numerical simulations of several growth-starvation cycles in environments defined by their starvation time ( for deterministic ones ) or by their average starvation time ( for stochastic ones ) . For the latter , starvation times were a stochastic variable exponentially distributed . The spectrum of genotypes was discretized , using 44772 genotypes for stochastic environments ( 1092 values of α and 41 values of c ) and 4141 for deterministic environments ( 101 values of α and 41 values of c ) . The values of c range between c = 0 . 05 and c = 0 . 45 with a sampling of 10−2 . In stochastic environments , the values of α were chosen with a sampling of 10−4 between α = 0 and α = 0 . 1 and with a sampling of 10−2 between α = 0 . 1 and α = 1 . This irregular sampling was chosen to avoid abrupt jumps in the winning c in environments that select for strategies with a small investment in spores . In deterministic environments , however , the sampling in α was homogeneous , with a sampling step of 0 . 01 between α = 0 and α = 1 . In deterministic environments a single run was used ( randomness introduced by the initial condition was tested and shown to be irrelevant in determining the stationary state ) while in stochastic environments averages were taken over 20 independent realizations . In both cases simulations were run until t = 108 hours and the winning genotype defined as the most abundant at that time . This was done for computational feasibility , longer simulations shown that in the long-term only one genotype survives . In stochastic environments the variance of this measurement is very low and the mean value coincides with the result of each single run . S1 Fig and S2 Fig show the short time evolution of some genotypes in deterministic and stochastic environments . Initial abundances of each genotype were independently drawn from a standard log-normal distribution and subsequently normalized so that the entire population contained 108 cells . An initial resource pulse of magnitude 108 was added and the trajectories governed by Eq ( 1 ) was integrated using a Runge-Kutta numerical method until resources are exhausted ( R reaches a zero value ) . The depletion of the resources triggers the onset of starvation and the population instantaneously splits between aggregators and non-aggregators and aggregators instantaneously differentiate between spores and stalk cells . A fraction α of cells aggregate and the remaining 1 - α remains as solitary ( non-aggregated ) cells . Finally , the population of aggregators is multiplied by a factor s = 0 . 8 that accounts for the spore:stalk cell differentiation . The starvation onset is followed by a starvation phase in which both populations decline due to the death of spores and non-aggregated cells . To evaluate the final population size of both classes of cells at the end of the starvation phase , due to the fact that Eq ( 3 ) has an analytical solution , the functions are evaluated at t = T ( if the environment is stochastic , this time is previously sorted from an exponential distribution ) , which significantly speeds up the simulations . Simulations are stopped at the starvation onset , when resources have been exhausted and the population splits between aggregators and non-aggregators but before entering in the starvation phase in which cells and spores die . Genotypes underwent one growth and after resources consumption , the populations split between a fraction α of aggregators and ( 1-α ) of non-aggregators . Let Xα , csp be the amount of spores produced by genotype ( α , c ) , which is obtained by multiplying the number of aggregators by a factor s = 0 . 8 to account for the stalk:spore cell differentiation . The pairwise chimeric success of genotype ( α , c ) against genotype ( α , c′ ) is given by Xα , csp/ ( Xα , csp+Xα′ , c′sp ) and the average chimeric success of genotype ( α , c ) is given by: CS ( α , c ) =1Nst−1∑α′ , c′Xα , cspXα , csp+Xα′ , c′sp ( 5 ) where the sum runs over all the possible values of α and c except for α = α′ and c = c′ simultaneously and Nst is the total number of strains considered in the average . To construct the linear hierarchies ( Fig 4 ) , the 31 winning genotypes obtained from Fig 2B are ordered according to the environment they evolved in . Label 0 corresponds to the winning genotype from the environment with λT = 10 hour and label 30 to the winning genotype that evolved in the environment with λT = 104 hour . All strains are mixed in pairs , they undergo a growth-starvation cycle , and pairwise chimeric success is measured . To define a dominance / subordination relationship between the genotypes we set one of the genotypes in the mix as the reference genotype and its partner as the mixed genotype . If the reference genotype makes more spores than the mixed genotype , then it has a higher chimeric success , which is indicated by a red square in Fig 4 . If the reference genotype makes fewer spores it will have a lower chimeric success ( blue squares in Fig 4 ) , and both genotypes will be equally successful if the number of spores is the same ( gray squares in Fig 4 ) .
|
Fitness in social microbes is often measured in terms of reproductive success in the social stage , with little regard to other stages of the life cycle ( e . g . solitary ) or to the ecological context . This approach can lead to seemingly paradoxical results that point to complex social interactions ( e . g . , social cheating ) among individuals in the population . However , recent experimental studies in Dictyostelium discoideum , one of the most studied social microbes , have highlighted various tradeoffs among previously ignored non-social traits that should affect fitness . We develop an ecologically-motivated socially-neutral model for the life cycle of D . discoideum that combines these proposed traits and tradeoffs and proposes new ones to determine whether existing observations can be explained without the need to invoke social interactions . We confirm this expectation and conclude that the complexities of social behavior can only be understood in the appropriate ecological context , when considering a complete description of the life cycle .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"death",
"rates",
"cell",
"death",
"ecology",
"and",
"environmental",
"sciences",
"dictyostelium",
"fruiting",
"body",
"demography",
"cell",
"cycle",
"and",
"cell",
"division",
"dictyosteliomycota",
"cell",
"processes",
"microbiology",
"dictyostelium",
"spore",
"cells",
"developmental",
"biology",
"model",
"organisms",
"theoretical",
"ecology",
"population",
"biology",
"microbial",
"growth",
"and",
"development",
"research",
"and",
"analysis",
"methods",
"dictyostelium",
"discoideum",
"microbial",
"physiology",
"slime",
"molds",
"life",
"cycles",
"protozoan",
"models",
"dictyostelium",
"growth",
"and",
"development",
"microbial",
"ecology",
"people",
"and",
"places",
"population",
"metrics",
"cell",
"biology",
"ecology",
"biology",
"and",
"life",
"sciences",
"dictyostelium",
"protists",
"organisms"
] |
2016
|
Lack of Ecological and Life History Context Can Create the Illusion of Social Interactions in Dictyostelium discoideum
|
Ixodes scapularis is the most medically important tick species and transmits five of the 14 reportable human tick borne disease ( TBD ) agents in the USA . This study describes LC-MS/MS identification of 582 tick- and 83 rabbit proteins in saliva of I . scapularis ticks that fed for 24 , 48 , 72 , 96 , and 120 h , as well as engorged but not detached ( BD ) , and spontaneously detached ( SD ) . The 582 tick proteins include proteases ( 5 . 7% ) , protease inhibitors ( 7 . 4% ) , unknown function proteins ( 22% ) , immunity/antimicrobial ( 2 . 6% ) , lipocalin ( 3 . 1% ) , heme/iron binding ( 2 . 6% ) , extracellular matrix/ cell adhesion ( 2 . 2% ) , oxidant metabolism/ detoxification ( 6% ) , transporter/ receptor related ( 3 . 2% ) , cytoskeletal ( 5 . 5% ) , and housekeeping-like ( 39 . 7% ) . Notable observations include: ( i ) tick saliva proteins of unknown function accounting for >33% of total protein content , ( ii ) 79% of proteases are metalloproteases , ( iii ) 13% ( 76/582 ) of proteins in this study were found in saliva of other tick species and , ( iv ) ticks apparently selectively inject functionally similar but unique proteins every 24 h , which we speculate is the tick's antigenic variation equivalent strategy to protect important tick feeding functions from host immune system . The host immune responses to proteins present in 24 h I . scapularis saliva will not be effective at later feeding stages . Rabbit proteins identified in our study suggest the tick's strategic use of host proteins to modulate the feeding site . Notably fibrinogen , which is central to blood clotting and wound healing , was detected in high abundance in BD and SD saliva , when the tick is preparing to terminate feeding and detach from the host . A remarkable tick adaptation is that the feeding lesion is completely healed when the tick detaches from the host . Does the tick concentrate fibrinogen at the feeding site to aide in promoting healing of the feeding lesion ? Overall , these data provide broad insight into molecular mechanisms regulating different tick feeding phases . These data set the foundation for in depth I . scapularis tick feeding physiology and TBD transmission studies .
Ticks surpass all arthropods in transmission of a greater variety of pathogens including fungi , viruses , bacteria , and protozoa [1 , 2] . In livestock production , ticks and tick-borne diseases ( TBD ) have caused annual losses in billions of US dollars globally [3 , 4] . Recently , ticks have gained the attention in public health policy with a recent publication that advocated for One Health solutions listing 17 human TBDs among sources of human health concerns [5] . Ixodes scapularis , commonly known as the deer tick or blacklegged tick , is among the most medically important tick species and transmits 5 of the 14 human TBD agents in the USA: Borrelia burgdorferi [6] , Anaplasma phagocytophilum [7] , Borrelia miyamotoi [8] , Babesia microti [9] , and Powassan virus disease [10] . Likewise , close relatives of this tick including I . pacificus on the west coast of the USA and I . ricinus in Europe are vectors of important human TBD agents including B . burgdorgferi , B . miyamotoi , and A . phagocytophilum [11] . On this basis , the I . scapularis genome was sequenced [12 , 13] and these data has provided opportunities for in depth studies of biological adaptations that make ticks successful vectors of pathogens . These data were postulated to facilitate studies that will reveal weaknesses that can be targeted for development of novel tick control methods [13] . In absence of effective vaccines against TBDs , controlling ticks using acaricides remains the most reliable method [14 , 15] . Although effective in the short term , limitations of chemical acaricides such as selecting resistant tick populations , costs in new acaricide development , environmental and food contamination have necessitated the search for alternative tick control methods [4 , 16 , 17] . Immunization of animals has been advocated as a sustainable alternative tick control method [18] . The major limitation toward global adoption of anti-tick vaccines as an alternative tick control method is availability of effective target antigens . We are interested in understanding tick feeding physiology as a means to identify physiologically important proteins that can be targeted for anti-tick vaccine development . Ticks accomplish feeding by lacerating the vasculature of the surrounding host tissue and sucking up blood that bleeds into the feeding lesion [19–21] . This destructive feeding style triggers the host defense mechanisms such as hemostasis , inflammation and immune responses . However , ticks counteract the host defense mechanisms by secreting pharmacologically active molecules in saliva to modulate host defenses [19 , 20 , 22–24] . In addition to blood meal acquisition , tick saliva proteins are also involved with the transmission and acquisition of TBD agents [25] . Reports of reduced pathogen transmission to repeatedly tick infested animals that developed resistance to tick feeding [26–29] provide credence to the importance of tick saliva proteins in vector tick competence . Thus , identification of tick saliva proteins will provide a basis for development of novel methods to interfere with tick feeding and prevention of pathogen transmission . With the advent of next-generation sequencing ( NGS ) technologies , tick salivary gland transcriptomes have been described [23 , 30–39] . However , the major limitation to these data is that it does not inform on which transcripts that encode for proteins are secreted in tick saliva . In an interesting approach to identify secreted tick salivary proteins ( TSPs ) Radulovic et . al . , [40] and Lewis et . al . , [41] used antibodies to 24–48 h tick saliva proteins [42 , 43] to immunoscreen phage display cDNA expression libraries to identify 24–48 h Amblyomma americanum and 24 h I . scapularis immunogenic tick saliva proteins . Similar immunoscreening approaches were used to identify immunodominant I . scapularis tick saliva proteins [44–47] . In a related study , saliva of I . scapularis was analyzed by Edman degradation identifying 15 proteins [48] . Recently proteins in saliva of ixodid ticks from replete fed Rhipicephalus sanguineus [49] , partial and replete fed Rhipicephalus microplus [50] , three and five day fed Dermacentor andersoni [51] , and replete fed adult and nymph Haemaphysalis longicornis [52] were identified . In argasid ticks , a lone study identified saliva proteins from twice fed Ornithodoros moubata ticks with saliva collected after 4 months from feeding [53] . Whereas studies reviewed here identified proteins in saliva of ticks at one or two feeding time points , this study has described proteins that I . scapularis ticks likely inject into animals every 24 h during the first five days of feeding and toward the end of the tick feeding process . The catalog of I . scapularis tick saliva proteins in this study provides an in depth view at protein families and/or molecular systems that are at play at the I . scapularis tick and host interface .
All experiments were done according to the animal use protocol approved by Texas A&M University Institutional Animal Care and Use Committee ( IACUC ) ( AUP 2011–207 and 2011–189 ) that meets all federal requirements , as defined in the Animal Welfare Act ( AWA ) , the Public Health Service Policy ( PHS ) , and the Humane Care and Use of Laboratory Animals . I . scapularis ticks were purchased from the tick rearing facility at Oklahoma State University ( Stillwater , OK , USA ) . Prior to feeding on rabbits , female ticks were paired with males to mate . Ticks were considered mated once males were detached from the females . Routinely , ticks were fed on rabbits as previously described [43] . Mated I . scapularis ticks were restricted to feed onto the outer part of the ear of New Zealand rabbits with orthopedic stockinet’s glued with Kamar adhesive ( Kamar Products Inc . , Zionsville , IN , USA ) . A total of 84 adult I . scapularis ticks ( 42 per ear ) were placed into tick containment apparatus on three rabbits and allowed to attach . To collect tick saliva , female ticks partially fed for 24 h ( n = 43 ticks ) , 48 h ( n = 40 ticks ) , 72 h ( n = 40 ticks ) , 96 h ( n = 40 ticks ) , 120 h ( n = 40 ticks ) as well as apparently fully fed but not detached from the host ( BD , n = 8 ticks ) and spontaneously detached ticks ( SD , n = 6 ticks ) were rinsed in Milli-Q water and dried on a paper towel . Rinsed ticks were placed dorsal-side down on double-sided tape on a glass slide . Salivation was induced by injecting 1–3 μL of 2% pilocarpine hydrochloride in phosphate buffered saline ( PBS , pH 7 . 4 ) on the ventral side adjacent to the fourth leg coxa using a 34 gauge/ 0 . 5 inches/ 45° angle beveled needle on a model 701 Hamilton syringe ( Hamilton Company , Reno , NV , USA ) . Subsequently , saliva was collected every 15–30 min using a Hamilton syringe for approximately 4h at room temperature . Saliva of I . scapularis ticks ( at least 2 μg total protein per run X3 ) for each specific feeding time point was digested in solution with trypsin . Saliva were diluted in 8 M urea/0 . 1 M Tris , pH 8 . 5 , reduced with 5 mM Tris ( 2-carboxyethyl ) phosphine hydrochloride ( TCEP , Sigma-Aldrich , St Louis , MO , USA ) and alkylated with 25 mM iodoaceamide ( Sigma-Aldrich ) . Proteins were digested overnight at 37°C in 2 M urea/0 . 1M Tris pH 8 . 5 , 1 mM CaCl2 with trypsin ( Promega , Madison , WI , USA ) with a final ratio of 1:20 ( enzyme:substrate ) . Digestion reactions , in a final concentration of 0 . 15 μg/mL , were quenched with formic acid ( 5% final concentration ) and centrifuged for debris removal . Reversed phase pre-columns were prepared by first creating a Kasil frit at one end of a deactivated 250 μm ID/360 μm OD capillary ( Agilent Technologies , Santa Clara , CA , USA ) . Kasil frits were prepared by dipping 20 cm capillary in 300 μL Kasil 1624 ( PQ Corporation , Malvern , PA , USA ) and 100 μL formamide solution , curing at 100°C for 3 h and adjusting the length . Pre-columns were packed in-house ( John Yates III's Laboratory , The Scripps Research Institute , La Jolla , CA , USA ) with 2 cm of 5 μm ODS-AQ C18 ( YMC America , Inc . , Allentown , PA , USA ) particles from particle slurries in methanol . Analytical reversed phase columns were fabricated by pulling a 100 μm ID/360 μm OD silica capillary ( Molex Polymicro Technologies , Austin , TX , USA ) to a 5 μm ID tip . The same packing material was packed until 20 cm directly behind the pulled tip . Reversed phase pre-columns and analytical columns were connected using a zero-dead volume union ( IDEX Corp . , Upchurch Scientific , Oak Harbor , WA , USA ) . Peptide mixtures were analyzed by nanoflow liquid chromatography mass spectrometry using an Easy NanoLC II and a Q Exactive mass spectrometer ( Thermo Scientific , Waltham , MA , USA ) . Peptides eluted from the analytical column were electrosprayed directly into the mass spectrometer . Buffer A and B consisted of 5% acetonitrile/0 . 1% formic acid and 80% acetonitrile/0 . 1% formic acid , respectively . The flow rate was set to 400 nL/min . Feeding time saliva samples ( 1 . 5 μg per injection ) were separated in 155 min chromatographic runs , as follows: 1–10% gradient of buffer B in 10 min , 10–40% of buffer B in 100 min , 40–50% of buffer B in 10 min and 50–90% of buffer B in 10 min . Column was held at 90% of buffer B for 10 min , reduced to 1% of buffer B and re-equilibrated prior to next injection . The mass spectrometer was operated in a data dependent mode , collecting a full MS scan from 400 to 1 , 200 m/z at 70 , 000 resolution and an AGC target of 1 x 106 . The 10 most abundant ions per scan were selected for MS/MS at 17 , 500 resolution and AGC target of 2 x 105 and an underfill ratio of 0 . 1% . Maximum fill times were 20 and 120 ms for MS and MS/MS scans , respectively , with dynamic exclusion of 15 s . Normalized collision energy was set to 25 . The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium [54] via the PRIDE partner repository with the dataset identifier PXD003214 . Tandem mass spectra were extracted from Thermo RAW files using RawExtract 1 . 9 . 9 . 2 [55] and searched with ProLuCID [56] against a non-redundant database containing an Ixodidae database from National Center for Biotechnology Information ( NCBI , www . ncbi . nlm . nih . gov ) ( 62 , 246 entries ) concatenated with Oryctolagus cuniculus from Uniprot ( www . uniprot . org ) reference database ( 21 , 148 entries ) and reverse sequences of all entries . Database sequence redundancies were removed by FastaDBXtractor module from PatternLab for Proteomics platform [57] . Searches were done using Integrated Proteomics Pipeline–IP2 ( Integrated Proteomics Applications , Inc . , San Diego , CA , USA ) . The search space included all fully-tryptic and half-tryptic peptide candidates . Carbamidomethylation of cysteine was used as static modification . Data was searched with 50 ppm precursor ion tolerance and 20 ppm fragment ion tolerance . The validity of the peptide spectrum matches ( PSMs ) generated by ProLuCID [56] was assessed using Search Engine Processor ( SEPro ) module from PatternLab for Proteomics platform [57] . Identifications were grouped by charge state and tryptic status , resulting in four distinct subgroups . For each group , ProLuCID XCorr , DeltaCN , DeltaMass , ZScore , number of peaks matched and secondary rank values were used to generate a Bayesian discriminating function . A cutoff score was established to accept a protein false discovery rate ( FDR ) of 1% based on the number of decoys . This procedure was independently performed on each data subset , resulting in a false-positive rate that was independent of tryptic status or charge state . Additionally , a minimum sequence length of six residues per peptide was required . Results were post processed to only accept PSMs with <10ppm precursor mass error . BLASTP searches against several databases were performed to annotate the matched proteins . To check tick proteins identity , the following databases were used: non-redundant ( NR ) , Acari and refseq-invertebrate from NCBI , Acari from Uniprot , the GeneOntology ( GO ) FASTA subset [58] , MEROPS database [59] , and the conserved domains database of NCBI [60] containing the COG [61] , PFAM[62] , and SMART motifs [63] . To check rabbit proteins , the following databases were used: Oryctolagus cuniculus and refseq-vertebrates databases from NCBI , O . cuniculus from Uniprot , the GeneOntology ( GO ) FASTA subset [58] the conserved domains database of NCBI [60] , containing the COG [61] , PFAM [62] , and SMART motifs [63] . To functionally classify the protein sequences , a program provided by Dr . José M . C Ribeiro written in Visual Basic 6 . 0 ( Microsoft , Redmond , Washington , USA ) was used [34] . The functionally annotated catalog for each dataset was manually curated and input in a hyperlinked Excel spreadsheet ( S1 and S2 Tables ) . To determine the relative abundance of saliva proteins normalized spectral abundance factors ( NSAF ) were used . The NSAF value was validated as reliable in a label-free relative quantification approach [64–66] . Average NSAF of two or three replicates were used . To determine relative abundance , average NSAF for each protein functional class or an individual annotated protein was expressed as a percent ( % ) of total NSAF per time point . To visualize relative expression patterns on a heat map , % NSAF values were normalized using Z-score statics using the formula Z=X−μσ , where Z is the Z-score , X is the NSAF for each protein per time point , μ is the mean throughout time points , σ is the standard deviation throughout time points . Normalized NSAF values were used to generate heat maps using the heatmap2 function from the gplots library in R [67] . Amino acid sequences were used to construct a guide phylogeny tree using MacVector 12 . 7 . 3 ( MacVector Inc Cary , NC , USA ) software . Protein sequences were aligned using Muscle method in MacVector under default settings . Subsequently , the tree was constructed using the Neighbor Joining method with uncorrected ( “p” ) distance setting . To estimate bootstrap values , replications were set to 1000 .
We successfully harvested pilocarpine-induced saliva of I . scapularis ticks that were partially fed on rabbits for 24 , 48 , 72 , 96 and 120 h as well as those that were apparently engorged but not detached ( BD ) , and those that had engorged and spontaneously detached ( SD ) . During collection of saliva , we observed that saliva of 24 h fed ticks dried up quickly forming flakey white crystal-like residues , and to collect we dissolved these flakes in 2μL sterile phosphate buffered saline ( PBS , pH 7 . 4 ) batches . On the contrary , saliva droplet of ticks at subsequent feeding stages was visible within seconds to min after pilocarpine injection . S1 Table lists tick and rabbit proteins that were identified in I . scapularis saliva . The search of extracted tandem mass spectra against the tick and rabbit protein database using ProLucid [56] and filtering using SEPro [57] produced hits to 769 tick and 130 rabbit proteins respectively with at least one peptide match per protein ( S1 Table , please note the different tabs ) . When subjected to further analysis in BirdsEye View module from PatternLab for Proteomics platform [57] , 582 of the 769 tick proteins were determined to be authentic as they were detected in two or all of the three runs , while the remaining 187 proteins detected in only one of the three runs were considered low confidence hits and not further discussed ( S1 Table ) . Of the 130 rabbit proteins that were detected in I . scapularis tick saliva , 83 met the criteria for authentication . When subjected to auto-annotation [34] , 582 tick and 83 rabbit high confidence proteins respectively classified into 24 ( Table 1 ) and 18 ( Table 2 ) functional protein classes . Specifically Tables 1 and 2 summarizes cumulative numbers of proteins that were identified in each functional class , apparent relative abundance at each time point , and time points at where class were not detected [represented by zero ( 0 ) ] . Figs 1 and 2 gives a snap shot of relative abundance of tick ( Fig 1 ) and rabbit ( Fig 2 ) proteins in I . scapularis saliva every 24 h . In Fig 1 , it is apparent that majority of I . scapularis tick proteins in this study belong to four predominant functional protein classes starting with proteins of unknown function , followed by protease inhibitors ( PI ) , antimicrobial/immunity related , and heme binding proteins . This is followed by lowly abundant protein classes that account for 1–6% ( cytoskeletal , glycine rich , and protein modification machinery ) with the remaining protein classes being detected accounted for less than 1% . Of the four major protein classes , relative abundance of proteins of unknown function appear to increase with feeding , accounting for 33–58% of total protein between 24–120 h before dropping to 13% in saliva of fully fed but not detached ticks ( BD ) as well as fully fed and spontaneously detached . Similarly , heme binding proteins increased from ~14% at 24 h to ~24% at 96 h , before dropping to 10% at 120 h , coming back up to 24% in BD and dropping to 8% in SD . On the other hand , PIs and antimicrobial/immunity related peptides decreased in abundance with feeding with the former dropping from 24 . 7% at 24 h to 17–8% at 48–120 h respectively , but increasing to 21% in BD and dropping to 7% in SD . Similarly anti-microbial/immunity-related proteins decreased from 18% at 24 h to 17–5% at 48–120 h , before slightly rising to ~7% in BD and SD ( Table 1 and Fig 1 ) . Notable protein classes include proteases and lipocalins that appear to increase in abundance with feeding . Protease content increases from 0 . 5% at 24 h to 3–6% at 48 , 96 , and 120 h except for 72 h were content was at 0 . 5% , and 5–6% in BD and SD ( Fig 1 ) . Similarly lipocalin content increases from 0 . 3% at 24 h to ~6% at 120 h , not detected in BD , but accounted for ~3% of protein content in SD . Also notable in Table 1 and Fig 1 , tick housekeeping-like proteins appear to increase with feeding . Fig 2 summarizes relative abundance of rabbit proteins that were detected in I . scapularis saliva . It is interesting to note that of the 18 protein classes in Table 2 , four protein classes: heme/iron , hemoglobin/RBC degradation products , antimicrobial/immunity related , and keratin were found in all time points . It is notable that these four protein classes represented the most abundant rabbit proteins in tick saliva . Except at 72 h where rabbit heme/iron binding proteins accounted for 3% , this protein was among the most predominant in other time points accounting for 17–75% of total rabbit protein content in tick saliva . Similarly , hemoglobin/RBC-related proteins increased from 7 . 6% at 24 h to 50 . 2% in BD and 39 . 3% in SD . Immunity-related proteins of rabbits were most abundant at 48 , 96 , and 120 h saliva at 30% , 29 . 6% , and 27 . 6% respectively ( Fig 2 ) . Keratins detected at all time points could signal handling contamination of our samples . Another interesting observation in Fig 2 , fibrinogen the precursor to fibrin , which is involved in clot formation was detected toward the end of feeding , 0 . 25% in 96 h saliva increasing to 3 . 1% in SD . Could this suggest that the tick ingests fibrinogen during feeding and secretes it back into the host during the detachment phase to promote wound healing ? Hard ticks create a wound in host skin from which they suck the blood , however this wound is completely healed when ticks detach . We are interested in understanding mechanisms that regulate early stage tick feeding , and thus the subsequent discussion of data is biased toward non-housekeeping-like tick derived proteins that were found in saliva from 24/48 h . We have discussed rabbit derived proteins separately , but highlight similarities and differences where appropriate . We used Z-statistics normalization of NSAF values ( S3 Table ) to develop heat maps in Figs 3–5 . These data give insight into relative abundance of specific proteins during feeding: proteases ( Fig 3A ) , protease inhibitors ( Fig 3B ) , lipocalins/tick histamine-binding proteins/fatty acid binding proteins ( Fig 4A ) , anti-microbial/immunity-related ( Fig 4B ) , heme-binding proteins ( Fig 4C ) , anti-oxidants ( Fig 4D ) , proteins of unknown function ( Fig 5A ) , glycine rich proteins ( Fig 5B ) and extracellular matrix proteins ( Fig 5C ) . The I . scapularis genome encodes for at least 233 putatively active and 150 putatively inactive proteases belonging to serine , cysteine , aspartic , metallo , and threonine protease families [68] . In this study we found 33 proteases in four clans: serine- ( n = 3 ) , cysteine- ( n = 3 ) , aspartic- ( n = 1 ) , and metalloproteases ( n = 26 ) ( S2 Table ) . When searched against the Merops database [59] , the 26 metalloproteases belong to families M12 ( n = 15 ) , M20 ( n = 4 ) , M2 ( n = 2 ) , M28 ( n = 2 ) , M13 ( n = 1 ) , M17 ( n = 1 ) , and M49 ( n = 1 ) ( S2 Table ) , while serine , cysteine and aspartic proteases are classified in families S1 , C1 and A1 respectively . Most of the proteases here are likely associated with tick feeding regulation in that 75% ( 25/33 ) were detected between 24–120 h during tick feeding except for seven that were identified only in SD ( S2 Table ) . Nearly 40% of proteases in the I . scapularis genome are metalloproteases [68] . Whether or not the observation in this study that majority of proteases in I . scapularis tick saliva are metalloproteases reflects the protease composition in I . scapularis genome or it is a physiological event , is unknown at this point . Z-score statistic analysis and visualization of normalized NSAF values of the 26 metalloproteases ( Fig 3A ) show that M12 and M2 metalloproteases were likely secreted in high abundance between 24–120 h during feeding respectively , which could indicate the importance of these proteins in regulating the first five days to tick feeding . The remaining metalloproteases in families M17 , 20 , 28 and 49 , which were abundant in BD and SD ( Fig 3A ) are not likely associated with regulating tick feeding events . The observation that I . scapularis predominantly secreted family M12 metalloproteases during feeding is suggestive of the importance of this protein class in tick feeding physiology . Emerging evidence indicate that this is the case . A recombinant protein of M12 protease ( AAP22067 . 1 , MCC Fig 3A ) has gelatinase and fibrin ( ogen ) olytic activities [69] , which is a pro-tick feeding event . In a related study , RNAi silencing of AAM93625 . 1 ( MCC Fig 3A ) and AAT92201 . 1 ( MCF Fig 3A ) homologs , Metis 1 and 2 ( CAO000625 and CAO000626 ) in I . ricinus impaired blood meal feeding and egg laying with salivary gland protein extracts of these ticks not affecting host fibrinolysis [70] . In related studies , snake venom M12 proteases were associated with hemorrhaging , edema , hypotension , hypovolemia , inflammation and necrosis [71–73] some of which will promote tick feeding . It will be interesting to characterize the role ( s ) of tick saliva proteases identified in this study . The first line of host defense to tick feeding such as inflammation , platelet aggregation , blood clotting , complement activation , and cellular immunity are mediated by proteases that are controlled by protease inhibitors ( PI ) . From this perspective , it has been hypothesized that ticks could inject PIs into the host to evade host defense [18 , 74–77] . In this study , we identified 43 putative PIs ( S1 Table ) , which according to the Merops database belong in eight families: I2 ( Kunitz type serine protease inhibitors , n = 2 ) , I4 ( serine protease inhibitors , [serpins] , n = 11 ) , I8 ( TIL domain serine protease inhibitors , n = 13 ) , I25 ( cystatins , cysteine protease inhibitors , n = 4 ) , I31 ( thyropins , cysteine protease inhibitors n = 2 ) , I39 ( α-2 macroglobulin , A2M , n = 9 ) , I43 ( Kazal type serine protease inhibitors , n = 1 ) , and I68 ( carboxypeptidase inhibitors , TCI , n = 1 ) were identified in I . scapularis saliva ( S2 Table ) . It is notable that 84% ( 36/43 ) of PIs were detected in 24 and 48 h saliva ( S2 Table ) , suggesting the potential for these proteins to regulate early stages of tick feeding . The observation here that majority of PIs in this study are likely inhibitors of serine proteases could signal the potential that most host defense pathways to tick feeding are likely serine protease mediated . Similar to other protein classes in this study , relative abundance of PIs varied every 24 h ( Fig 3B ) . Serpins show three secretion profiles: SCA proteins are abundant in first 48 h and decrease with feeding , SCB are abundant at 96 h and increase in SD saliva , and SCC proteins increase in abundance from 24 to 120 h ( Fig 3B1 ) . Similarly , TIL domain PIs segregate in three clusters: those abundant during first 48 h of feeding but decrease with feeding in TCA , increase with feeding between 24–96 h in TCB , and those abundant in BD and SD saliva in TCC ( Fig 3B2 ) . In Fig 3B3 , alpha-2-macroglobulins segregate in two clusters: those secreted in abundance between 24–120 h in αCA , and 48-SD in αCB . In Fig 3B4 , cystatins cluster into CCA for those that increase in abundance with feeding and CCB for those that were secreted in high abundance at the 120 h time point . There is evidence that some of the PIs identified in this study regulate important tick feeding functions . For instance serpin EEC19556 . 1 in SCA ( Fig 3B1 ) is 98% identical to AID54718 . 1 , an inhibitor of trypsin and thrombin that also inhibited blood clotting and platelet aggregation [43] . Similarly I . ricinus serpin ABI94056 , the homolog of I . scapularis serpin EEC14235 . 1 in this study ( Fig 3B1 SCB ) is an immunosuppressant , anti-inflammatory , and anti-hemostatic serpin [78–80] . In other studies , I . scapularis cystatin AAY66685 . 1 in this study ( Fig 3B4 CCA ) known as Sialostatin L2 and its close relative Sialostatin L have immuno-modulatory functions , and suppressed cytokine production in absence [81–85] or presence of B . burgdorferi [86] . It will be exciting to understand role ( s ) of PIs in I . scapularis feeding identified in this study . Lipocalins/HBP and FABPs belong to the calycin superfamily of hydrophobic ligand binding extracellular proteins [87–89] . The lipocalin protein family to which HBPs belong is a large group of proteins that bind and transport small hydrophobic molecules , and also associated with multiple functions including regulation of inflammation through binding of pro-inflammation molecules such as histamine [90–92] . Likewise the FABPs bind and transport hydrophobic ligands including long chain fatty acids , eicosanoids , bile salts and peroxisome proliferators [93] . Tick lipocalins/histamine-binding proteins are thought to be involved with mediating the tick's evasion of the host's inflammation defense through sequestration of pro-inflammatory biogenic amines , lipids , histamine , serotonin and prostanoids [94] . Tick histamine binding proteins ( tHBP ) are a subset of lipocalins with two histamine-binding pockets [91] . Of the 18 proteins in S2 Table and Fig 4A , 14 are annotated as tHBPs , three as lipocalins , and one as FABP-like . Similar to other proteins , I . scapularis appears to selectively inject tHBPs/lipocalins into the host at specific time periods , with two tHBPs detected at 24 h in LCC , one at 48 h and three proteins each at 72 h and 96 h in LCD ( Fig 4A ) . The highest numbers of tHBPs/lipocalins were identified at 120 h in LCA ( n = 12 ) of which half were exclusive to this time point ( S2 Table and Fig 4A ) . It is notable that two tHBPs and one each of lipocalin and FABP-like identified in this study were exclusive to SD saliva in LCB , which could suggest that these proteins are involved with events at the end of tick feeding . A limited number of studies suggest that lipocalins/HBPs/FABP indeed perform tick-feeding functions . Three R . appendiculatus tHBPs were predicted to suppress inflammation during blood feeding as revealed by its ability to outcompete histamine receptors [90] . In other studies , D . reticulatus tHBP bound histamine and serotonin [95] , and Ornithodoros . moubata tHBP , referred to as moubatin , demonstrated inhibition of collagen induced platelet aggregation [96] . In a recent study , lipocalins/HBPs/FABPs were identified among 24–48 h A . americanum immunogenic tick saliva proteins [40] suggesting that these proteins are part of the tick saliva proteins that confer anti-tick resistance in repeatedly infested animals . It is notable that in Radulovic et al . , [40] , alongside lipocalins/HBPs , a leukotriene B4-like protease was also found among 24–48 h A . americanum immunogenic tick saliva proteins . It is interesting to note that I . ricinus , tHBP referred to as LIR6 bound leukotriene B4 [97] , a pro-inflammatory mediator and a potent neutrophil chemoattractant . The tick feeding style of tearing up host tissue and sucking up blood from a wounded feeding site exposes the host to microbial infections . From this perspective ticks were postulated to inject anti-microbial peptides into the feeding site to prevent the feeding site from being infected [23 , 48] . Multiple anti-microbial peptides have been characterized in ticks , a majority of which are defensins [98–106] , microplusin/microplusin-like [107–109] and hebreain/hebreain-like [110] . In this study seven of the 15 anti-microbial peptides in S2 Table and Fig 4B are microplusin-like , a single lysozyme , and the rest , are characterized by pathogen-recognition domains ( n = 7 ) . Fig 4B shows three secretion patterns , where ACA proteins were abundant during 24–120 h , ACB were only present in 48 h and ACC proteins increase from 48–96 h but highly abundant in BD and SD saliva . Except for microplusin [107] , which was shown to stop Micrococcus luteus and Cryptococcus neoformans growth [111] , nothing is known on the role ( s ) of most of the anti-microbial peptides in this study . It is notable that majority of anti-microbial peptides in this study are apparently injected into the host within the first 48 h of feeding ( n = 11 ) ( S2 Table and Fig 4B ) . Understanding functions of some of these antimicrobial peptides will reveal microbes that I . scapularis want to keep out of the feeding site . When fully fed , hard ticks are estimated to imbibe host blood that is more than 100 times the their original weight [112] . Catabolism of this huge amount of blood generates high amounts of iron and heme [113–115] . Both iron and heme are needed for normal cell function [113 , 114] . However , if left unsecured , both iron and heme can cause cell damage through promotion of oxidative stress [116 , 117] . Ticks are postulated to prevent iron and heme mediated tick cell damage through expression of iron and heme binding proteins , which play two roles: bind and distribute to cells for normal physiology , and sequester excess iron or heme and prevent oxidative stress triggered cell damage [115] . One of the most notable observations in this study is that although heme-binding proteins represented ~2 . 6% ( 15/582 ) of proteins identified , they accounted for ~11–24% of total protein abundance ( Table 1 and Fig 1 ) . This could suggest that heme metabolism is potentially a “must-not-fail” tick physiological function . The observation that all 15 heme binding proteins in this study are likely injected into the host from within 24–48 h of the tick starting to feed ( S2 Table ) suggests that this mechanism is important from the start of tick feeding . In Fig 4B three secretion patterns are observable: HCA increases in abundance in 120 h-BD proteins , HCB abundant in 48 and 96 h , and HCC abundant in first 48 h but decrease with feeding . It is notable that the five heme binding proteins that were detected at all time points ( S2 Table ) cluster together in HCA ( Fig 4C ) with the exception of EEC13578 . 1 . These proteins account for up to 38% of total NSAF within this class , which could suggest their significance in tick feeding physiology . It is interesting to note that both iron and heme-binding proteins were also detected in high abundance in saliva of D . andersoni [51] , R . microplus [50] , and H . longicornis [52] . However only the latter was detected in this study . Whether or not this is unique to I . scapularis or that iron-binding proteins were injected at below detectable levels needs further investigation . Published evidence has suggested that the tick may detoxify heme/iron through sequestration in digestive cells ( hemosomes ) [118 , 119] and hemolymph [120–122] . Data in this study and others [40 , 41 , 50 , 52 , 123] that show secretion of heme binding proteins in tick saliva suggest a third possibility of eliminating heme through tick saliva . Given that heme has pro-inflammatory functions[124] , secretion of these proteins in tick saliva may be associated with heme sequestration , and thus allowing tick evasion of the host's inflammation defense . Iron sequestration is among the mammalian host's anti-microbial defense . To counter the host's iron sequestration defense , microbes have developed elaborate ways to bind iron from the environment [125–127] and directly uptake heme , which is then digested to release associated iron [128] . From this perspective it is possible that secretion of heme binding proteins is the tick's strategy to make heme available to transmitted pathogens at the tick-feeding site . It is important to note here that B . burgdorferi , the most important I . scapularis transmitted human TBD agent , may not require iron to colonize the host [129] . Tissue injury caused by tick feeding such as disrupting host tissue and then sucking blood from the wounded area will lead to production of reactive oxygen species ( ROS ) , which will in turn damage host tissue and/or transmitted TBD agents [130 , 131] . Thus , it is expected that ticks would inject anti-oxidants into the feeding site as observed in this study . Fig 4D summarizes relative abundance of 36 putative anti-oxidant proteins , 23 of which were identified only in SD saliva ( S2 Table ) , and are likely associated with events toward end of tick feeding . The remaining 13 proteins were identified between 24 h-BD and are likely associated with tick feeding regulation . The heat map in Fig 4D show that different anti-oxidants were detected in high abundance at different time points: ANCA in BD and SD , ANCB at 96 h , ANCC in SD , ANCD at 48 and 72 h , ANCE at 24 h and , ANCF at 120 h . It is interesting to note that some of the data in this study are consistent with previous observations . Glutathione peroxidase ( AAK97814 . 1 ) previously found among immuno-dominant proteins in engorged I . scapularis [132] is among the 23 anti-oxidants that were found in SD saliva only ( S3 Table and Fig 4D ) . The role ( s ) of antioxidants in tick physiology remain mostly unknown . In a recent study , thioredoxin peroxidase gene expression increased in organs of B . burgdorferi infected I . ricinus ticks [133] suggesting involvement in tick and pathogen interaction . It is interesting to note in this study thioredoxin peroxidase protein in non-infected ticks decreased with feeding ( S3 Table ) . It will be interesting to determine if anti-oxidant proteins identified from this study may play roles at the tick-host interface in TBD acquisition and transmission . More than 30% of tick sequences in public databases are of unknown function [30 , 32–37 , 48 , 134–138] . In this study we have identified 129 tick saliva proteins ( TSP ) of unknown function ( S2 Table ) . For clarity secretion profiles of the 112 TSPs of unknown function are summarized in Fig 5A , while the remaining 17 glycine-rich proteins , which are thought to be involved in tick cement formation [139] are shown in Fig 5B . It is interesting to note that in S2 Table , 93 . 7% ( 105/112 ) of TSPs were detected in 24–120 h saliva which could indicate that these proteins are important to tick feeding physiology . The remaining 6 . 3% ( 7/112 ) were exclusive to BD and SD stages and are likely associated with events towards end of feeding . Some proteins were found at one time point: 48 ( n = 12 ) , 72 ( n = 7 ) , 96 ( n = 5 ) and 120 h ( n = 14 ) saliva ( S2 Table ) . More than half ( n = 62 ) of TSPs of unknown function were detected within the first 48 h of feeding . These could be crucial for tick feeding initiation and progression . Patterns in Fig 5A suggest that the tick may potentially selectively inject different proteins into its host every 24 h . In this way , the tick could successfully evade host immunity and acquire a blood meal . Seven clusters ( UCA-UCG ) of TSP of unknown function are observed ( Fig 5A ) . Most notable is that TSP of unknown function that are highly abundant at 24 h ( UCG Fig 5A ) , decrease with feeding indicating that these proteins could serve as pivotal proteins in commencing the tick feeding process . Other secretion patterns include proteins that are abundant at 48 , 72 , 96 , and 120 h in UCE , UCC , UCF , and UCA respectively , as these proteins could be important in maintaining different phases of the tick feeding process . Proteins in UCD and UCA could play important roles towards end of feeding such as in wound healing and detachment from its host or serve as markers for completion of tick feeding . Like several other hard ticks , I . scapularis ticks secrete cement to securely anchor onto host skin during the prolonged tick-feeding period [1 , 139 , 140] . Chemical analysis studies have shown that tick cement has a high content of glycine-rich proteins [139] . On this basis , we speculate that glycine rich proteins in S2 Table could be associated with tick cement formation . The first layer of the tick cement cone is deposited within 5–30 min of the tick attaching , while the second layer starts to form from 24 h post attachment [139] . It is interesting to note that majority ( n = 13 ) of the glycine rich proteins were identified in high abundance in 24 and 48 h saliva ( S3 Table and Fig 5B ) . Secretion patterns of glycine rich proteins shown in Fig 5B suggest that the tick alternates secretion of these proteins during feeding . Most notably the proteins in GCA are most abundant in 24 h , GCB in 48 h , GCC in 96 h , GCD in BD-SD , and GCE in 72–120 h saliva ( Fig 5B ) . The importance of glycine rich proteins detected in abundance towards the end of feeding is unknown at this point . However , there is a possibility for these proteins representing products of degenerated salivary glands . It will be interesting to determine the function of these proteins towards the end of feeding . When subjected to phylogeny analysis , 40 . 2% ( 45/112 ) of TSP of unknown function are unique in that they segregate individually , followed by 7 . 1% ( 8/112 ) that cluster in pairs , and the remaining 52 . 7% ( 59/112 ) segregate in five clusters ( C ) A-E ( Fig 6 ) . According to previously described classifications of I . scapularis proteins [136] , CA , CB , and CD clusters are respectively classified as basic tail ( group 1 , n = 15 ) or tailless proteins ( group 2 , n = 10 ) , GPIIb/IIIa antagonist ( group 9 , n = 7 ) , and 7–9 kDa family ( group 7 , n = 11 ) . TSPs in CC cluster ( n = 7 ) have insulin binding-like proteins motifs [141] , while CE cluster proteins are leucine rich ( n = 9 ) as revealed by sequence inspection . On the basis of amino acid motifs , Ribeiro et . al . , [136] classified basic tail and basic tailless proteins into types I-III . Of the 25 CA proteins , 44% ( 11/25 ) and 20% ( 5/25 ) fit to basic tail types I and II protein respectively , and the remaining 36% ( 9/25 ) fit to basic tailless proteins . An interesting observation from our data is that proteins that segregated together in the ( Fig 6 ) , were identified at different time points ( Fig 7A–7E ) suggesting that the tick could be selectively secreting these proteins during feeding . Similar to Figs 3–5 , we used Z-statistics normalization of NSAF values ( S3 Table ) to develop heat maps in Fig 7A–7E . Basic tail or tailless proteins segregated into five clusters according to secretion patterns starting with the lone protein in BCA that is abundant in SD , followed by proteins in BCB , BCC , BCD and BCE that are respectively abundant in 96 , 48 , 120 , and 72 h saliva ( Fig 7A ) . Likewise in Fig 7B , GPIIb/IIIa antagonist protein cluster in three groups: abundant from 72 h ( GPCA ) , 24 and 48 h ( GPCB ) , and 48 h only ( GPCC ) . In Fig 7C , except for one protein , which is abundant in SD saliva ( ICB ) , majority of these proteins are abundant in 24–72 h saliva ( ICA ) . In Fig 7D and 7E , 7–9 kDa and Leucine rich proteins were identified at variable levels throughout feeding . Putative GPIIb/IIIa in GPCB ( Fig 7B ) cluster are characterized by "RGD" motif and can potentially block platelet aggregation by blocking activated platelets from binding to fibrinogen [142] . In a recent study peptides containing the "NGR" motif prevented resting platelets to bind to fibrinogen [143] . It is interesting to note that four ( AAY66799 , AAY66507 , AAY66621 , and AAY66504 ) basic tail saliva proteins have this motif . Whether or not these proteins can functionally block platelet aggregation of resting platelets needs verification . If functional , these could play key roles in tick feeding success in that at the start of tick feeding , the tick will encounter resting platelets . Surprisingly none of the four NGR motif proteins were detected in 24 h saliva when we expect resting platelets at the feeding site . Interestingly , except for AAY66507 . 1 ( UCE ) detected in 48 and 96 h saliva , the other three were detected at single time points: AAY66799 . 1 at 72 h ( UCC ) , AAY66621 . 1 at 96 h ( UCF ) , and AAY66504 . 1 at 120 h ( UCB ) . Similar to glycine-rich proteins , extracellular matrix proteins likely participate in tick cement formation and/or cell adhesion function . In this study we found 9 extracellular proteins ( S2 Table ) that included cuticle and chitinase-like proteins . Two secretion patterns are observed in Fig 5C , where ECB proteins were abundant from 24–96 h and ECA proteins were abundant in BD and SD saliva . No proteins were detected in 120 h saliva from this class . It is interesting to note that both the active and inactive forms of chitinase were identified in the first 48 h . The former is highlighted by a peritrophin-A chitin-binding domain , which is involved in remodeling the chitinous tick exoskeleton , particularly the mouthpart [144 , 145] . The latter is highly identical to A . americanum tick feeding stimuli responsive acidic chitinase [134] , which when silenced by RNAi caused ticks to loosely attach onto host skin [146] . Blast2seq alignments revealed that the two I . scapularis inactive chitinases ( EEC01936 . 1 and JAB70416 . 1 ) identified in both 24 and 48 h saliva are respectively 64 and 65% identical to A . americanum inactive chitinase ( AIR95100 . 1 ) . Whether or not I . scapularis inactive chitinases serves similar function during tick feeding needs further investigation . In addition to anti-oxidants discussed above , housekeeping-like proteins identified in this study include those associated with metabolism of lipids ( n = 15 ) , carbohydrates ( n = 20 ) , intermediate ( n = 1 ) , energy ( n = 45 ) , nucleotides ( n = 14 ) and amino acids ( n = 20 ) ( S2 Table ) . Others are classified as cytoskeletal ( n = 32 ) , proteasome machinery ( n = 10 ) , protein modification ( n = 49 ) , protein synthesis ( n = 24 ) , protein export ( n = 10 ) , nuclear regulation ( n = 7 ) , signal transduction and apoptosis ( n = 8 ) , transcription machinery , ( n = 8 ) , and transporters and receptors ( n = 19 ) ( S2 Table ) . Cumulative NSAF as an index for protein abundance suggests that majority of housekeeping-like proteins were secreted toward the end of tick feeding in BD and SD saliva , respectively ( Table 1 ) . The tick salivary gland starts to degenerate toward the end of tick feeding and is almost completed within four days of the tick detaching [140 , 147] . Given that most housekeeping genes function inside the cell , one may argue that the high abundance of these proteins in BD and SD saliva may represent progressive SG degradation toward end of tick feeding . However , recent immuno-screening of phage display expression libraries with antibodies to 24 h I . scapularis [41] and 24–48 h A . americanum [40] tick saliva proteins that identified housekeeping-like indicates that secretion of some of the housekeeping proteins starts way before tick salivary gland degeneration , and thus , these proteins likely play important role ( s ) in tick feeding regulation . One remarkable tick adaptation is that although ticks feed from a wounded area in the host's skin , the feeding site is completely healed when ticks complete feeding and detach from host skin . There is a possibility that some of the proteins identified in BD and SD could be associated with speeding up wound healing . It is interesting to note that some cytoskeletal proteins including actin [148–152] , profilin [153 , 154] , alpha tubulin [154] , calponin [155 , 156] , non-muscle myosin [149 , 157 , 158] , thymosin [159] , and tropomyosin [160] identified at high abundance in BD and SD saliva were associated with different aspects of wound healing . Could secretion of these proteins at high abundance be the tick's way to help the host heal ? When ticks feed on blood , they uptake thousands of host proteins . The observation in this study that I . scapularis secreted 83 out of thousands of host proteins suggests that the tick has a mechanism to selectively secrete host proteins in its saliva . Similar to secretion dynamics of tick-derived proteins , the tick appears to selectively secrete different rabbit proteins at different tick feeding time points ( S1 Table ) . It is potentially possible that similar to tick-derived proteins , host proteins in tick saliva perform functions that are unique to different tick feeding phases . Proteins identified from 24/48 h saliva and other time points ( immunity/antimicrobial function , heme/iron metabolism , hemoglobin , nuclear regulation , extracellular matrix , and collagen alpha-1 chain ) , likely aide the tick to feed . On the other hand , proteins identified in BD and SD saliva such as fibrinogen and protease inhibitors are likely associated with events toward the end of tick feeding . For instance , functionally annotated antimicrobial peptides: antimicrobial protein CAP18 [161 , 162] identified in all samples except 120 h saliva , neutrophil gelatinase-associated lipocalin [163 , 164] , neutrophil granule protein [165–167] , protein S100-A12 [168] , neutrophil antibiotic [165] , and lysozyme C [169 , 170] that were identified in 48 h and other stage saliva ( S2 Table ) could aid the tick to clear microbes from the tick feeding site . It is interesting to note that we identified both tick- and rabbit- derived antimicrobial peptides at the same time points . It is most likely that these antimicrobial peptides target different microbes with tick-derived proteins clearing tick-derived microbes , whereas host-derived proteins clear microbes from the host . Cell free hemoglobin ( Hb ) was shown to possess antimicrobial activity through oxidative shock [171 , 172] , and thus there is a possibility that Hb detected in tick saliva could be providing antimicrobial function [173 , 174] at the tick-feeding site . In another study peptides derived from hemoglobin digestion by tick proteases have been described as antimicrobial peptides [173 , 175 , 176] . It is also possible that secretion of Hb could just be an indication of blood meal digestion . Similar to ticks ( Fig 1 ) , rabbit derived heme/iron metabolism associated proteins were the highly abundant at all time points ( Fig 2 ) . A notable difference is that whereas we exclusively identified heme-binding proteins for ticks , we identified a majority of iron binding proteins for rabbits ( n = 6 ) and one heme binding protein ( S2 Table ) . Could this mean that , I . scapularis uses host proteins to remove excess iron though its saliva ? If so , it could be that I . scapularis tick-derived heme binding proteins are responsible for removing heme , but the tick engages host-iron binding proteins to remove excess iron . Except for haptoglobin [177] , which was detected in SD saliva , all other iron binding proteins: serum albumin , histidine rich glycoprotein , lactotransferrin , and serotransferrin as well as the heme binding protein , hemopexin and serum albumin were identified from 24/48 h saliva and other stages during feeding ( S2 Table ) . Based on our data , I . scapularis apparently could use host proteins to eliminate excess iron from the host starting within 24–48 h . It is also interesting to note that human serum albumin was shown to suppress tumor necrosis factor-alpha ( TNF ) and complement component C5a triggered neutrophil respiratory burst [178 , 179] . It is possible the increased concentration of serum albumin at the tick-feeding site could serve other functions . Given that the host uses iron sequestration as the defense mechanism against microbes [180–183] , it is possible that the tick's manipulation of the host to pump back iron into the feeding site could be an adaptation to aide TBD agents to colonize the host , with exception of organisms such as B . burgdorferi , which do not need iron for proliferation [129] . It is interesting to note that in this study we detected rabbit fibrinogen in BD and SD saliva . Fibrinogen is the source for fibrin needed to strengthen the blood clot [184 , 185] . Could it be that the tick pumps back fibrinogen into the host to aide in sealing off the feeding site at the end of tick feeding ? Given that high abundance of keratins are expressed in the skin [186 , 187] , there is a possibility that keratin proteins identified in I . scapularis tick saliva could be due to sample handling or rabbit skin contamination . It is important to note that all keratin types that were identified in I . scapularis tick saliva in this study are associated with different layers of the skin [188 , 189] , and thus there is a high chance we identified remnants on tick mouthparts . However the eight keratin proteins identified in this study represent less than a quarter of the 27 skin keratins [189] . Does the tick selectively inject keratins , and for what purpose is an interesting question for future research . The unique contribution of this study is that , we have for the first time attempted to identify tick- and host- derived proteins that are found in I . scapularis tick saliva every 24 h through the first five days of feeding as well as toward the end of feeding . This study provides identities of I . scapularis tick saliva proteins associated with regulation of: ( i ) early tick feeding events such as tick attachment onto host skin and creating the feeding lesion , which precede tick transmission of TBD agents , ( ii ) slow feeding phase when most TBD agents are transmitted and the tick prepares for rapid feeding phase , and ( iii ) rapid feeding phase when the tick feeds to repletion and detaches from the host . The impact of these data on future in depth tick feeding physiology studies is vast . For instance , transmission of most TBD agents occur at least 36–48 h post tick attachment [190–194] . What happens if we immunize against 24–48 h tick saliva proteins , is TBD agent transmission stopped ? On the other hand we have identified proteins that were apparently secreted at all time points . In future studies , it would be interesting to determine if these proteins regulate "must have" pathways ? It will be interesting to validate the importance of such proteins using the RNAi silencing approach . Some proteins were found at single , two or three time points , could these regulate functions unique to that tick-feeding period ? An interesting recurring pattern observed in these data is that some functionally similar but antigenically unique proteins were identified at different feeding time points . We speculate that this could be the tick's strategy to protect essential pathways from immune response attack . For instance , host immune response against 24 h proteins will not affect functions of functionally similar but antigenically unique proteins at later feeding time points . Essentially the host immune defense against tick feeding will restart every so often , and in the end it will not be effective . In this way key tick feeding physiological functions will continue uninterrupted . Could this mechanism be the tick's equivalent to antigenic variation used by parasites such as Trypanosomes to evade host immunity [195–197] ? What happens if we target as a cluster of functionally similar but antigenically unique proteins that are injected into the host at different time points ? We would like to caution the reader on the inherent limitations of this study . First , in LC-MS/MS approaches , there is a possibility that predominant proteins will mask discovery of lowly expressed but important proteins , and thus the list of I . scapularis tick saliva proteins presented here may not be exhaustive . Second , we sequenced proteins in saliva that was collected by pilocarpine stimulation , and whether or not all detected proteins are secreted under physiological conditions remains to be investigated . However , we are encouraged by our findings that 13% ( 76/582 ) of I . scapularis tick saliva proteins in this study were reported in other tick saliva proteomes and immuno-transcriptome studies ( S4 Table ) . Of the 76 proteins , 12 and 13 proteins were found among tick saliva immunogenic proteins that bound antibodies to 24 h I . scapularis [41] and 24–48 h A . americanum tick saliva proteins [40] respectively . Additionally one protein was identified in I . scapularis nymphs as an immunogenic protein that bound to human serum from exposure to tick bites [198] . The remaining proteins were found in saliva proteomes of R . microplus ( n = 28 , [50] ) , H . longicornis ( n = 22 , [52] ) , D . andersoni ( n = 2 , [51] ) , O . moubata ( n = 5 , [53] , sequencing of I . scapularis tick saliva by Edman degradation ( n = 4 ) [48] , and others were verified as secreted in western blotting studies [42 , 43 , 199–204] . I . scapularis proteins in S4 Table could represent highly conserved tick saliva proteins that regulate important functions , which if disrupted could affect the tick . These proteins could represent priority candidates in future studies . We would like to note that some of the protein sequences in this study are from I . ricinus and other tick species . Majority of these protein sequences have homologs in I . scapularis , which were eliminated as redundancies when we collapsed the local database . I . ricinus proteins in this study represent highly conserved proteins among Ixodes spp ticks .
|
Ixodes scapularis , the blacklegged tick or the deer tick , is among the most medically important ticks that transmit human tick borne disease agents including the spirochaete Lyme disease agent . To develop new tick control methods , a deeper understanding of how the tick feeds is needed . We collected saliva from rabbit fed blacklegged ticks every 24 h through five days of feeding and towards the end of feeding . We used novel proteomic technologies ( Liquid Chromotography–Tandem Mass Spectrometry ) to sequence and identify proteins in tick saliva . The main finding is that ticks apparently selectively inject functionally similar but unique proteins every 24 h , which we speculate is the tick's way to avoid the host's defense to protect important tick feeding functions from host immune system . The host immune responses to proteins present in 24 h I . scapularis saliva will not be effective at later feeding stages . This will influence how to design effective anti-tick vaccine antigens to stop disease agent transmission . This is the first comprehensive study of proteins in blacklegged tick saliva that provides insight into the molecular mechanisms that are at play at the tick feeding site every 24 h .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"and",
"Discussion"
] |
[] |
2016
|
Ixodes scapularis Tick Saliva Proteins Sequentially Secreted Every 24 h during Blood Feeding
|
North African populations are distinct from sub-Saharan Africans based on cultural , linguistic , and phenotypic attributes; however , the time and the extent of genetic divergence between populations north and south of the Sahara remain poorly understood . Here , we interrogate the multilayered history of North Africa by characterizing the effect of hypothesized migrations from the Near East , Europe , and sub-Saharan Africa on current genetic diversity . We present dense , genome-wide SNP genotyping array data ( 730 , 000 sites ) from seven North African populations , spanning from Egypt to Morocco , and one Spanish population . We identify a gradient of likely autochthonous Maghrebi ancestry that increases from east to west across northern Africa; this ancestry is likely derived from “back-to-Africa” gene flow more than 12 , 000 years ago ( ya ) , prior to the Holocene . The indigenous North African ancestry is more frequent in populations with historical Berber ethnicity . In most North African populations we also see substantial shared ancestry with the Near East , and to a lesser extent sub-Saharan Africa and Europe . To estimate the time of migration from sub-Saharan populations into North Africa , we implement a maximum likelihood dating method based on the distribution of migrant tracts . In order to first identify migrant tracts , we assign local ancestry to haplotypes using a novel , principal component-based analysis of three ancestral populations . We estimate that a migration of western African origin into Morocco began about 40 generations ago ( approximately 1 , 200 ya ) ; a migration of individuals with Nilotic ancestry into Egypt occurred about 25 generations ago ( approximately 750 ya ) . Our genomic data reveal an extraordinarily complex history of migrations , involving at least five ancestral populations , into North Africa .
The census size of Mediterranean North Africa exceeds 160 million people [1] , but relatively little is known about the genetic makeup of these populations and the demographic history of migration between North Africa and neighboring regions . Mediterranean North Africans are often grouped with Near Eastern populations because populations in both regions speak primarily Afro-Asiatic languages , like Arabic , and phenotypic attributes , like lighter skin pigmentation , differentiate many North Africans from sub-Saharan Africans . Recently , geneticists have attempted to replicate disease associations identified in Europeans and Near Eastern groups with North African populations , reflecting a hypothesis of shared genetic ancestry , with mixed results [2]–[5] . In this paper , we present analysis of autosomal single nucleotide polymorphism ( SNP ) array data for seven North African populations ( see Materials and Methods ) , distributed along an east-to-west transect across the continent . We clarify the population structure of North Africa and explicitly interrogate the history of gene flow into North Africa from the Near East , Europe and sub-Saharan Africa . Prior genetic studies , largely from uniparentally inherited markers , have not resolved the location origin of North African populations or the timing of human dispersal ( s ) into North Africa . Analyses based on the frequencies of a small number of autosomal genetic polymorphisms and uniparental markers have shown that the genetic landscape follow an east-west pattern with little to no difference between Berber- and Arab-speaking populations [6] , [7] . Mitochondrial data , for example , indicate an early back-to-Africa migration [8]–[10] , but Y-chromosome markers largely support a Neolithic expansion and historic period gene flow throughout the Mediterranean [11] ( though see [12] ) . Do current North Africans retain genetic continuity with the first modern human occupants of northern Africa from more than 50 , 000 years ago ( ya ) or was northern Africa primarily repopulated during the Holocene by herding and farming populations from elsewhere ? Evidence of Neolithic migration from the Near East is supported by the introduction of domestic animals like cows , sheep and goats to North Africa . But the indigenous development of ceramics in Saharan Africa by 9 , 000 ya is also suggestive of an incipient form of agriculture or pastoralism , prior to any demic diffusion from the Near East [13] . Less controversial is the observation that many North African populations now speak Arabic and that this language shift occurred primarily after the Arabic conquest 1 , 400 ya . This Arabic shift is well documented , but it remains unknown how deeply recent migrations ( <2 , 000 ya ) from the Arabian and Iberian Peninsulas shaped the genetic diversity of current North African populations . In addition , sub-Saharan influence has been detected in North African samples by all types of genetic markers analyzed , although it is unknown how recent this gene flow might have been [14]–[16] . Initial autosomal SNP analysis of the Algerian Mozabites indicated they carry ancestry from Europe , the Near East and sub-Saharan Africa; neighbor-joining phylogenetic analysis suggested that Mozabites branch off with Out-of-African populations , but are an outgroup to all Near Eastern populations in the Human Genome Diversity Panel ( HGDP-CEPH ) [17] . In short , the origins of North African populations and the number of subsequent migrations from neighboring regions have been poorly resolved . Genomic models of admixture in human populations have largely been confined to cases of two-way admixture such as African-American [18]–[20] and some Hispanic-Latino groups such as Mexican-Americans [21]–[23] . However , a two-population model is likely inappropriate for North African populations ( as it is for some Caribbean groups such as Puerto Ricans [24] ) given multiple putative migrations proposed in earlier studies . Moreover , while African-Americans and Hispanic/Latino have ancestries from highly divergent source populations , North Africans may have ancestry from more closely related populations , for example Europeans and Near Easterners . We extend a principal component analysis-based ( PCA ) method of local ancestry assignment [20] in order to allow for three possible ancestral populations . With haplotypes for various ancestries inferred from PCA-based assignment , we model the time and mode of migrations from neighboring regions into North Africa .
In order to characterize population structure across North Africa , we combined our genotype data for the seven North African populations with population samples from western Africa , eastern Africa , Europe and the Near East ( see Materials and Methods ) . A representative subset of these samples is displayed in Figure 1 and Figure 2 . We applied both classic multidimensional scaling ( MDS ) with an LD-reduced set of 280 K SNPs on the identity by state ( IBS ) matrix and an unsupervised clustering algorithm , ADMIXTURE [25] , to explore patterns of population structure . In ADMIXTURE , we explored k = 2 through 10 ancestral populations to investigate how assumptions regarding k impact our inference of population structure in North Africa . Log likelihoods for successively increasing levels of k continue to increase substantially as k increases ( Figure S1B ) . However , visualization of k = 10 indicates that very high order clusters pulled out related individuals in the Tunisian Berber sample ( Figure S1 ) ; for this reason we focus on k = 2 through 8 . Our North African population samples are clearly differentiated from other African populations ( Figure 1 , Figure 2 ) . MDS component 1 separates sub-Saharan Africans from populations that currently reside outside of Africa ( OOA ) , and the North African populations cluster closest to the Near Eastern Qatari . A subset of individuals are intermediate between the North and sub-Saharan African samples ( Figure 2 , Figure S2 ) . At k = 2 ( ADMIXTURE ) , 80% of the ancestry in North African individuals is assigned to a cluster defined by its maximum frequency in Near Eastern and European populations ( Figure 1 ) . MDS component 2 differentiates western from eastern sub-Saharan Africans ( Figure 2A ) . MDS component 3 differentiates populations thought to have a high degree of autochthonous ancestry ( i . e . Tunisian Berbers and Saharawi ) from populations outside of Africa . Interestingly , the MDS component 3 appears to be largely independent of the amount of sub-Saharan ancestry ( Figure 2B ) and North Africans are dispersed along the MDS component 3 axis , with the Tunisian Berbers occupying the extreme end of this gradient . A gradient also appears in the higher k ancestral population plots of the ADMIXTURE analysis ( Figure 1 ) . Assuming 4 or more ancestral populations ( k = 4 through 10 , Figure S1 ) there is a cline of putative autochthonous North African ancestry decreasing in frequency from Western Sahara eastward to Egypt . We refer to this North African ancestral component as the “Maghrebi” throughout the remainder of the paper , reflecting the primary geographic distribution of this ancestry in the Maghreb: West Sahara , Morocco , Algeria and Tunisia . The west-to-east decline in Maghrebi assignment is only interrupted by the Tunisian Berbers , who are assigned nearly 100% Maghrebi ancestry . The Tunisian Berbers further separate as a distinct population cluster at k = 8 . An opposite cline of ancestry appears to originate in the Near East ( i . e . Qatari Arabs ) and decreases into Egypt and westward across North Africa ( k = 6 , 8 ) . At k = 6 through 8 , all North African populations except for Tunisians have sub-Saharan ancestry , present in most individuals , though this ancestry varies between 1%–55% . Interestingly , eastern populations ( i . e . Libya and Egypt ) share ancestry assigned to both the Bantu-speaking Luhya and the Nilotic-speaking Maasai , whereas western populations share ancestry mainly with the Luhya . Of note is that the South Moroccan and western Saharan populations contain considerable variation across individuals in the amount of sub-Saharan ancestry ( see also [14] , [26] ) , consistent with recent admixture . We estimate Fst by comparing each of the North African populations to the Tuscans and Qatari respectively . Estimates range between 0 . 035–0 . 063 ( Figure 3 ) . In order to quantify population divergence among these groups , we use the relationship between Fst and the effective size “Ne” to estimate the divergence time “t” ( see Materials and Methods ) . Since this model neglects migration , we expect our results to form a lower bound on the population divergence time , as similar levels of population divergence would require a longer separation in the presence of migration . Additionally , the model assumes that populations have had similar demographic histories ( i . e . if there was a bottleneck , all populations were affected equally ) ; as all populations derive the majority of their ancestry from an Out-of-Africa ancestral population , and the OOA bottleneck is the primary signature in OOA populations , we believe this assumption is valid . Estimates for population Ne were taken from Li et al . [17] . All estimates of population divergence between the North Africans and the European/Near Eastern samples predate the Holocene . We then attempted to obtain more accurate estimates of divergence time by controlling for recent migration . We calculated a second set of Fst estimates using cluster-based allele frequencies from ADMIXTURE among the Maghrebi , European and Near Eastern ancestries , when we considered higher order k = 5∶8 ancestral clusters . As indicated in Figure 3 , population divergence between the Maghrebi and the European and Near Eastern populations occurred between 18 , 000–38 , 000 ya . The bounds here represent variation in ancestral k estimates and assumptions regarding Ne , as Near Eastern populations have a greater estimated Ne than European . Although these divergence time estimates may not be precise , as they do not adequately model ancient migration , they do suggest that the population divergence between the ancestral Maghrebi population and neighboring Mediterranean populations occurred at least 12 , 000 ya and indeed more likely predated even the Last Glacial Maximum . Given the complex patterns of admixture apparent from population structure analyses , we asked if populations differed in the proportion of DNA that individuals within populations shared identically . We estimate the amount of DNA shared identically by descent ( IBD ) using the GERMLINE software [27] , with a 5 cM threshold to eliminate false positive IBD matches . The estimated cumulative amount of IBD between pairs of individuals within each population is illustrated in Figure 4 with the Tunisians , Saharawi , and North Moroccans . Most of our North African populations shared little IBD or displayed an exponential-like decline in the cumulative amount of IBD , indicating that the great majority of individuals in these populations were only distantly related ( i . e . had less IBD than predicted in third cousins ) . However , the Tunisian Berber population displayed an excess of pairs of individuals sharing 200–1200 cM IBD . This bimodal distribution indicates that many 1st and 2nd cousin genetic equivalent pairs were present in this sample , even though donors declared themselves to be unrelated during the sampling process . Analysis of long runs of homozygosity ( ROH ) indicate that the Tunisian population averaged almost twice as much of their genome is in ROH than other North African populations , 230 Kb versus 120 Kb respectively ( Figure S3 ) . The pattern of ROH and pairwise IBD in the Tunisian Berbers is likely the result of endogamy due to geographic isolation or cultural marriage preferences . Our cluster-based analysis identified five distinct ancestries in North Africa that we refer to as: Maghrebi , European , Near Eastern , eastern and western sub-Saharan Africa . In order to test whether sub-Saharan African ancestry was an ancient or recent migration signature , we considered the length of sub-Saharan haplotypes . First , to assign local , ancestry-specific haplotypes across a genome , we implemented a new principal component-based admixture deconvolution approach ( PCADMIX ) for three ancestral populations ( see Materials and Methods , Figure S4 ) [20] , [28] . We focus on admixed populations at either end of North Africa , specifically our population samples of South Moroccans and Egyptians . PCADMIX requires predefined ancestral groups . For this purpose , we assume South Moroccans have ancestry from three primary sources: Maghrebi ancestry ( e . g . Saharawi ) , eastern Bantu-speakers ( e . g . Luhya ) and European ( e . g . Spanish Basque ) ( Figure 5A ) . We similarly assume Egyptians have ancestry from four primary source populations: Maghrebi ( e . g . Saharawi ) , eastern Nilotic-speakers ( e . g . Maasai ) , Near Eastern Arabs ( e . g . Qatari ) and European ( e . g . Spanish Basque ) . These source populations reflect the ancestry assigned in our clustering algorithm analysis ( Figure 1 ) . According to our ADMIXTURE results , two distinct sub-Saharan ancestries are present in Egyptian individuals at k = 6∶10; these two ancestry components are highest in the Kenyan Luhya and Maasai populations . However , the “Luhya” ancestry is present at very low proportions , below 10% at k = 6 and below 5% at k = 8 and there is also “Luhya” ancestry detectable in Maasai populations . Thus , we chose the Maasai as the best ancestral sub-Saharan population for extant Egyptians . If our choice of source populations for an admixed individual is accurate ( i . e . the source populations are reasonable representatives of an ancestral population ) we expect similar estimates of ancestry proportions between PCADMIX-ADMIXTURE when ancestry in PCADMIX is assigned with a posterior probability threshold of 0 . 8 ( , Figure 5 , Figure S5 ) . We used the Saharawi as our proxy Maghrebi population , since the high relatedness in the Tunisian samples is likely to cause reduced ability to infer Maghrebi tracts in more diverse populations . Our sample of Tunisian Berbers retains the highest amount of Maghrebi ancestry , without substantial evidence of admixture with sub-Saharan , European or Near Eastern populations . However , their bimodal mean IBD distribution ( Figure 4A ) indicates a high proportion of 1st–2nd cousin equivalents and suggest that our sample of Tunisian Berbers comes from an isolated , endogamous population with diversity that is likely reduced relative to other Maghrebi populations . Thus , although their low degree of non-Maghrebi admixture might make them ideal as a Maghrebi source population , reduced haplotypic diversity means that we are likely to under-call true Maghrebi segments from other , more diverse populations . This expectation was borne out in our PCA-based admixture deconvolution of southern Moroccans when comparing Tunisian versus Saharawi as a Maghrebi source population ( Figure S6 ) . We note that when using either the Tunisian Berbers or the Saharawi , the Maghrebi component in other individuals ( e . g . Egyptian , South Moroccan ) tended to be underestimated in comparison to the ADMIXTURE proportions ( Figure 5A , Table S2 ) . We also infer independent admixture proportions in the Algerian , South Moroccan and Saharawi samples by running LAMP [29] to estimate local ancestry using 3 source populations: Tunisians , Basque and Luhya; with LAMP we also observe a likely excess of inferred European ancestry in the Algerian , South Moroccan and Saharawi samples ( 7 ) . The length of tracts assigned to distinct ancestries in an individual is informative regarding the time and mode of migration from one ancestral population into another . After a migrant chromosome enters a population , the length of the migrant ancestry tract is broken down over time due to the process of recombination . We use a maximum likelihood approach developed by Pool and Nielsen [30] to estimate the time of change in migration rate between populations based on the length and number of migrant tracts in the absorbing sink ( or “admixed” ) population . We first consider a continuous migration model where migration occurs at a constant rate from T generations ago to present day . We assume that there has been no migration between the source and sink populations prior to the initial time of a migration into the admixed population . We tabulate the number of migrant tracts in the Moroccans and Egyptians , where each migrant tract has a posterior probability >0 . 8 . To reduce biases due to our lower sensitivity to short tracts , we only modeled tracts longer than 3 cM , and considered assigned tracts with posterior probability >0 . 8 . With a 3 cM cutoff we expect to capture 50% of tracts from 55 generations ago and 10% of tracts from 130 generations ago ( see Materials and Methods ) . Unassigned short tracts ( i . e . the “undecided” regions , Figure 5B and Figure 6 ) within a long continuous migrant segment can be artificially shortened by spikes of low posterior probability . Unassigned tracts that were situated within a tract of one ancestry and which maintained a posterior probability >0 . 5 for the same neighboring ancestry were considered to as one long ancestry tract . We focus on the sub-Saharan African migrant tracts in South Moroccans ( Figure 5B ) and Egyptians ( Figure 6 ) . These tracts tend to be highly diverged from other ancestries in the population ( Fst>0 . 10 ) and populations with similar divergence resulted in accurate haplotype assignment in prior testing [28] . Under a model of constant migration from the Bantu-speaking Kenyans and southern Moroccans started about 41 generations ago ( ga ) ( 95% CI: 39–44ga ) assuming there was no migration occurring prior to this period . The confidence interval calculations , obtained by resampling sub-Saharan migrant tracts with replacement , do not take into account possible biases caused , for example , by the model assumption of a fixed migration rate . We hypothesized that the distribution of sub-Saharan African tracts in the Moroccans and Egyptians might better reflect a single episode or “pulse” of migration . In order to test this hypothesis , we modify Pool and Nielsen's [30] approach to conform to a pulse model ( see Materials and Methods ) . We compared the log likelihoods summed over all migrant tracts under constant and pulse migration models for each population maximized over the relevant parameters , and present the model with the higher log likelihood ( Table 1 ) . Estimates of the time of migration are more recent under a pulse model . The younger estimate occurs because the model fit must account for relatively long migrant tracts in the data; under a constant migration model these tracts represent recent migrants , but for a single episode of migration , long tracts can only be accounted for by recent migration of the entire sample . In order for the average migrant tract length to be equal in the two models , migration must have started more than twice as long ago in the constant migration model compared to the pulse model ( Table 1 ) . Our Egyptian sample of Nilotic segments ( derived from Maasai ) has a better log likelihood under a pulse migration model , estimated as time since admixture of 24ga ( 95% CI: 23–26ga ) rather than 51ga under a constant migration model ( Table 1 ) .
By sampling multiple populations along an approximate transect across North Africa , we were able to identify gradients in ancestry along an east-west axis ( Figure 1 and Figure 2 ) . Notably , even northwestern populations with very high proportions of Maghrebi ancestry , such as the Tunisians and Saharawi , still cluster with Out-of-Africa populations in the population structure analyses ( Figure 1 ( k = 2 ) , Figure 2 ) . This observation of clustering formed the basis for further analyses to distinguish between two alternative demographic models . First , North Africans could be closer to OOA populations due to extensive gene flow , likely from the Near East , over the past ∼50 Kya . Second , North Africans could be closer to OOA populations if the two groups had diverged more recently than either had split with sub-Saharan Africans . We can reject a simple model of long-term continuous gene flow between the Near East and North Africa , as evidenced by clear geographic structure and non-zero Fst estimates . Fst estimates between the inferred Maghrebi cluster and sub-Saharan Africans are two to three-times greater than Fst between the Maghrebi and Europeans/Near Easterners ancestral clusters ( Table S3 ) . We then address whether this population structure was recent or ancient . Although Fst estimates from ascertained data may be biased , as rare alleles are under-represented in the site frequency spectrum , comparison of African-European Fst from resequencing data and the Affymetrix 500 K platform showed only a negligible difference [31] . Assuming reasonable effective population sizes for North African Maghrebi and neighboring populations [17] , we first showed that all North African populations are estimated to have diverged from OOA groups more than 12 , 000 ya ( Figure 3 ) . After accounting for putative recent admixture ( Figure 1 ) , the indigenous Maghrebi component ( k-based ) is estimated to have diverged from Near Eastern/Europeans between 18–38 Kya ( Figure 3 ) , under a range of Ne and k values . We hence suggest that the ancestral Maghrebi population separated from Near Eastern/Europeans prior to the Holocene , and that the Maghrebi populations do not represent a large-scale demic diffusion of agropastoralists from the Near East . With model parameters for divergence approximately estimated , we then ask whether North African ancestral populations were part of the initial OOA exit and then returned to Africa [8] , or if an in situ model of population persistence for the past 50 Kya is more likely ( with variable episodes of migration from the Near East ) ? We can address this question only indirectly with contemporary samples; however , several auxiliary observations point toward the former hypothesis . Substantially elevated linkage disequilibrium in all of these North African population samples , compared to sub-Saharan populations [32] , is consistent with a population bottleneck . Hellenthal et al . [30] also observed that the reduction in the number of haplotype founders required to reconstruct the Mozabite population , as compared to other African populations , could be explained by a population bottleneck . If North African ancestral populations persisted in situ , then we need to invoke two population bottlenecks , one in the ancestors of North Africans ( including the Berbers ) and one for OOA groups . Alternatively , the “OOA” bottleneck would need to occur in North Africa , rather than when groups moved out of the continent [33] . The second possibility appears at odds with most published models of the movement of modern humans outside of Africa . A scenario where North African Maghrebi ancestry is the result of in situ population absorbing Near Eastern migrants would likely need the following premises to explain the results here and elsewhere: a ) an Out-of-Africa migration [concurrent with bottleneck] occurs 50–60 Kya , geographically dividing North African and Near Eastern populations; b ) North Africans experience a separate bottleneck; c ) gene flow maintains similarity between the two geographically distinct populations; d ) the gene flow then ceases or slows roughly between 12–40 Kya in order to allow sufficiently distinct allele frequency distributions to form . In contrast , we find it more parsimonious to describe model where: a ) an OOA migration occurs [concurrent with a bottleneck]; b ) OOA populations and North Africans diverge between 12–40 Kya when a migration back-to-Africa occurs . These models should be further tested with genomic sequence data , which have better power to detect magnitude and timing of bottlenecks , and to estimate the true joint allele frequency spectrum . More recently , the substantial , east-to-west decline of Near Eastern ancestry ( Figure 1A ) could represent a defined migration associated with Arab conquest 1 , 400 ya or merely gene flow occurring gradually among neighboring populations along a North African | Arabian Peninsula transect . Although we observe a declining amount of Maghrebi ancestry from northwest-to-northeast , it is possible that other geographically North African samples ( e . g . Egyptians further south than the sampled Siwa Oasis ) do not conform to this geographic cline . Finally , we also observe European ancestry that is not clearly accounted for by the inclusion of a Near Eastern sample . Additional migration coming from Europe might be plausible , though the origin and the period where it took place cannot be determined with the present data . The less than 25% European ancestry in populations like Algerians and northern Moroccans could trace back to maritime migrations throughout the Mediterranean [34] . Alternatively , the Qatari could represent a poor proxy for an Arabic source population , causing additional diversity to be assigned European ( e . g . European ancestry tracts were not reliably assigned as European with PCADMIX ) . In summary , although paleoanthropological evidence has established the ancient presence of anatomically modern humans in northern Africa prior to 60 , 000 ya [35] , the simplest interpretation of our results is that the majority of ancestry in modern North Africans derives from populations outside of Africa , through at least two episodes of increased gene flow during the past 40 , 000 years ( Figure 1 , Figure 2 , Figure 3 ) . Multiple local ancestry assignment methods , including PCADMIX , require thinning genotype datasets to remove alleles in high linkage disequilibrium between populations [29] , [36]; this step discards information regarding haplotype patterns that tend to be more informative than genotypes when using data biased by SNP ascertainment [37] . HAPMIX incorporates both LD information and uncertainty in phase inference for haplotypes [18] , but the software is currently limited to a two-population model . Our ancestral proportions of European and sub-Saharan ancestry for many North Africans at k = 2 ( Figure 1 ) are similar to those obtained with HAPMIX by Price et al . [18] for the HGDP Algerian Mozabites , assuming a two-population mixture of northern Europeans and Yoruba . However , our results show that increasing the number of possible ancestral populations reveals multiple , diverse ancestries ( e . g . Maghrebi , Near Eastern , Nilotic ) and that the proportion of sub-Saharan African assignment decreases as these other ancestries are accounted for . This decrease in assigned sub-Saharan ancestry in North African samples , from a k = 2 model , is consistent with an interpretation that Maghrebi or Near Eastern diversity that is not present in the panel populations is more likely to be assigned to the more diverse , Sub-Saharan African ancestry . Using a two-population admixture model , Price et al . [18] estimated the time of migration from sub-Saharan Africa into the Mozabites to have begun about 100 generations ago ( or more ) . Our results suggest that sub-Saharan African and Maghreb admixture is considerably more recent , 24–41 generations ago ( and even the upper 95% CI estimate under either model is 55ga , Table 1 ) . The discrepancy between these two estimates may result from our incorporation of multiple source populations , our use of non-linear models to estimate migration timing and the elimination , in Price et al . [18] , of individuals with megabase long African segments . We use a two-population model of migration where we measure the number and length of migrant tracts observed in the admixed population . However , as argued earlier , North African populations have absorbed migrants from multiple episodes of migration . We use three- and four-population admixture deconvolution to identify the tracts from these separate migrations . One complication with this approach is the possibility that source populations that contribute migrants to North Africa are themselves exchanging migrants . For example , Near Eastern populations expanded into European continent during the Neolithic , and even an isolated population like the Spanish Basque may have genomic segments that trace back to the Neolithic expansion [38] , [39] . In this case , estimation of the time of migration of Arabic individuals into North Africa would be biased by Basque segments of Arab ancestry that were contributed by Europeans , but are locally assigned to Arabic ancestry . We confine our migration estimates to those from sub-Saharan populations into North Africans because there has likely been relatively little recent gene flow between sub-Saharan Africans and the European/Near Eastern populations . Moorjani et al . [40] present evidence for recent gene flow ( <100 generations ago ) from Africa to the Near East and Europe . But , they hypothesize it might be due to North African migrations , rather than sub-Saharan Africa . Assuming a 30-year generation time [41] , the proposed migration of sub-Saharans to southern Morocco at about 1 , 200 years ago coincides with the rise of the Ghana Empire , involved in the trans-Saharan slave trading , and the “Great Berber Uprising” which established Berber kingdoms throughout Morocco . We use a Bantu-speaking population from Kenya as a source population for this migration , as North African individuals with sub-Saharan ancestry appeared to be closer to the Luhya than the Nigerian Yoruba ( Figure 1 , Figure 2 and Figure S2 ) . However , there are likely other western African populations genetically similar to Kenyan Bantu-speakers . We do not interpret this association as an explicit migration from Kenya to southern Morocco . We also use the length of Nilotic tracts in Egyptians to ask if sub-Saharan ancestry ( apparent in Figure 1 and Figure 6 ) also appears to be a recent introduction . Under a pulse model of migration , a significant increase in gene flow likely occurred ∼700 ya , after the Arabic expansion into North Africa 1 , 400 ya . Our migration results are in agreement with previous studies based on mtDNA analysis where gene flow into eastern and western North Africa appeared to have different sub-Saharan population sources [10] , [16] . Our genome-wide dense genotyping data from seven North African populations allow us to address outstanding questions regarding the origin and migration history of North Africa . We propose that present-day ancestry in North Africa is the result of at least three distinct episodes: ancient “back-to-Africa” gene flow prior to the Holocene , more recent gene flow from the Near East resulting in a longitudinal gradient , and limited but very recent migrations from sub-Saharan Africa . Population structure in North Africa is particularly complex , and future disease or phenotypic studies should carefully account for local demographic history . However , the rich history of gene flow can also help empower genome-wide association mapping via admixture mapping techniques [42] . For example , the variable but relatively long haplotypes of sub-Saharan ancestry are amenable to admixture mapping approaches developed for African-American samples . In conclusion , North African populations retain a unique signature of early “Maghrebi” ancestry , but North African populations are not a homogenous group and most display varying combinations of five distinct ancestries .
A total of 152 individuals representing seven different North African locations and the Basque Country were included in the present study . Informed consent was obtained from all of them . Samples were genotyped on the Affymetrix 6 . 0 chip , and after quality control filtering for missing loci and close relatives , 125 individuals remained: 18 from North Morocco , 16 from South Morocco , 18 from Western Sahara , 19 from Algeria , 18 from Tunisia , 17 from Libya and 19 from Egypt . Further information on the samples may be found in Table S1 . Moreover , 20 individuals from the Spanish Basque country were included in the analysis . Data are publicly available at: bhusers . upf . edu/dcomas/ . In order to study the population structure and the genetic influence of migrants in the region a database was built including African and European populations from HapMap3 [43] , western Africa [20] , and 20 Qatari from the Arabian Peninsula [44] as Near Eastern representatives . Written informed consent was obtained from the participants and analyses were performed anonymously . The project obtained the ethics approval from the Institutional Review Board of the institution involved in the sampling ( Comitè Ètic d'Investigació Clínica - Institut Municipal d'Assistència Sanitària ( CEIC-IMAS ) in Barcelona , Spain ) . An unsupervised clustering algorithm , ADMIXTURE [25] , was run on our seven new North African populations , Spanish Basque , Near Eastern Qatari , western Africans , HapMap3 Kenyan Luhya , Maasai and Italian Tuscans . Nine ancestral clusters ( k = 2 through 10 ) in total were tested successively . Log likelihoods for each k clusters are available in Figure S1B . Fst based on allele frequencies was calculated in ADMIXTURE for each identified cluster at k = 8 . Given the high heterogeneity in Qatari population , we present individuals with the lowest sub-Saharan , European and North African ancestries and higher Near Eastern ancestry , based on ADMIXTURE . Multidimensional scaling ( MDS ) was applied to the pairwise IBS Matrix of 279 , 528 SNPs using PLINK 1 . 07 software [45] . The top three MDS components were plotted together using R 2 . 11 . 1 . Population divergence estimates from the cluster-based allele frequencies from ADMIXTURE ( k = 5–8 ) were obtained using [46]:The cluster-based allele frequencies will be less biased by recent migration between populations . Estimates of population divergence , though potentially older if migration is unaccounted for in the Fst estimate , are unlikely to be younger if the range of Ne sizes is realistic . In previous work , imputation accuracy was tested in a sample of Algerian Mozabites and other populations from the Human Genome Diversity Project ( HGDP-CEPH ) [37] . Among all the African populations , the Mozabites had the poorest imputation accuracy when the sub-Saharan Yoruban sample was used to predict allele states [37] . For this reason , we used multiple populations for phase inference . North African , Qatari and Basque genotypes were phased using BEAGLE 3 . 0 software [47] . Phased haplotypes from three HapMap3 populations ( i . e . Maasai , Yoruba , and Tuscans ) were used as seeds for haploype inference; each HapMap3 population was randomly sub-sampled for 30 individuals each in order to prevent over-representation of haplotypes from a single geographic region . The Basque , Qatari and all North African populations were phased with the same three seed populations to prevent discrepancies based solely on different haplotype seeds . We estimate the amount of DNA shared identically by descent ( IBD ) using the GERMLINE software [27] , with a 5 cM threshold to eliminate false positive IBD matches . All 5 cM or greater segments shared IBD between pairs of individuals were summed , and histograms created for sharing within each North African population . Local ancestry was assigned with a new PCA-based method , PCADMIX . This method uses phased genotype data ( i . e . , haplotypes ) to determine exact posterior probabilities along each chromosome . PCADMIX relies on Principal Components Analysis ( PCA ) to quantify the information that each SNP contributes to distinguishing the ancestry of a genomic region . PCADMIX is publicly available at sites . google . com/site/PCADMIX . We use Singular Value Decomposition in R to perform PCA on the phased genotypes of the ancestral representatives . We project admixed individuals on the basis of principal components , and compute the observed ancestry “score” for a haplotype i in the jth window as the weighted average Ljgij , where gij is a column vector of the haplotype's alleles ( coded as 0 or 1 ) in window j , standardized by the mean and standard deviation of that SNP's frequency in the ancestral populations . Lj is a matrix for which the entry in the kth row , lth column is the loading of SNP l in the window on principal component k . We use a forward-backward algorithm to identify the probability of ancestry at each window , conditional on the ancestry scores . For the forward-backward algorithm in our HMM , we used a haploid version of the transition and emission probabilities in the Viterbi algorithm of Bryc et al . [20] . The transition probability is defined by p , the probability of recombination between windows , and qj , the frequency of the target population's chromosomes in the admixing ancestral pool . First , ancestral populations are thinned for SNPs with r2<0 . 8 in order to remove highly linked alleles from different populations , which can lead to spurious ancestry transitions . Second , chromosomes for each individual in a population are artificially strung together to create two haploid genomes for the individual; this step increases the amount of information used for PCA , and it is of special relevance given that Europeans , Near Easterners and North African are differentiated with an Fst of only ∼0 . 05 . Then , PCA on a number k≤3 of ancestral populations is performed and the admixed population is projected into the determined k≤3 PCA space . PC loadings are used as weights in a weighted average of the allele values in a window of 40 SNPs . These window scores are then used as observed values in a HMM to assign posterior probabilities to the ancestry in each window ( where chromosome were considered separately ) . Information on using PCADMIX in Egyptians is available in Figure S8 . Additional performance testing and details of the implementation for this approach are available in [28] , Texts S1 , S2 , S3 and Figure S9 . We tabulated the length and number of genomic tracts ( i . e . phased haplotypes ) assigned to particular population ancestries for the South Moroccan and Egyptian population samples ( see above for PCA-based local ancestry assignment ) . We used a posterior probability threshold of 0 . 8 , optimized for concordance with ADMIXTURE ancestry proportions ( Figure 5A ) . The maximum likelihood estimate of the time of migration is sensitive to the minimum detectable length of migrant tracts . That is , as migrant tracts recombine with non-migrants and become smaller in size , we are less likely to detect them . Histograms of the cumulative number of migrant tracts of different lengths , for all individuals , were visualized ( Figure S10 ) and we observe a reduction in the number of short migrant tracts in the 0 . 5 to 1 . 5 cM bins , inconsistent with constant or punctual migration model . Rather , this reduction can be understood as a reduction in our ability to detect short migrant segments due to insufficient SNP density or haplotype variation that is not present in our source population . We therefore choose a 3 cM threshold as the minimal length of migrant tracts to be considered . Theoretically , under an isolation followed by migration model and with a 3 cM tract length threshold , we have power to detect relatively recent migrations occurring within the past generations [30] . We modify Pool and Nielsen [30] equation 10 , with for the likelihood that a segment is of length Morgans given that it is longer than the cutoff length in a model with constant migration rate starting at time in a chromosome of length . Similarly , we estimated a likelihood of for punctuated migration occurring generations ago , which neglects chromosomal edge effects , an approximation justified by the fact that for a large majority of tracts .
|
Proposed migrations between North Africa and neighboring regions have included Paleolithic gene flow from the Near East , an Arabic migration across the whole of North Africa 1 , 400 years ago ( ya ) , and trans-Saharan transport of slaves from sub-Saharan Africa . Historical records , archaeology , and mitochondrial and Y-chromosome DNA have been marshaled in support of one theory or another , but there is little consensus regarding the overall genetic background of North African populations or their origin and expansion . We characterize the patterns of genetic variation in North Africa using ∼730 , 000 single nucleotide polymorphisms from across the genome for seven populations . We observe two distinct , opposite gradients of ancestry: an east-to-west increase in likely autochthonous North African ancestry and an east-to-west decrease in likely Near Eastern Arabic ancestry . The indigenous North African ancestry may have been more common in Berber populations and appears most closely related to populations outside of Africa , but divergence between Maghrebi peoples and Near Eastern/Europeans likely precedes the Holocene ( >12 , 000 ya ) . We also find significant signatures of sub-Saharan African ancestry that vary substantially among populations . These sub-Saharan ancestries appear to be a recent introduction into North African populations , dating to about 1 , 200 years ago in southern Morocco and about 750 years ago into Egypt , possibly reflecting the patterns of the trans-Saharan slave trade that occurred during this period .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genome",
"evolution",
"genome",
"scans",
"population",
"genetics",
"gene",
"pool",
"genome",
"analysis",
"tools",
"effective",
"population",
"size",
"population",
"biology",
"genetic",
"polymorphism",
"biology",
"evolutionary",
"genetics",
"genetic",
"drift",
"haplotypes",
"genetics",
"gene",
"flow",
"genomics",
"evolutionary",
"biology",
"evolutionary",
"processes",
"genetics",
"and",
"genomics",
"human",
"genetics"
] |
2012
|
Genomic Ancestry of North Africans Supports Back-to-Africa Migrations
|
Ecological and laboratory studies have demonstrated that temperature modulates West Nile virus ( WNV ) transmission dynamics and spillover infection to humans . Here we explore whether inclusion of temperature forcing in a model depicting WNV transmission improves WNV forecast accuracy relative to a baseline model depicting WNV transmission without temperature forcing . Both models are optimized using a data assimilation method and two observed data streams: mosquito infection rates and reported human WNV cases . Each coupled model-inference framework is then used to generate retrospective ensemble forecasts of WNV for 110 outbreak years from among 12 geographically diverse United States counties . The temperature-forced model improves forecast accuracy for much of the outbreak season . From the end of July until the beginning of October , a timespan during which 70% of human cases are reported , the temperature-forced model generated forecasts of the total number of human cases over the next 3 weeks , total number of human cases over the season , the week with the highest percentage of infectious mosquitoes , and the peak percentage of infectious mosquitoes that on average increased absolute forecast accuracy 5% , 10% , 12% , and 6% , respectively , over the non-temperature forced baseline model . These results indicate that use of temperature forcing improves WNV forecast accuracy and provide further evidence that temperature influences rates of WNV transmission . The findings provide a foundation for implementation of a statistically rigorous system for real-time forecast of seasonal WNV outbreaks and their use as a quantitative decision support tool for public health officials and mosquito control programs .
West Nile virus ( family Flaviviridae , genus Flavivirus , WNV ) is the leading cause of domestically acquired arthropod-borne viral ( arboviral ) disease in the United States ( U . S . ) [1] . It was first identified in North America in New York City during the summer of 1999 [2] and by 2003 had spread throughout the continent [3] . Following this expansion , the virus did not disappear but instead settled into a pattern of endemic zoonotic transmission between vector mosquitoes and avian hosts . Spillover transmission to humans has since continued; however , during 2012 human WNV cases surged to numbers not seen since 2003 suggesting that it will continue to produce unpredictable local and regional outbreaks throughout the U . S . [3] . Vector control agencies monitor mosquito and viral activity and use this information to guide mosquito and WNV control measures [3 , 4]; however , due to the nonlinearity of WNV transmission dynamics [5] it is non-trivial to project future WNV activity from current , observed conditions . Indeed , seasonal WNV outbreaks vary considerably in size and scope [3 , 6] such that , even after an outbreak has begun , it remains difficult to predict the future characteristics of the epidemic curve [6] . Accurate forecasts of WNV outbreak characteristics are thus needed so that public health response and mosquito control efforts can be more effectively coordinated and implemented [7] . In a recent study , we showed that accurate and reliable predictions of seasonal WNV outbreaks can be made using a parsimonious mathematical model representing the transmission dynamics of WNV among mosquitoes and birds , as well as spillover to humans [8] . The dynamic model was recursively optimized using the ensemble adjustment Kalman filter ( EAKF ) , a data assimilation technique , and observational estimates of the proportion of infectious mosquitoes and the number of reported human cases . This model system accurately forecast mosquito infection rates prior to the week of mosquito peak infection , and accurately predicted the seasonal total number of human WNV cases up to 9 weeks prior to the last reported case . The model-inference system thus provides a basic , parsimonious structure for forecasting WNV; however , it is natural to wonder whether inclusion of additional physical environmental factors within the model structure , which have been shown to affect WNV transmission dynamics , would improve forecast skill [8] . Ecological and laboratory studies have demonstrated that physical environmental factors ( e . g . , temperature , precipitation , hydrology , and humidity [9–26] ) influence WNV transmission dynamics and spillover infection of humans , often in complex , nonlinear fashion . For example , rainfall increases near-surface humidity , which enhances mosquito flight activity and host-seeking behavior and alters the abundance and type of aquatic habitats available to the mosquito for egg deposition and sub-adult growth [16] . Thus , for some species , above average rainfall can increase vector populations , but extreme rainfall events can decrease survival by flushing larvae from aquatic habitats [17] . On the other hand , years with below average rainfall may concentrate water resources for both avian hosts and mosquitoes creating optimal conditions for WNV amplification [18] . Additionally , warmer temperatures affect mosquito development rates , shorten the duration of the gonotrophic period [12] , and decrease the extrinsic incubation period of the virus [9–11]; however , at high temperatures , vector survival decreases [13 , 14] . These myriad non-linear effects make it difficult to estimate how environmental variables influence overall WNV risk . Here , we explore whether the addition of a biological parameter depicting the relationship between temperature and the extrinsic incubation period will improve our ability to forecast WNV . The challenge is that the inclusion of too many processes results in a high-dimensional model structure , which , given the limited observational data streams available , may be difficult to optimize . We use the relationship between temperature and the extrinsic incubation period to expand our previously developed parsimonious model to include an environmental factor , a temperature-forcing parameter , that modulates the zoonotic transmission of WNV between mosquito vectors and avian hosts . We couple this mechanistic model with the ensemble adjustment Kalman filter ( EAKF ) [27] for data assimilation and generate retrospective forecasts for a total of 110 outbreak years in 12 different U . S . counties ( Fig 1 ) . The accuracy of the forecasts that use historical climatology for temperature forcing , see Fig 2 for example , are compared with WNV forecasts generated with our baseline system , which lacks temperature forcing and has one less parameter to optimize [8] . We show that temperature forcing improves forecast accuracy for total human case numbers , the week of peak infectious mosquito abundance , and the magnitude of that abundance for much of the WNV season , whereas the total number of infectious mosquitoes is more accurately predicted using the baseline forecast model .
Retrospective WNV predictions were generated for 12 different counties representing a total of 110 outbreak years ( Fig 1 , see Methods ) . For each annual outbreak , the model-EAKF system was initiated four weeks prior to the first positive mosquito observation . Each week , observations of human WNV cases and , when more than 300 mosquitoes were sampled , mosquito infection rates were assimilated using the EAKF , and a forecast was generated by integrating the posterior model ensemble to the end of the outbreak season . Forecasts were generated each week from the first detection of infectious mosquitoes to the end of the year using both the baseline system [8] , which lacks temperature forcing , and the temperature-forced system ( for example forecasts see Fig 3 and supplementary information ( SI ) S1 and S2 Figs ) . Forecast accuracy was evaluated for both the short-term , 1–4 weeks in the future , and the season . Short-term forecast were deemed accurate if the ensemble mean trajectory was within ±25% or ±1 case , whichever was larger , of the number of human cases reported for each of the next 4 weeks . Seasonal forecast accuracy was assessed using 4 metrics: total human WNV cases , total infectious mosquitoes , peak infectious mosquitoes and peak timing . Forecasts were deemed accurate if the ensemble mean trajectory was within ±25% or ±1 case , whichever was larger , of the first metric , within ±25% of the next two metrics , and within ±1 week of the fourth metric . Forecast accuracy across all outbreaks , regions and seasons , was assessed for both the baseline model and temperature-forced model as a function of calendar week . Forecasts were further grouped by prediction lead-time , here defined as the week of forecast generation minus the week of predicted peak mosquito infection . Both the baseline model and temperature-forced model produced accurate short-term forecasts ( see S3 and S4 Figs ) . For weeks 31–40 , when 70% of human cases were reported ( see S5 Fig ) , the temperature-forced model forecasts accurately predicted human case numbers 1 , 2 , 3 , and 4 weeks in advance 74 , 63 , 57 and 52% of the time , respectively . Without temperature forcing , the baseline model was accurate 1 , 2 , 3 , and 4 weeks in advance 73 , 60 , 52 , and 48% of the time , respectively . For forecasts with predicted leads between -1 to 4 weeks past the peak week of mosquito infection , when 52% of human cases were reported , the temperature-forced model on average improved absolute forecast accuracy for human case numbers 1 , 2 , 3 , and 4 weeks ahead by 6 , 8 , 10 and 12% , respectively , over the baseline model . For the seasonal forecasts , the baseline model generated more accurate forecasts as a function of calendar week for total number of human cases , peak timing , and peak magnitude early in the season but by the middle of the season ( the end of July ) , weeks 31 , 30 , and 31 , for these 3 metrics , respectively , the temperature-forced model forecasts were more accurate ( Fig 4 ) . Between week 31 and week 40 , when 70% of human cases were reported , the temperature-forced model forecasts on average improved absolute accuracy by 10 , 12 , and 6% for the total number of human cases , peak timing of infectious mosquitoes , and peak magnitude , respectively , over the baseline model forecasts . Only one-fifth of outbreaks among mosquitoes peaked prior to week 31 , and only one-quarter of human cases were reported prior to week 31 ( see S5 Fig ) ; thus , for the majority of outbreaks the temperature-forced model produced more accurate forecasts of peak timing , total number of human cases , and the peak mosquito infection rates . On the other hand , the baseline model was generally more accurate forecasting the total number of infectious mosquitoes recorded during an outbreak . Prior to the predicted peak , the baseline system forecast the number of human cases , the peak mosquito infection rate , and the seasonal mosquito infection rate more accurately than the temperature-forced system ( Fig 5 ) ; however , peak timing was more accurately forecast using the temperature-forced system at all leads . Though the temperature-forced model was less accurate predicting seasonal mosquito infection rates , it generated more accurate forecasts of spillover transmission to humans . At 0 , 1 and 2 weeks past the predicted week of peak mosquito infection , the temperature-forced model forecasts of total human WNV cases were accurate 63 , 70 and 73% of the time , respectively , whereas the baseline model was accurate 32 , 35 and 49% , of the time , respectively . Only one-quarter of human cases were reported before the week of peak mosquito infection . Consequently , forecasts of human cases near the predicted peak of mosquito infection are prior to reporting of the majority of human cases and thus provide considerable advanced warning ( see S5 Fig ) . Seasonal forecast accuracy was also compared to local historical average outbreak conditions to determine if the system could simply forecast accurately whether an outbreak was earlier or later than normal , or larger or smaller than normal . The average outbreak for each county was defined as the mean value for the 4 metrics ( total human WNV cases , total infectious mosquitoes , peak infectious mosquitoes and peak timing ) for all years excluding the forecast year . Both forecasting approaches were greater than 65% accurate predicting all 4 metrics 3 weeks prior to the predicted week of peak mosquito infection , see S6 Fig . Fig 6 shows absolute error for the 4 forecast metrics in which the predictions have been binned in 3-week groupings . We used the Wilcoxon signed-rank test to evaluate which forecasts had significantly lower error for a given week ( see S1 Table ) . The baseline model had significantly lower error than the temperature-forced model for human cases , peak magnitude and total infectious mosquitoes prior to weeks 31 , 31 and 35 , respectively . However , during the part of the season when the majority of spillover transmission occurs , the temperature forced model had significantly lower error than the baseline model for human cases , peak magnitude and remaining weekly seasonal infectious mosquitoes for weeks 31 to 47 , 34 to 52 and 35 to 50 , respectively . From week 31 until the end of the outbreak the temperature-forced model had statistically significant lower error than the baseline model in its forecasts of mosquito infection peak timing . The short-term forecast accuracy for human cases was also evaluated using the Wilcoxon signed-rank test ( see S2 Table ) . The temperature-forced forecast had statistically significant lower error for the remainder of the season beginning week 33 , 32 , 31 and 31 for the 1 , 2 , 3 and 4 week forecasts , respectively . We further evaluated differences in forecast accuracy based on geographic location ( northern v . southern ) and precipitation levels ( wetter v . drier ) ( see S1 Text for details ) . For 3 of the 4 sub-groups we found similar results in which the temperature-forced model improved retrospective forecast accuracy for 3 of the 4 metrics evaluated relative to the baseline model ( see S7–S10 Figs ) . Using the temperature-forced model , counties experiencing more precipitation were more accurately forecast than drier counties ( see S11 Fig ) , whereas there were no consistent differences between northern versus southern counties ( see S12 Fig ) . We additionally simulated and compared 6 different temperature-forcing structures in order to better understand the impact of temperature on WNV disease dynamics ( see Methods ) . The 6 scenarios are: 1 ) the baseline model without temperature forcing , 2 ) our principal temperature forcing: daily climatology temperature forcing , 3 ) observed daily temperature forcing ( i . e . , real weather , which is not feasible for use in real time forecasting as such data would not be available ) , 4 ) observed seasonal average temperature forcing , 5 ) permuted observed temperature values for a given year and location; and 6 ) permuted temperature values from the historical 1981–2000 record ( see S13 Fig for examples of each temperature time series ) . The climatology and real weather-forced models both performed better than the baseline model ( see Fig 7 and S14 Fig ) . Real weather forcing was more accurate than the baseline model for peak timing , peak magnitude and total human WNV cases during the majority of the outbreak season ( S3 Table ) . However , forecast accuracy was best with climatology temperature forcing ( S4 Table ) , indicating that a smoothed temperature forcing function generates more accurate forecasts . The seasonal average and permuted temperature forcings degraded forecast accuracy , indicating that a seasonal cycle and realistic weather patterns are needed to improve forecast accuracy over the baseline model .
Our findings demonstrate that a simple WNV model , iteratively optimized with data assimilation methods and weekly observations of mosquito infection rates and human WNV cases , can produce accurate forecasts of mosquito infection rates , infectious mosquito biting pressure and human cases in a variety of locations around the U . S . ( Fig 1 ) . Though WNV transmission dynamics vary by location [10 , 28–32] , the simple forecasting framework presented here was designed for broad application , and , in representing modulation of the extrinsic incubation period , temperature forcing improves forecast skill over the non-temperature forced model for much of the WNV season . Specifically , the temperature-forced forecast model improves prediction accuracy for 4 of the 5 metrics evaluated: short-term ( 1–4 week ) human cases , the total number of human cases , peak timing of infectious mosquitoes , and peak magnitude of infectious mosquitoes . These improvements manifest prior to when the majority of human cases are reported and prior to the peak week of mosquito infectiousness . Also , the results were not sensitive to the tolerance ( e . g . ±25% ) used to define accuracy ( see S15 Fig ) . The forecasts of the total number of infectious mosquitoes on average were more accurately predicted using the baseline model . When broken down by calendar week it is apparent the baseline model is more accurate forecasting total infectious mosquitoes early in the season , but around week 30 , when mosquito infection rates have risen , the temperature-forced model is more accurate ( see S16 Fig ) . Estimated mosquito-bird contact rates differed between the baseline model and the temperature-forced model ( S17 Fig ) . This discrepancy is not surprising given the imposed differences in the model forms ( see Eqs 6 and 7 in the Methods ) . In contrast estimates of the probability of spillover from mosquitoes to humans were more aligned for the two models ( see S18 Fig ) . For both mosquito-bird contact and spillover average parameter values varied between counties , and counties with greater population density tended to have higher spillover rates . For the majority of counties spillover increases later in the season , and this increase occurs around the time when mosquito-bird contact rates decreases . These shifts represent changes in mosquito feeding preference from avian to mammalian hosts [33–35] . For two counties , Allen County and Iberia Parish , we did not see an increase in human spillover due to the majority of years having few ( 0 or 1 ) human WNV cases reported . The other counties had higher numbers of human cases . In expanding the WNV forecast system , we had to choose which physical environmental effects to incorporate . While precipitation , hydrology , humidity and temperature all influence WNV dynamics [9–26] , inclusion of all these factors would increase the dimension of the model structure , which , given the limited observational data streams available , might make optimization more difficult . On the other hand , the model needs to contain sufficient dynamics to generate a characteristic WNV outbreak in free simulation . Addition of temperature-forcing appears to provide additional realistic biological constraint of WNV transmission dynamics and allows more accurate forecast during the heart of WNV outbreaks . In addition to affecting viral development , temperature forcing may also impact host seeking behavior , gonotrophic period , and vector survival . For instance , estimates of mosquito-bird contact rates and mosquito birth/mortality rates differ between the two model forms ( see S17 and S19 Figs ) . These differences are consistent with previous studies , which have shown that warmer temperatures influence mosquito development rates , shorten the duration of the gonotrophic period and decrease the extrinsic incubation period of the virus , increasing disease transmission , but at the same time decrease vector survival [9–14] . Thus , the temperature-forced model appears to more accurately depict transmission dynamics and improves forecast accuracy . Although the number of counties included in this study is not geographically exhaustive , the findings provide evidence that the methods presented can be flexibly applied to a diversity of regions to produce accurate forecasts . The 12 counties included here represent a diverse set of locations with differences in population density , ecosystems ( e . g . . wet/dry , warm/cold ) , primary vectors , and mosquito monitoring practices . In spite of these differences the EAKF was still able optimize the forecast model for simulation and skillful forecast . We believe that the method is applicable to regions ( e . g . the Northern Plains ) not represented in this sample , provided appropriate mosquito monitoring and disease surveillance systems are in place . We also expect the temperature-forced model to function better than the baseline model given the improvements in forecast accuracy apparent across latitude and precipitation levels ( see S1 Text and S7–S10 Figs ) . The greater forecast accuracy for wet versus dry counties may be indicative of a need to represent hydrologic effects . Precipitation in dry counties may have a larger impact on vector abundance than in wet counties , where breeding habitats are more plentiful and thus not as dependent on rainfall [21] and irrigation practices [22] . Further , in drier counties , mosquito collections were more variable . Low mosquito trap numbers corrupt inference of the proportion of infected mosquitoes in a given week , and when collections fell below 300 in a week , these observations were discarded ( see Methods and S1 Text ) . As a consequence , drier counties had more gaps in their mosquito infection data records . In the future , more robust collection data may improve forecast accuracy in drier regions . Mosquito population dynamics vary with local conditions [20] and may play an important role determining WNV spillover risk . We tested incorporating vector abundance by using entomological risk rather than prevalence as the estimate of mosquito infection rate . This approach produced forecasts with similar , though somewhat lower accuracy for most outcomes ( see S1 Text and S20 and S21 Figs ) . In the future , particularly should forecasts be generated at a smaller spatial scale , it will be important to reexamine inclusion of population dynamics in the model framework . Among the different temperature-forcing approaches compared , we found evidence that observed seasonal temperature variability is important but inter annual differences are not . Both the climatological and daily-observed temperature-forced forecasts were more accurate than the baseline model for peak timing , peak magnitude and total human WNV cases during the majority of the outbreak season . In contrast , forcing with a seasonal average temperature or permuted weather performed worse than the baseline model . These findings indicate that a local seasonal temperature cycle improves forecast accuracy by giving the model additional biologically relevant structure . It is interesting that forecasts with climatological temperature forcing were more accurate than those generated using observed temperature conditions . This finding indicates that short-term fluctuations of temperature due to synoptic variability may actually degrade forecast accuracy . Whether this effect is due to the transience of these signals , which corrupts filter optimization , or the simplicity of the model , which may not appropriately represent the effects of these fluctuations on transmission dynamics , is not clear . Overall , inclusion of temperature forcing improves the forecast skill of our parsimonious WNV model and provides further evidence that temperature modulates rates of WNV transmission . Though these forecast models do not represent the full complexity of WNV transmission dynamics , including effects such as ongoing mosquito control efforts , within county spatial heterogeneity [28] , bird migration [31] , variable susceptibility among different hosts [36] and vectors [29] , WNV strain variability [30] , vertical transmission [29] , and mosquito overwintering patterns [13] , all of which can affect WNV transmission dynamics and spillover infection to humans , the system does generate accurate predictions of a number of outbreak metrics . As more data become available , inclusion of additional effects in the model structure , such as differences in risk based on population exposure probabilities , climatic zone and mosquito species , may be possible and may enable more accurate and nuanced forecast . The present forecasts , if operationalized , could provide public health officials , mosquito control programs , and parks departments a quantified estimate of WNV spillover transmission risk . Such enumeration is important because decision makers often rely on a limited number of heuristic principles instead of assessing probabilities when making operational decisions . In general , these heuristic principals are useful , but sometimes lead to severe and systematic errors [37] . More quantitative decision tools , such as these forecasts , could thus provide a systematic platform for implementing response and control efforts . As real-time forecast of WNV is operationalized , potential challenges will arise due to the need for robust timely data sets of both mosquito monitoring and human cases . In response to the emergence of WNV , an electronic surveillance system for arboviral disease , ArboNET , was developed by the Centers for Disease Control and Prevention ( CDC ) in 2000 , and the CDC classified human cases of WNV as a nationally notifiable disease . ArboNET requires state and local health departments to report weekly human WNV case counts , along with infection of mosquitoes , birds and other animals , in order to monitor WNV activity across the country; however , this passive system has long lag times between when the data are generated and reported . In a retrospective study of Colorado hospital discharge data between 2003 and 2005 only 77% of hospital WNV cases were ever reported , and of those cases only 51% were reported within 7 days [38] . If real-time WNV forecast is to be operationalized , electronic monitoring systems , such as ArboNET , will need to shorten the delay between when data are collected and reported . In addition to accelerated human case reporting , some areas of the US might benefit from more active mosquito monitoring . Mosquito monitoring practices vary around the country and are influenced by local socioeconomic factors , the tax base , and the public/political will to budget for mosquito surveillance [4] . Previous studies have shown that increasing mosquito trap density decreases measurement uncertainty and remains a cost effective way to monitor WNV activity [4] . If some mosquito abatement districts were to increase their monitoring budgets , additional mosquito traps early in the season might provide important data for improving forecast accuracy , and model studies could potentially provide information on how to design a monitoring program for optimal surveillance . In addition to a more active monitoring program , operational forecasting would benefit from shorter lags between when mosquitoes are trapped and test results are received . The lag associated with tests results varies by abatement district . Large districts such as the city of Chicago or Maricopa County run in-house laboratories that provide same day or one-to-two day testing lags , whereas others ship samples to state laboratories , which subsidize testing , but have substantially longer lags: 7 to 10 days [39 , 40] . For example , New York state lab fully subsides WNV testing for abatement districts [39] and the Louisiana state lab only charges $1 . 50 per WNV sample [40] . If St . Tammany Parish , LA , conducted in-house WNV sampling , each sample would cost approximately 10 times more than the state lab [40] . More work is needed to understand the costs associated with these lag times and the value of real-time data relative to the impact associated with each human cases of WNV , which is expensive and debilitating . For instance , the economic impact of an outbreak in Louisiana was significant with the median cost per human case of WNV around $11 , 000 ( ranging from $830 -$218 , 600 ) [41] . In California the costs associated with neuroinvasive WNV and WNV fever are $75 , 600 and $1 , 430 [42] , respectively ( 2016 US$ ) . State subsidies might be better spent in providing counties with more resources for onsite testing that reduces the time between when mosquitoes are collected and a decision maker has quantified information pertaining to the outbreak . The general under reporting of human WNV infections [38 , 43 , 44] is an observational bias , and we implicitly account for this bias in the forecasting model . That is , the parameter , η , which estimates mosquito-to-human transmission ( see Methods ) , is necessarily only estimating recorded human cases . As long as roughly 4% of all human WNV infections are recorded each year [45]—and this rate appears relatively stable—the forecasts should perform reliably well . The human WNV case predictions are thus for the 1 in 25 cases that are clinically identified . However , the existence of a larger sub-clinical set of human WNV infections , whose numbers correlate with the clinically identified cases , must also be recognized . Should resources be provided to generate timely data streams , real-time WNV forecasts could be operationalized . These real-time forecasts could be used as a decision support tool by public health officials , mosquito control programs , and parks departments to help target control of infectious mosquito populations , alert the public when WNV spillover transmission risk is elevated , and identify when to intensify blood donor screening .
|
West Nile virus ( WNV ) is the leading cause of domestically acquired arthropod-borne viral disease in the United States . Here we show that accurate retrospective forecasts of mosquito infection rates and human WNV cases can be generated for a variety of locations in the U . S . Incorporation of temperature forcing into a baseline dynamic model improves our ability to accurately forecast WNV outbreaks and provides further evidence that temperature modulates rates of WNV transmission . These findings provide a foundation for implementation of a statistically rigorous system for real-time short-term and seasonal forecast of WNV . Such a decision support tool would help public health officials and mosquito control programs target control of infectious mosquito populations , alert the public to future periods of elevated WNV spillover transmission risk , and identify when to intensify blood donor screening .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"atmospheric",
"science",
"pathogens",
"microbiology",
"vertebrates",
"animals",
"viruses",
"seasons",
"mathematics",
"rna",
"viruses",
"forecasting",
"statistics",
"(mathematics)",
"insect",
"vectors",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"birds",
"medical",
"microbiology",
"epidemiology",
"mathematical",
"and",
"statistical",
"techniques",
"microbial",
"pathogens",
"disease",
"vectors",
"insects",
"arthropoda",
"mosquitoes",
"eukaryota",
"climatology",
"west",
"nile",
"virus",
"flaviviruses",
"meteorology",
"viral",
"pathogens",
"earth",
"sciences",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"physical",
"sciences",
"amniotes",
"statistical",
"methods",
"organisms"
] |
2018
|
Use of temperature to improve West Nile virus forecasts
|
Understanding how populations of neurons encode sensory information is a major goal of systems neuroscience . Attempts to answer this question have focused on responses measured over several hundred milliseconds , a duration much longer than that frequently used by animals to make decisions about the environment . How reliably sensory information is encoded on briefer time scales , and how best to extract this information , is unknown . Although it has been proposed that neuronal response latency provides a major cue for fast decisions in the visual system , this hypothesis has not been tested systematically and in a quantitative manner . Here we use a simple ‘race to threshold’ readout mechanism to quantify the information content of spike time latency of primary visual ( V1 ) cortical cells to stimulus orientation . We find that many V1 cells show pronounced tuning of their spike latency to stimulus orientation and that almost as much information can be extracted from spike latencies as from firing rates measured over much longer durations . To extract this information , stimulus onset must be estimated accurately . We show that the responses of cells with weak tuning of spike latency can provide a reliable onset detector . We find that spike latency information can be pooled from a large neuronal population , provided that the decision threshold is scaled linearly with the population size , yielding a processing time of the order of a few tens of milliseconds . Our results provide a novel mechanism for extracting information from neuronal populations over the very brief time scales in which behavioral judgments must sometimes be made .
Firing rates of many primary visual cortical cells are tuned to the orientation of visual stimuli [1] . This dependence of neuronal firing rates on the stimulus implies that information about the stimulus can be decoded from the spike count . The trial to trial variability of firing limits the accuracy with which a stimulus can be estimated from the neuronal spike count [2]–[4] . To decrease this variability and increase the accuracy of the rate code , studies have typically used responses measured over several hundred milliseconds [1] , [2] , [5] . However , increasing evidence indicates that the central nervous system can process complex information on very short time scales . Visual psychophysical and evoked potential studies have shown that human subjects can classify natural scenes or emotional facial expressions on the basis of 100–150 ms of processing [6]–[12] . Evidence for fast processing of visual stimuli also exists from behavioral and electrophysiological experiments in monkeys [13]–[15] . A recent study by Stanford et al . [15] shows that monkeys can make perceptual decisions regarding the color of stimuli after about 30 ms of processing time . Evidence for fast coding also exists for the auditory system [16] , [17] and the somatosensory system [18] , [19] . The overall theme deriving from these studies is that sensory systems are able to process the gist of a scene rapidly [20] . It has been suggested that the temporal structure of the neuronal response and in particular , response latency , is the source of fast decisions in the brain [18] , [19] , [21]–[33] . However , the accuracy of codes based on these responses has not been studied in the visual system systematically . A common approach to measuring response latency is to define it as the transition from spontaneous firing to stimulus-dependent firing , e . g . , by detecting the time at the which the PSTH ( Post Stimulus Time Histogram ) reaches half of its maximal firing rate [26] . This attempts to estimate the ‘pure’ latency component of the response , but it involves defining that quantity by a different number of spikes for each condition . For instance , latency might be defined by the time to the first ten spikes at the preferred stimulus and to the first spike at a non-preferred stimulus . Thus , in this approach the criterion for neural response time depends on the stimulus , making it impractical for decoding: the readout parameters cannot scale in a stimulus dependent manner , as that requires the readout to know the stimulus in order to estimate it . Recently , we proposed a simple spike latency code readout [34] , the temporal Winner-take-all ( tWTA ) . The tWTA determines the external stimulus by the label , e . g . preferred orientation , of the cell that fired the first spike in the population . It avoids attempting to estimate ‘pure’ onset latency and instead takes a pragmatic approach in which each cell's time to first spike will depend both on its latency and the strength of its response . Formally , consider a population of N neurons coding for the orientation of a visual stimulus , . Let us denote by the time of the first spike of neuron i , with preferred orientation , following some reference signal tref . The tWTA algorithm estimates the orientation of the external stimulus as the preferred orientation of the neuron which fired first with respect to tref:This definition can be generalized to estimate the stimulus by the preferred orientation of the cell that fired the nth spike first , or to incorporate a competition between groups , ‘columns’ of cells ( see below ) . Here we investigate neural coding on brief time scales by applying the tWTA to simultaneously recorded populations of neurons in the primary visual cortex of macaque monkeys responding to the orientation of visual stimuli .
We first investigated the tuning of first spike times to stimulus orientation . Figure 1A presents eight raster plots showing the response of the same V1 neuron to eight different orientations of the visual stimulus . Qualitatively , both response strength and response latency seem tuned to the stimulus . Measuring latency by simply calculating the mean time to the first spike is problematic because stimuli that evoke weak responses may result in no spikes on some trials . A more principled approach is to incorporate both response time and probability of firing by computing the probability density function and the corresponding cumulative distribution function of the first spike latency . Figure 1B ( upper panel ) shows the cumulative distribution function , ; i . e . , the probability of firing the first spike before time for a given orientation ( is measured with respect to the onset of the external stimulus ) . It is convenient to think of the level curves of this function , , as tuning curves of the neuron . For instance , Figure 1C shows the level curve ( red circles , fits shown by the solid red line and the dashed line in Figure 1B ) , which indicates the time at which there was a 50% chance that the neuron had fired its first spike , for each orientation . Typically , the level curves have unimodal orientation tuning , with a single minimum which we define as the latency-based preferred orientation of the cell . Note that although the choice of the 0 . 5 level curve is arbitrary , similar results were obtained for other criteria . For comparison , the conventional rate-tuning curve of the same neuron is shown in Figure 1D ( black circles represent mean firing rates over the entire response , solid curve represents fitted von-Mises function , stimulus duration was 400 ms; see Materials and Methods ) . The rate tuning is also characterized by a unimodal curve that peaks at the rate-based preferred orientation . Figure 2 shows three additional examples of V1 responses in each column . Eight raster plots for eight orientations are depicted at the top row for each cell . The stimulus dependence of the temporal structure of neural response can be seen from the PSTHs at the second row . The latency tuning curve , in terms of 0 . 5 level curve of first spike time cumulative distribution , is shown on the third row , and the conventional rate tuning curve appears on the fourth row for comparison . Examining the PSTHs of each cell , one can see that response strength has a considerable contribution to first spike latency , in our definition . For example , in cell B it is mainly the firing rate that is tuned to stimulus orientation . Nevertheless , due to the high firing rate near the preferred orientation , the first spike times tend to be shorter near that orientation . It is also evident that the temporal structure of the response is tuned to the stimulus as well . The modulation of the entire temporal structure ( and not a simple temporal shift ) limits the ability to extract the ‘pure’ latency tuning . However , as mentioned above , it is the distribution of the nth spike time that governs the tWTA readout accuracy; hence , the definition of spike latency used here . The middle and bottom panels of Figure 1B depict the cumulative distribution function for the second and third spike times , respectively; the green and blue traces in Figure 1C show the corresponding latency tuning curves ( level curves at 0 . 5 ) . The level curve for the cumulative distribution of the nth spike time indicates a tradeoff: the curves are delayed in time as n increases , but tuning becomes more pronounced . To quantify this behavior we characterized each tuning curve by a ‘DC’ component , denoted by A , which represents the mean latency across all orientations , and by the ‘modulation amplitude’ , denoted by B ( see Materials and Methods ) . Figures 3A and B show the dependence of the mean ( A ) and the modulation amplitude ( B ) of the spike-time tuning curve as a function of n , averaged across the population ( dataset 3 in Table 1 ) . The delay is evident from the linear increase of A with the spike number , while the increase of tuning amplitude , B , indicates that the tuning becomes more pronounced as n increases . A scatter plot showing the mean latency of the first spike against the tuning modulation of the first spike indicates that they are correlated ( Figure 3C; correlation coefficient 0 . 85 ) . This is a manifestation of an empirical result that the first spike latency at the preferred orientation ( A–B ) is approximately constant , and thus neurons with larger modulation amplitudes also have larger mean latencies . Note , that because ( A–B ) is the fitted latency at the preferred orientation and is expected to be positive , we would expect that in general A will be larger than B . We find that , typically , the rate-based preferred orientation is very close to the latency-based preferred orientation . Figure 3D shows the distribution of the difference ( in absolute value ) between the rate and the latency preferred orientations of cells with a tuned first spike latency . In about 90% of the cells this difference is less than 20° . In summary , the latency to the first spike is stimulus dependent: it is shortest for the same orientation that evokes the highest firing rate in the cell . Defining response latency by the first two or three spikes , rather than the first single spike , results in tuning with the same preference but with deeper modulation . Thus , spike latency appears to contain useful information about stimulus orientation . Because the brain does not have direct access to information about when a stimulus was presented , a reference signal is required to extract information about stimulus orientation from the first spike latency . Such a reference signal can be reported by neurons which are sensitive to the mere onset of the stimulus . An ideal onset neuron is expected not only to have a uniform spike time latency for all orientations , but also a low spontaneous firing rate , to prevent false alarms . In fact , several neurons in the data showed weak orientation tuning of their first spike latency as well as a low spontaneous firing rate . Figure 4A shows a scatter plot of the spontaneous firing rate against the modulation amplitude , B , of the latency tuning curve for a single dataset ( dataset 3 in Table 1 ) . We categorized neurons as onset detectors if their modulation amplitude was less than 15 ms and their spontaneous firing rate was less than 5 spks/sec ( gray shading in Figure 4A ) . Typically , we had 10–25 onset detectors in a dataset ( 10–25% of the population [35]; see Table 1 ) . Because the parameters A and B are correlated , these neurons also tend to have an earlier latency ( Figure 3C , red dots ) . In a given trial , onset time was determined using a simple coincidence detection mechanism . Stimulus presence was detected if the group of onset cells fired at least m spikes during a time interval of T ms , and stimulus onset was estimated by the first crossing time of this threshold . A high value of the threshold m results in a very low false-alarm rate but compromises the probability of hit , whereas a low value of m increases the hit probability but also the false-alarm rate . By varying the m criterion we can quantify the Receiver-Operating Characteristic ( ROC ) curve of this onset detection mechanism; i . e . , the dependence of the hit probability on the false alarm rate ( Figure 4B ) . Note that , in contrast to standard two alternative forced choice tasks , in a detection task there are no well-defined trials of ‘no stimulus’ , and the stimulus may be absent over a wide range of time intervals . The mean number of false alarms will scale linearly with the duration in which they are counted . Hence , in a detection task , false alarm is measured in rate of occurrence and not in probability . Unless otherwise stated , throughout this paper we use the following parameters for onset detection: a time window of T = 20 ms , with a criterion of 4 standard deviations above the mean number of spikes in this time interval during spontaneous firing . This choice takes into account the need for a fast detection of the onset ( Figure 4C ) while maintaining a high hit probability and a low false-alarm rate . The distribution of estimated onset times ( relative to stimulus onset ) with this criterion is depicted in Figure 4D . Because the detection of stimulus onset involves a simple integration of spikes emitted by onset detectors , it can be realized in a straightforward way in an integrate-and-fire neuron , producing a similar distribution of onset times ( Figure S1 ) . We have shown that first spike latency contains information about stimulus orientation and that there is a distinct subset of neurons whose responses can be used as a timing reference signal . To read out the information embedded in the neural response latencies , we used a temporal Winner-Take-All ( tWTA ) mechanism , with respect to the above onset mechanism [34] . The complete definition of the method used to compute tWTA performance is provided in Materials and Methods . The performance of the tWTA is affected by the spontaneous firing rates of the neurons , since the mechanism can erroneously identify a spontaneous spike as an informative one . This effect is reduced by taking a more general readout , the n-tWTA , in which the identity of the stimulus is determined by the cell or group of cells that fired the first n spikes with respect to the reference signal . This may come at the expense of the time it takes to make a decision . However , if the number of spikes , n , is less than or equal to the group size , N , then the mean decision time of the n-tWTA will be less than the mean first spike time of a single cell , keeping the mechanism fast . As a first test of the tWTA accuracy we quantified how well it can discriminate between two orientations based on single cell responses . We consider the case where one of the orientations is the cell's preferred orientation θ0 ( as defined by its latency tuning curve ) and the other orientation is θ0+Δθ . The tWTA decision rule is to associate the shorter latency with the stimulus at the preferred orientation of the cell and the longer latency with the other stimulus . The probability of correct discrimination , PC , using the n-tWTA was calculated from the probability density function , fn ( θ , t ) , of the n'th spike latency , as estimated from the data with time relative to the external stimulus onset ( see Materials and Methods ) . Similar to psychometric curves in psychophysical experiments , the curve that describes the probability of a correct response as a function of the orientation difference Δθ is termed the neurometric curve of the cell . Figures 5A , C and E show the neurometric curves of 3 single cells . The red , green and blue curves correspond to the n-tWTA readout for n = 1 , 2 , and 3 , respectively . For comparison , we show the neurometric curve of the conventional rate code readout in black ( the firing rate was estimated from the total number of spikes fired by the cell during the entire response ) . Typically , as n increases , the performance improves and approaches that of the rate code . Figures 6A and B compare the accuracy of the first spike latency code , in terms of probability of correct discrimination , and the rate code , for a relatively fine discrimination task ( Figure 6A; 22 . 5 deg ) and for a coarse one ( Figure 6B; 90 deg ) . Latency and rate code accuracy are correlated and , for the coarse discrimination task , the latency code performance is often comparable to that of the rate code . The cumulative distributions of the accuracy of the different codes in these two tasks are shown in Figure 6C . Figures 5B , D and F show the accuracy of the rate code as a function of the time used for the discrimination for three example cells ( same cells as in Figure 5A , C and E ) . For comparison we plot the accuracy of the n'th spike latency code readout at its mean decision time ( see Materials and Methods ) . On brief timescales , the latency code readout is superior to that of the conventional rate code . To quantify this effect , we show in Figure 6D the cumulative distribution of the difference between the accuracy of the n-tWTA and the accuracy of the rate code , as computed at the mean decision time using the n'th spike latency . As is clear from the figure , this difference is always positive , emphasizing the superiority of the latency code on brief timescales . The responses we measured were evoked by drifting gratings . We also recorded and analyzed additional data using flashed static gratings of brief ( 50 ms ) and long ( 300 ms ) durations . These data provided qualitatively similar results ( see Figure S2 ) . Decisions in the central nervous system are expected to involve large numbers of cells . In large populations , the n-tWTA discrimination in a two alternative forced choice paradigm can be thought of as a competition between two ‘columns’ towards a threshold of firing n spikes . To study the dependence of n-tWTA accuracy on the population size we divided the tuned neurons ( B>15 ms ) into artificial columns of equal orientation width according to the latency-based preferred orientation of the cells ( see Materials and Methods ) . For each pair of columns , we measured the probability of correct discrimination as a function of the number of cells in the population ( see Materials and Methods ) . Importantly , unless stated otherwise , the spike latencies in each trial were measured with respect to the onset detection mechanism described above . Thus , the analysis uses only information that is present in the brain , and , in principle , can be performed by an appropriate neuronal mechanism ( see Discussion ) . Figures 7A , B and C show the n-tWTA probability of correct discrimination for three representative pairs of columns as a function of the number of cells in each column , N . The pairs of columns differ in terms of the difference between the preferred orientations , . The blue curve depicts the performance of the naïve tWTA ( n = 1 ) readout . Initially , for small N , tWTA performance increases with N . However , beyond a critical size of , tWTA performance saturates . Theory has shown that two factors may limit tWTA performance . The first is correlations in the first spike latencies of different cells and the second is the spontaneous firing of the cells [34] . We find that although first spike latency is correlated ( Figure S3 ) , its effect on tWTA accuracy is negligible ( Figure S4; Text S1 ) . The dominant factor that limits accumulation of information from large populations is the spontaneous firing . Clearly , adding more cells also results in adding more spontaneous spikes which interfere with informative spikes ( see [34] for a detailed analysis ) . This effect can be reduced by increasing the decision threshold criterion; i . e . , by increasing n . We next analyzed the performance of the n-tWTA readout , which takes the winning group to be the first to fire n spikes . Different curves in Figure 7A , B and C correspond to different values of the decision threshold , n , in the n-tWTA readout . In this regime , as n is increased the maximal performance is also increased . Figures 7D , E , and F show the performance of the best n-tWTA for each N ( that is , the value of the uppermost curve in a vertical cross-section above this N ) . The inset shows the corresponding value of n , , as a function of the population size N . As the population size , N , grows , it pays to consider more spikes in the readout . Moreover , for these values of population size we obtain that is approximately linear in N . Figures 7G , H , and I show the mean decision time of the n-tWTA readout , relative to the onset signal ( decision times higher than 200 ms are truncated ) . As expected , for a given decision threshold , n , increasing the number of neurons reduces the decision time significantly . The important point is that the average waiting time for the nth spike in a population of N∼n cells is around the average waiting time for the first spike of a single cell ( black filled circles ) , which is typically in the range of 40–80 ms . Thus , considering both more spikes and more neurons ( N∼n ) can substantially improve reliability without compromising the decision time . In the preceding analysis we measured response timing relative to an internal stimulus onset detection mechanism . We wondered whether performance could be improved by making use of the absolute timing of stimulus onset . In principle , this could decrease the detrimental effect of spontaneous firing [34] . To evaluate this we used an artificial reference signal ( i . e . not based on neural responses ) which varied from 0 to 120 ms relative to the external stimulus onset . Spike times were then measured relative to this reference signal ( spikes before the signal were ignored ) . Figure 8 shows the accuracy of the tWTA readout ( n = 1 ) as a function of the onset time . Estimating the onset too early causes the readout mechanism to consider more spontaneous spikes which only contribute noise . Overestimating the onset time results in a loss of informative spikes . The performance is thus non-monotonic . Since most cells start responding about 60 to 90 ms following stimulus onset , tWTA accuracy peaks at about this time , at a performance level comparable to that achieved using the internal onset detection signal . For comparison , Figure 8 also shows the mean time ( ±1 standard deviation ) of our onset detection mechanism for the same dataset . As can be seen , the onset detection mechanism matches the range of times that produce optimal performance . We conclude that the speed and accuracy of our decoding is similar to that which would be achieved by making use of absolute information as to when the stimulus was presented . We next studied the issue of tWTA accuracy in a multiple ( M ) -alternative-forced-choice task using the following setting . All the tuned neurons ( B>15 ms ) in each dataset were divided into M ‘columns’ according to their preferred orientation , as depicted in Figure 9A ( see Materials and Methods ) . Note that the number of cells in different groups is not identical and that dividing them into many groups may result in some that contain no cells . The orientation label of each column was defined as the center of that column . The n-tWTA decision in a competition among M columns was defined as the orientation label of the first column to reach a threshold of n spikes . The resolution of this decision is inversely related to the number of alternatives , . Figure 9b shows the probability of correct discrimination of the n-tWTA as a function of in one of the datasets . Different curves correspond to different values of n . The dashed line represents chance value , which is inversely proportional to the number of alternatives . As the decision threshold , n , is increased , n-tWTA performance improves . This improvement is more significant for coarse discrimination tasks; i . e . , for large . To gain more insight , Figure 9C depicts the distribution of errors in a fine discrimination task ( ) using the tWTA ( n = 1 ) . The error distribution is very broad and there are relatively many large errors . These large errors are related to spontaneous firing and reflect the fact that discrimination at fine resolutions involves a competition among many groups ( 180 in this case ) . In a substantial fraction of the trials the winning group is the first to fire a spontaneous spike , which carries no information about the stimulus; hence errors in these cases are distributed uniformly . Using the n-tWTA readout with n = 2 decreases this effect and makes the distribution narrower , as depicted in Figure 9d . Nevertheless , the decision is still based on a competition between one “correct” group and many ( M−1 = 179 ) “incorrect” groups . The chances that one of the “incorrect” groups will fire its first two spikes before the “correct” group are still high and the distribution of errors is still relatively wide . With larger groups of neurons in each bin ( i . e . with more samples than that provided by our microelectrode arrays ) , the decision threshold , n , could be increased so as to improve performance for these more difficult discriminations . Nevertheless , for our dataset , we can conclude that n-tWTA can perform coarse discriminations remarkably quickly and with high accuracy .
Coding of visual attributes by spike latencies was studied previously in the context of contrast processing [26] , [31] , where it was demonstrated that higher stimulus contrast results in shorter response latency . However , some confusion exists in the literature as to the tuning of first spike latency to the orientation of visual stimuli . Whereas Celebrini et al [32] reported tuning of spike latency of V1 neurons to orientation , Gawne et al . [26] claimed that stimulus orientation mainly modulates response strength and only weakly affects response latency [26] . We have shown that first spike latencies of V1 neurons are tuned to the orientation of external stimuli . This tuning is typically unimodal and the minimal latency is close to the orientation that evokes the maximal firing in the cell . The apparent discrepancy with Gawne et al . is due to different definitions of response latency . In their study , Gawne et al . [26] defined response latency to be the time at which the PSTH reaches half of its peak . The utility of this measure is that it attempts to estimate changes in the ‘pure latency’ in a manner that is unaffected by the changes in the firing rate of the cell . However , since firing rate is modulated by orientation , this definition may measure the latency to a single spike at the null orientation and the latency to ten spikes at the preferred orientation . Hence , using this definition should result in flatter latency tuning curves . Indeed , when applying this definition to our data , we found little modulation of latency with orientation ( Figure S5 ) . Moreover , since response strength , the temporal structure of the PSTH , and response latency itself may all be modulated by the stimulus , it is very difficult to obtain a reliable estimate of ‘pure latency’ tuning based on finite amounts of data . Here we took a more pragmatic approach . Since we are interested in the issue of decoding neural responses on brief time scales , we studied latency tuning using the probability density function of first spike time , which is the quantity that governs tWTA accuracy . Our results thus hold regardless of whether differences in first spike latency arise entirely from differences in response strength , or whether there is some tendency for neurons' absolute latency to vary with stimulus conditions . To extract the information embedded in spike latencies , a reference signal is required . Note that a reference signal is also required for decisions based on spike count in order to determine the start of the counting window . In the general case of latency coding , the onset signal gives a natural reference for measuring latency . However , in our case we do not use the absolute response time , but instead only use relative timing , i . e . , who fired first . In this case , an important feature of the onset signal is to filter out spontaneous spikes that are not stimulus dependent and hence carry no information ( see Figure 8 ) . In the case of ‘active sensing’ , the intrinsic signal of the motor command [36] can , in principle , serve as the onset signal . However , in the case of ‘passive sensing’; e . g . , when a child suddenly jumps in front of your car , the onset signal must be estimated from the responses of sensory neurons . Here we suggested a principle by which stimulus onset is estimated by the group of cells that are not tuned to the information that must be processed rapidly . We showed that a simple summation of the responses of ‘onset’ neurons during short time intervals can provide a reliable reference signal , with sufficient accuracy to allow for accurate identification of stimulus orientation . The onset cells were characterized by weak first spike latency tuning , to limit stimulus dependent bias of the estimated onset time , and low spontaneous firing rates to reduce the false alarm rate . Because the tuning modulation and the mean latency are correlated ( Figure 3C ) , these cells also tend to have an early response . However , even if the onset signal arrives slightly after the tuned neurons started to fire , the performance is only mildly decreased ( Figure 8 ) . In terms of the identity of the onset cells , one possibility is that these are inhibitory interneurons , which are known to be responsive but poorly tuned [37] , [38] . Since these neurons do not project downstream , this would imply that onset detection is performed locally . A similar approach has been applied in the past for the estimation of the onset of auditory stimuli by Chase & Young [24] . The main differences are twofold . One , Chase & Young used a ‘pseudo population’ signal whereas we use simultaneous recordings of real neural populations . Two , we used the responses of a distinguished subclass of cells with weakly tuned first spike latency for our onset signal , whereas Chase & Young pooled the responses of all the cells . It remains an open question whether the brain employs a latency-based readout like the tWTA . Nevertheless , the utility of the tWTA in our study has been to enable us to investigate and quantify the information embedded in spike time latency . Let us consider , for example , the case of a two alternative forced choice discrimination task , based on a competition between two neurons . At the time of the first spike the tWTA decision is identical to that of the conventional rate-based readout . The advantage of a latency-based readout is clear when both neurons fired one spike in the counting window . In those cases the latency based readout can extract information from the temporal structure of the response , whereas there is no information in the total spike count . A rate code readout will perform better when more spikes were fired , but this results in a slower readout . A recent study reported that the minimal processing time required for visual perceptual decisions in the monkey is about 30 ms [15] . This brief time scale is on par with the processing time of the latency readout , i . e the mean decision time following the internal onset signal ( see e . g . Figure 7I ) . To test more directly if a candidate readout mechanism is used by the brain one would need to correlate the behavior of animals with the relevant aspects of neural activity . In a recent study [39] , activity of single neurons in monkey V1 was measured together with reaction times for visually guided saccades . It was shown that first spike latency was correlated with behavior whereas firing rate was not , suggesting that spike latency may indeed serve as a source of information for fast decisions in the brain . As noted above , the implementation of the n-tWTA readout requires an integration process and a threshold decision mechanism . In this sense , n-tWTA competition is very similar to the ‘race to threshold’ mechanism suggested by Mazurek et al [40] , in which the decision in a two alternative task is determined by integrating ‘evidence’ ( spikes ) for the two competing alternatives to reach a decision threshold ( n spikes ) . The decision mechanism involves a winner-take-all type competition , which is an algorithm that others have also used to decode neural response [41]–[43] . Winner-take-all competition can be implemented using reciprocal inhibition between the integrators that represent the different alternatives [44] , [45] ( Figure 10 ) . Each inhibitory neuron accumulates evidence for the corresponding alternative and fires when it crosses a threshold . Higher threshold values reflect a stricter decision criterion and correspond to higher values of n in the n-tWTA readout . The integration time constant of the neurons should be on the order of the relevant time scale for decisions ( ∼10–30 ms ) . The circuit also requires a gating mechanism that triggers the integration process based on the reference signal . One qualitative way to implement such a gating mechanism is using NMDA synapses [33] for the tuned inputs ( Figure 10A ) . The inputs from the onset cells are first integrated by a coincidence detector , which in turn excites the inhibitory cells through AMPA synapses ( as shown in Figure S1 , such a coincidence detector can be implemented using a simple integrate-and-fire neuron ) . Only when this detector is active , the inhibitory cells become depolarized and the magnesium block of the NMDA synapses is removed , allowing for integration of the tuned inputs . When the onset cells are silent , the NMDA synapses do not allow inputs from the tuned populations to be integrated . The gating mechanism can also be implemented using a disinhibition pathway ( Figure 10B ) . In this case the onset cells are assumed to be inhibitory . Their inputs are integrated by a neuron which inhibits the competing neurons . Thus , the competing neurons are released from inhibition only when the onset cells are active , allowing the ‘race to threshold’ to begin . Previous studies have proposed more sophisticated mechanisms to combine information from the first spikes of different neurons in a large population . These methods include rank order [30] , [46] and synfire chains [47] . The utility of tWTA is that its simplicity enables statistical analysis of its accuracy , whereas sophisticated readout mechanisms that rely on specific combinations of firing orders cannot be tested with finite data on the order of a few hundred repetitions per stimulus condition . Furthermore , these readouts may be more difficult to implement in biological circuits . Recently first spike latency code has been analyzed in the framework of fast discrimination of sound source location in the auditory system [48] . There are several interesting similarities and differences worth noting . In both systems , many cells exhibit tuning of their first spike latency to the stimulus . Tuned cells are typically characterized by a unimodal latency tuning curve that peaks close to the preferred stimulus of the cell , as defined by the rate tuning curve . In addition , the accuracy of first spike latency readout is typically comparable though somewhat inferior to the accuracy of the conventional rate code in single tuned cells in both systems . The main differences between the systems are the higher spontaneous firing rates in visual cortex and the poorer performance of V1 neurons for orientation discrimination . To overcome the detrimental effect of spontaneous spikes , we developed here a novel onset detection mechanism , based on pooling the responses from a set of simultaneously recorded neurons . The use of simultaneous data from array recordings rather than single units also enabled us to investigate the accuracy of latency coding at the population level without the use of artificial pseudo populations of neurons . In summary , our study demonstrates that the orientation tuning of first spike latencies enables accurate discrimination of orientations on brief time scales . Spontaneous firing limits the resolution of the decision . However , larger populations can afford better resolution . Furthermore , in many cases when fast decisions are essential , it is important that the probability of correct response will be high but coarse resolution may suffice . This may be a general principle used by the nervous system when fast decisions are essential . For example , when an object suddenly appears on the road while we are driving , all we need to know is its rough location . In most cases we react before we realize whether this object is a child , a dog or just a plastic bag . These finer details can be sorted out later as more spikes are accumulated using readout mechanisms that take into account the entire neural response .
All procedures were approved by the Institutional Animal Care and Use Committee at the Albert Einstein College of Medicine of Yeshiva University , and were in compliance with the guideline set forth in the United States Public Health Service Guide for the Care and Use of Laboratory Animals . The methods we use to record from neural populations have been described in detail [49] . In short , we recorded from anesthetized ( sufentanil citrate , typically 6–18 microg/kg/hr , adjusted as needed for each animal ) , paralyzed ( vecuronium bromide , 0 . 1 mg/kg/h ) macaque monkeys ( macaca fascicularis ) . Vital signs were monitored continuously to assure adequate anesthesia and the well-being of the animal . The pupils were dilated with topical atropine and the corneas protected with gas-permeable hard contact lenses . Supplementary lenses were used to bring the retinal image into focus . Neural activity was recorded using the Cyberkinetics “Utah” Array ( Cyberkinetics Neurotechnology Systems ) , using methods reported previously [49] , [50] . The array consists of a 10×10 grid of silicon microelectrodes ( 1 mm in length ) spaced 400 µm apart , thus covering 12 . 96 mm2 . The array was inserted roughly 0 . 6 mm into cortex using a pneumatic insertion device [51] , resulting in recordings confined mostly to layers 2–3 . Signals from each microelectrode were amplified and bandpass filtered ( 250 Hz to 7 . 5 kHz ) . Waveform segments that exceeded a threshold ( periodically adjusted using a multiple of the rms noise on each channel ) were digitized ( 30 kHz ) and sorted off-line . Sorted units included both well-isolated single units and small multiunit clusters . Neuronal receptive fields were roughly 2–5° from the fovea . Visual stimuli were displayed at a resolution of 1024×768 pixels and a video frame rate of 100 Hz on a calibrated CRT monitor . Stimuli were oriented drifting gratings presented in a circular aperture surrounded by a gray field of average luminance ( 8 orientations in 4 datasets and 36 orientations in one dataset ) . Stimuli were presented binocularly , for 300–400 ms , and separated by 500–800 ms intervals during which we presented an isoluminant gray screen . Stimulus orientation was block randomized , and each stimulus was presented 200–400 times ( see Table 1 for details ) . In 4 datasets the initial phase of the drifting grating was identical across trials . To test whether our results were skewed by this , we collected and analyzed additional data using initial phases that were randomized across trials . We obtained similar results from this dataset ( see Figure S6 ) . To verify that our results also generalize to static images , we collected and analyzed responses to static gratings presented for 50 or 300 ms ( dataset 6 in Table 1 ) . We obtained similar results from this dataset ( see Figure S2 ) . The rate tuning curves represent the mean firing rate across all trials at each orientation . The firing rate in a trial was calculated using a time window from stimulus onset to 300 ms after stimulus offset . The tuning curves are well fitted using the Von-Mises function:where θ is the stimulus orientation and φ is the rate-preferred orientation of the cell . To generate latency tuning curves for a neuron we first estimate the probability density function of the first spike latency of this neuron , f1 ( θ , t ) . This is done by computing the histogram of the first spike times over trials and then normalizing it . Note that because in some trials there may be no spikes , the integral of the probability density function may not be 1 but slightly below . The spike times are measured with respect to the external stimulus onset and the histogram is generated using bins of 10 ms from time 0 to 300 ms after stimulus termination . The corresponding cumulative distribution , F1 ( θ , t ) , is generated by direct numerical integration of the density function . A similar procedure is applied to obtain the nth spike time probability density , fn ( θ , t ) , and cumulative distribution , Fn ( θ , t ) , for general n . The latency tuning curve of the n'th spike is defined as the level curve at 0 . 5 of the corresponding cumulative distribution function . These level curves are fitted using a cosine function of the form:where θ is the stimulus orientation and φ is termed the latency preferred orientation of the cell . Parameter A represents the mean latency and B represents the modulation of the tuning . Here , the reference time is chosen to be the onset of the external stimulus , but in principle other external reference times can be used , e . g . 20 ms after stimulus onset . We note that in the cosine fit , changing the reference time will change the value of A but not B . The arbitrary choice of the reference is also why a simple cosine function is more appropriate here than the von-Mises function . Choosing the reference such that at some orientations the latency is zero requires parameter k at the von-Mises function to diverge to infinity . In addition , if the latency is negative with respect to the reference at some orientations , the von-Mises function will not fit at all , as it is purely positive . Because in some trials there may be no spikes , error bars for the latency tuning curves cannot be simply calculated from the standard error of the mean associated with the spike times . In order to generate error bars , we first calculated the standard errors of the mean for the cumulative distribution , F . This can be done by noting that F is the mean of a Bernoulli variable and thus its variance is . The standard error of the mean is therefore: , where K is the number of trials . We then calculated the level curves at 0 . 5 for F+SEM ( F ) and for F-SEM ( F ) , and used them to generate lower and upper error bars , respectively . These error bars are depicted in Figure 1C and in subsequent plots of spike latency tuning . In each dataset we identify a group of cells that can serve for the detection of stimulus onset . These cells are characterized by poor tuning and low spontaneous firing rates . The spontaneous firing rates are estimated from the recordings during the inter-stimulus interval ( ISI ) after each stimulus . From each ISI we remove the first 300 ms , assuming that after this period the cell returned to its spontaneous rate ( i . e . any post-response adaptation of spontaneous rate would have dissipated ) . The tuning is characterized by the modulation amplitude , B , of the cosine fit to the first spike latency tuning curve . In each dataset , the cells with a spontaneous rate lower than 5 spks/sec and with a modulation lower than 15 ms , were labeled as onset detectors . Using this definition , the number of onset detectors in a dataset is roughly 10–25% of the population ( see Table 1 ) . The onset signal in each trial is generated using coincidence detection . We used a running time window of T ms and looked for the first time in which there were at least m spikes in this window ( but see also Figure S1 ) . The onset time is then defined as the end of this window . To set m , we first estimated the mean and standard deviation of the number of spikes that these cells fire in a time window T during spontaneous firing . We then set the threshold m to be Nv standard deviations above this baseline value . By varying Nv for a given T we generated ROC curves for the onset detection process . In subsequent analyses we used T = 20 ms and Nv = 4 standard deviations . This onset signal was used as the reference time tref for measuring spike latencies in the tWTA . The discrimination accuracy of single cells is computed in the context of a Two-Interval 2-Alternative-Forced-Choice paradigm . We assume that the cell is presented with two stimuli , one at orientation θ1 and the other at orientation θ2 , where θ1 is the preferred orientation of the cell . The probability that the tWTA will yield the correct response is the probability that the latency of the response to θ1 will be shorter than the latency of the response to θ2 . To find this probability , we multiply the probability that the neuron first fired at time t in response to θ1 by the probability that it did not fire before t in response to θ2 , and then we sum over all possible times , t ( the time is measured with respect to the onset of the external stimulus ) . Formally , this is given by the following integral:However , recording time is finite . Our data contains only 300–400 ms of stimulus presence and the following 700–800 ms of inter-stimulus time; hence , in some cases the decision threshold is not reached during our recording time . In practice we assume that after time T0 , that contains the stimulus presence time and the initial 300 ms of the following inter stimulus period , the neuron returns to its spontaneous firing rate . Assuming Poisson firing with mean rate λ after time T0 , we obtain:It is also important to note that f and F are estimated from the data using time bins of Δt . The spikes from the responses to θ1 and θ2 may fall within the same time bin , leading to correct discrimination at chance level . Correcting for this effect we obtain:Finally , for general n , the correction that stems from the spontaneous firing after response termination is more complicated due to all the combinations of spike trains that have to be taken into account . The general expression is then:where the coefficients are given by:and is the probability that neuron i fired m spikes up to time T in response to stimulus θi . The probability of correct response Pc is the mean of a Bernoulli variable and the corresponding standard error of the mean can be calculated as , where K is the number of trials . To prevent possible interaction between the discrimination accuracy analysis and the latency tuning analysis , we separated each dataset into a training and test set , each consisting of half of the trials ( randomly chosen ) . The training set was used for estimating the latency preferred orientation of the cell . The test set was then used for constructing the neurometric curve , based on the preferred orientation from the training set . To calculate the mean decision time we first compute the probability that decision will be made between t and t+ Δt , and then compute its mean . To study the dependence of n-tWTA accuracy on the population size we divided the neurons into several artificial columns of equal orientation width ( for datasets with 8 orientations we divided into 8 columns of 22 . 5° width and for the dataset with 36 orientations ( dataset 3 in Table 1 ) we divided into 9 columns of 20° width ) . Each neuron was assigned to the column with the closest orientation to its own preferred orientation ( the number of neurons in such a column ranged from 1 to 14 ) . For each pair of columns , we then constructed a neurometric curve , which measures the probability of correct response as a function of the number of neurons , N . Given two subsets of N cells from each column , we simply went over all trials with the orientation of the first column and then over all trials with the orientation of the second . In each trial , the subset that first fired the n'th spike after the onset signal from the onset neurons was the n-tWTA . If the time of the n'th spike was the same for both subsets we tested whether one of the subsets fired additional spikes in the same bin and took the winner as the subset that had more spikes . The average number of correct responses using the n-tWTA gave an estimate of the probability of correct response for these two subsets of cells . For a given N we averaged this value over 1000 realizations of the subsets of neurons . The decision time in a given trial was the time relative to the onset signal and we calculated its mean and standard error of the mean across all trials . To investigate discrimination among multiple alternatives , the neurons were divided according to their preferred orientation into M groups of equal orientation width , Δθ . For convenience , we set one group to be centered at the stimulus orientation ( e . g . , if M = 18 and the stimulus orientation is 45° , the centers will be at 5° , 15° , 25° , … , 175° ) . On a given trial , the group that was first to fire n spikes was the n-tWTA . If several groups fired the n'th spike at the same time we chose among them in a random manner . The error in the trial was the ( signed ) difference between the orientation of the winning group and the stimulus orientation . The probability of correct response was calculated as the average number of times in which the correct group was the winner .
|
How can humans and animals make complex decisions on time scales as short as 100 ms ? The information required for such decisions is coded in neural activity and should be read out on a very brief time scale . Traditional approaches to coding of neural information rely on the number of electrical pulses , or spikes , that neurons fire in a certain time window . Although this type of code is likely to be used by the brain for higher cognitive tasks , it may be too slow for fast decisions . Here , we explore an alternative code which is based on the latency of spikes with respect to a reference signal . By analyzing the simultaneous responses of many cells in monkey visual cortex , we show that information about the orientation of visual stimuli can be extracted reliably from spike latencies on very short time scales .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"computational",
"neuroscience",
"biology",
"sensory",
"systems",
"neuroscience"
] |
2012
|
Fast Coding of Orientation in Primary Visual Cortex
|
Hepatitis E virus ( HEV ) causes acute hepatitis in many parts of the world including Asia , Africa and Latin America . Though self-limiting in normal individuals , it results in ~30% mortality in infected pregnant women . It has also been reported to cause acute and chronic hepatitis in organ transplant patients . Of the seven viral genotypes , genotype-1 virus infects humans and is a major public health concern in South Asian countries . Sporadic cases of genotype-3 and 4 infection in human and animals such as pigs , deer , mongeese have been reported primarily from industrialized countries . Genotype-5 , 6 and 7 viruses are known to infect animals such as wild boar and camel , respectively . Genotype-3 and 4 viruses have been successfully propagated in the laboratory in mammalian cell culture . However , genotype-1 virus replicates poorly in mammalian cell culture and no other efficient model exists to study its life cycle . Here , we report that endoplasmic reticulum ( ER ) stress promotes genotype-1 HEV replication by inducing cap-independent , internal initiation mediated translation of a novel viral protein ( named ORF4 ) . Importantly , ORF4 expression and stimulatory effect of ER stress inducers on viral replication is specific to genotype-1 . ORF4 protein sequence is mostly conserved among genotype-1 HEV isolates and ORF4 specific antibodies were detected in genotype-1 HEV patient serum . ORF4 interacted with multiple viral and host proteins and assembled a protein complex consisting of viral helicase , RNA dependent RNA polymerase ( RdRp ) , X , host eEF1α1 ( eukaryotic elongation factor 1 isoform-1 ) and tubulinβ . In association with eEF1α1 , ORF4 stimulated viral RdRp activity . Furthermore , human hepatoma cells that stably express ORF4 or engineered proteasome resistant ORF4 mutant genome permitted enhanced viral replication . These findings reveal a positive role of ER stress in promoting genotype-1 HEV replication and pave the way towards development of an efficient model of the virus .
Hepatitis E is a feco-orally transmitted positive strand RNA virus that causes acute and sporadic hepatitis in human and other animals [1] . It is also emerging to be a major cause of infection in organ transplant patients worldwide [2] . Though self-limiting in normal individuals , a peculiar characteristic of HEV is attributed to its ability to cause high mortality ( ~30% ) in infected pregnant women [3] . The viral genome consists of a 7 . 2 kb 5’-capped and 3’-polyadenylated RNA , which encodes three known open reading frames ( ORF ) ; ORF1 codes for non-structural proteins , ORF2 codes for the major capsid protein and ORF3 codes for an accessory protein that associates with multiple host proteins and is supposed to modulate host signaling pathways [1] . ORF3 also interacts with host tumor susceptibility gene 101 ( TSG 101 ) and plays an essential role in virus release [4 , 5] . ORF2 has been observed to bind to the viral genomic RNA [6] , induce endoplasmic reticulum ( ER ) stress [7 , 8] and inhibit NFκB activity [9] in human hepatoma cells , suggesting a possible regulatory role of the viral capsid protein . Seven genotypes of HEV have been reported; genotype-1 ( g-1 ) , genotype-2 ( g-2 ) exclusively infect human whereas genotype-3 ( g-3 ) , genotype-4 ( g-4 ) infect human , pig , deer , mongeese and rabbit . Infection by genotypes 5–7 have not been reported in human . Genotype-5 ( g-5 ) , genotype-6 ( g-6 ) infects wild boar and genotype-7 ( g-7 ) is known to infect camel [10 , 11] . Little is known about the life cycle of HEV owing to lack of a handy animal or cell culture model . Among the various genotypes , in vitro synthesized genome of g-3 and g-4 HEV replicates well in mammalian cell culture [12] . Attempts at achieving high replication efficiency of g-1 HEV in mammalian cell culture have not been successful [13 , 14] . Interestingly , in one of the g-3 HEV infected patients , human ribosomal S17 coding sequence was found to be inserted in the ORF1 region , which conferred growth advantage to the virus [15 , 16] . Molecular mechanisms underlying the above observation remain to be explored . Moreover , such an insertion appears to be a very rare occurrence . Efficient translation and replication are two crucial events in the life of an RNA virus , tight control of which is essential for survival of both virus and its host . These events are under strict surveillance by the host defence machinery as innate antiviral measures , most common being induction of ER stress and unfolded protein response , inactivation of eukaryotic translation initiation factor 2α and shut down of cap-dependent translation [17] . Degradation of viral double stranded RNA by innate immune effectors [18] and autophagy [19] also serves as host defence mechanisms against many viruses . Viruses on the other hand , employ clever strategies to exploit the adversities imposed by the host . Innate immune response is countered by various strategies such as inhibition of type I interferon production [20] , manipulation of pattern recognition receptor signaling [21] , IRF3 ( interferon regulatory factor 3 ) inhibition [22 , 23] and autophagy inhibition [24] . Translational restrictions are overcome by ribosome shunting , reinitiation , stimulation of eIF4F complex assembly , inhibition of elF2α phosphorylation [25] , internal ribosome entry site ( IRES ) mediated translation [26 , 27] and execution of both cap-dependent and cap-independent modes of translation [28 , 29] depending on the cellular state . The second step in the RNA virus life cycle pertains to genome replication . Most viruses encode regulatory proteins , RNA and/or miRNA that exploit host machineries to augment viral replication . Hepatitis C , Dengue and Polio viruses activate autophagy [30 , 31 , 32] . Hepatitis B virus inhibits proteasome activity [33] , which leads to increased viral replication . Hence , depending upon the host cellular condition , each virus seems to have evolved suitable survival strategies that permits its optimal growth . Since g-1 HEV does not replicate efficiently in mammalian cell culture , we wondered whether any particular cellular condition might enhance viral replication . Screening of various compounds known to alter cellular condition revealed a role of ER stress inducing compounds , thapsigargin and tunicamycin in enhancing g-1 HEV replication . Further studies led to the identification of a novel viral protein synthesized from an overlapping reading frame within ORF1 , which was named open reading frame 4 ( ORF4 ) . The role of ORF4 in viral replication was explored .
In order to identify the influence of a particular cellular condition on HEV replication , Huh7 cells were transfected with wild type capped genomic RNA ( WT HEV ) . 6 days post transfection , viral replication was measured by monitoring the level of sense and antisense RNA and estimating the percentage of cells expressing viral helicase and ORF2 . Note that helicase synthesis reflects ORF1 translation from genomic RNA whereas ORF2 is synthesized from the subgenomic RNA ( generated after replication ) . A replication deficient mutant genome ( GAA HEV ) , in which “DD” amino acids of RdRp ( a . a . 1551 , 1552 in ORF1 ) were altered to “AA”; was used to ensure specificity of both assays . Quantitative real-time PCR ( QRT-PCR ) of sense strand RNA level in GAA HEV transfected samples reflected the level of input RNA ( quantity of transfected RNA , Fig 1A ) . Sense strand RNA level was approximately four fold higher in WT HEV RNA transfected sample ( compared to GAA HEV ) , reflecting replication mediated increase . As expected , no antisense RNA was detected in GAA HEV transfected samples but was detectable at basal levels in WT HEV expressing samples . Upon treatment with known ER stress inducers; thapsigargin ( TG ) and tunicamycin ( TUN ) , sense and antisense RNA levels were further increased by 2–3 fold ( Fig 1A ) . Similarly treated samples were analyzed by immunofluorescence assay to measure the percentage of helicase and ORF2 positive cells ( Fig 1B ) . Representative images are shown ( S1 Fig ) . GAA HEV transfected sample contained 2% helicase positive and no ORF2 positive cells , in agreement with QRT-PCR data , confirming specificity of the assays ( Fig 1B ) . 20% helicase and 5% ORF2 positive cells were detected in DMSO treated WT RNA transfected cells . Thapsigargin and tunicamycin treated samples contained significantly higher percentage of helicase and ORF2 positive cells in WT HEV . Helicase positive cells were absent in thapsigargin and tunicamycin treated GAA HEV sample , ruling out the possibility of increased ORF1 translation by these compounds . We next tested whether thapsigargin and tunicamycin enhanced the replication of g-3 HEV using a Gaussia luciferase secreting replicon of g-3 HEV [16] . There was no increase in luciferase level upon treatment of the replicon expressing cells with thapsigargin and tunicamycin ( Fig 1C ) , suggesting that both compounds had a stimulatory effect only on g-1 HEV replication . Assuming that the mechanism underlying the observed stimulatory effect of ER stress on g-1 HEV replication is encoded in the viral genome , we analysed the g-1 HEV genome ( SAR 55 strain , Genbank ID: AF444002 . 1 ) using “ATGpr” , a software to predict potential open reading frames [34] . All known ORFs of HEV were predicted . An unknown ORF of 158 amino acids within ORF1 , located in +1 reading frame ( with reference to ORF1 , 2835–3308 bases from 5’ ) was also predicted , which was named ORF4 ( S1 Table , Fig 1D ) . An ORF coding for a truncated ORF1 protein was also predicted . In contrast , sequence analysis of other HEV genotypes did not reveal any ORF resembling that of ORF4 ( S1 Table ) . Next , we performed a bioinformatics analysis of several g-1 HEV genomic sequences available in public database to find out whether open reading frame 4 is present in all and whether the ORF4 protein sequence is conserved among the various isolates . Additionally , we analyzed the viral genomic sequence from five new cases of g-1 HEV infection , recently isolated by us at the All India Institute of Medical Sciences , New Delhi , India ( Genbank ID: KU168733- KU168737 , Fig 1E ) . All g-1 HEV genomes contain an ORF at the expected position of ORF4 , with a suboptimal Kozak sequence starting either at 2832 or 2834 nucleotides , from 5’ end ( Fig 1D ) . Three different patterns were observed with respect to termination of the putative ORF4; in 8 cases , it terminates at 3311 nucleotides ( 158 amino acids , full length ORF4 protein ) , in two cases , at 3277 nucleotides ( 147 amino acids ) and in remaining 9 cases , it terminates at 3256 nucleotides ( 139 amino acids , Fig 1E ) . ClustalW analysis of the putative ORF4 protein sequence of these isolates revealed ~80% conservation of amino acids ( Fig 1E ) . To verify ORF4 and/or ΔORF1 expression , an in vitro transcription-translation assay was performed using a TNT kit . Two bands corresponding to unprocessed and probably partially processed ORF1 protein ( ** ) were detected ( Fig 2A , top and middle ) . Two bands corresponding to ~20kDa and ~40kDa ( * ) were also observed ( Fig 2A , top ) . No such bands were detected in mock . Mutating the initiator methionine codon of ORF1 to Lysine ( ATG-AAA substitution , 26 ATG mut ORF1 ) resulted in disappearance of ORF1 specific bands without affecting 20kDa and 40kDa bands . Similarly , blocking ORF1 translation initiation by inserting a well characterised stem loop forming sequence [35] upstream of the initiator codon of ORF1 ( SL ins ORF1 ) abolished the bands corresponding to ORF1 without impacting 20kDa and 40kDa bands ( Fig 2A ) . Correlating “ATGpr” prediction with above data suggested that 20kDa band may correspond to translation product of ORF4 . 40kDa band could be a denaturation resistant dimeric form of ORF4 or an unrelated protein . In agreement with the above proposition , TNT of ORF4 coding sequence produced 20 and 40 kDa bands ( Fig 2A , pSGI ORF4 ) . A peptide based rabbit polyclonal antibody was generated against the putative ORF4 protein in order to identify the unknown bands . Functionality and specificity of the antibody was validated ( S2A and S2B Fig ) and aliquots of TNT samples were western blotted using this antibody . Only the 20kDa band was detectable by ORF4 antibody in WT ORF1 , 26 ATG mut ORF1 , SL ins ORF1 and pSGI ORF4 ( Fig 2A , bottom ) . Two sub optimal Kozak sequences containing initiation codons are present in the ORF4 coding region ( Fig 1D ) . Both were mutated to Lysine ( ATG-AAA ) in HEV ORF1 construct ( ORF4 ATG DM ORF1 ) , followed by TNT to confirm the identity of 20 and 40kDa bands . Both bands were absent in the autoradiogram and western , without affecting ORF1 level ( Fig 2A ) . As expected , inhibiting both ORF1 and ORF4 translation initiation by stem loop insertion and ATG-AAA substitution , respectively , resulted in disappearance of all bands ( SL ins ORF1 DM ORF4 ) . An immunofluorescence assay was conducted using ORF4 antibody to detect its expression in WT g-1 HEV genome transfected Huh7 cells . ORF4 signal was significantly higher in tunicamycin and thapsigargin treated cells compared to the DMSO control ( Fig 2B ) . Specificity of the signal was controlled by using tunicamycin treated DM HEV ( mutant g-1 HEV genome , in which ORF4 initiation codons are mutated to Lysine ) transfected cells , which failed to show ORF4 signal . Tunicamycin treated 26 ATG mut HEV or GAA HEV RNA transfected cells also expressed ORF4 , clearly ruling out any influence of ORF1 translation or replication on ORF4 production , respectively ( Fig 2B ) . In order to confirm that no ORF4 like protein is expressed in genotype-3 HEV ( g-3 HEV ) , in vitro transcribed genome of a luciferase replicon of g-3 HEV ( pSK HEV p6 luc ) or WT g-1 HEV was transfected into Huh7 cells , followed by thapsigargin treatment and subsequent immunofluorescence staining using anti-ORF4 or anti-Helicase antibodies . Helicase expression was detectable in both samples whereas ORF4 expression was detectable only in the case of g-1 HEV ( Fig 2C ) . Next , we analysed ORF4 expression in the five g-1 HEV infected patients , in which ORF4 sequence was conserved ( KU168733-KU168737 , Fig 1E ) . ORF4 expression was assessed indirectly by monitoring the level of anti-ORF4 antibody , if any . Purified GST-ORF4 protein was readily detected by serum from all 5 patients ( KU168733-KU168737 ) whereas serum from two healthy ( CS1-CS2 ) individuals were negative ( Fig 3A ) . A stable cell line of Huh7 constitutively expressing Flag-tagged ORF4 was generated ( ORF4-Huh7 ) to explore the role of ORF4 in HEV replication ( Fig 3B ) . WT HEV or GAA HEV genome was transfected into ORF4-Huh7 and its control ( pCDNA5-Huh7 ) . The level of sense and antisense RNA of WT HEV was approximately two fold higher in DMSO treated ORF4-Huh7 cells compared to control ( Fig 3C ) . As expected , GAA HEV mutant was unable to replicate . Tunicamycin treatment increased sense and antisense RNA by two fold in control and four fold in ORF4-Huh7 cells . DM HEV behaved like GAA HEV in control cells in the presence and absence of tunicamycin . In contrast to GAA HEV , DM HEV produced both sense and antisense RNA at levels equivalent to WT HEV in DMSO treated ORF4-Huh7 cells and remarkably , these levels remained unaltered in the presence of tunicamycin ( Fig 3C ) . Similar pattern was obtained in immunofluorescence analysis of helicase and ORF2 positive cells ( Fig 3D ) . Thapsigargin too displayed a pattern similar to tunicamycin ( Fig 3D ) . ORF1 translation is cap-dependent [1] . However , ORF4 could be translated in the absence of cap-dependent translation ( SL ins ORF1 , Fig 2A ) . Considering its location deep inside ORF1 , we wondered whether ORF4 synthesis was driven by an internal translation initiation mechanism . Bioinformatics analysis of viral RNA flanking ORF4 region using “Reg RNA” [36] indicated the presence of a putative IRES-like element between 2701–2787 bases ( Fig 1D , IRESl ) . Analysis of same sequence using “IRESite” [37] predicted weak homology with Equine Rhinitis A virus-1 IRES [38] . Secondary structure analysis of 2664–2845 bases encompassing the predicted IRES-like element using “mfold” [39] revealed the presence of three stem loops within 2701–2787 bases ( Fig 4A , sequence in cyan ) . Increase in sequence length ( 315 bases , 2619–2933 bases ) did not alter those stem loops , indicating their stability ( S3 Fig ) . A dual luciferase reporter assay was conducted to evaluate the functionality of the IRES-like element by placing it between Renilla and Firefly coding sequences ( Fig 4B ) . Three consecutive stop codons were introduced downstream of the Renilla coding sequence to ensure termination of cap-dependent translation of Renilla luciferase . 315 bases from HEV genome encompassing the IRES-like element ( HIRESl 315 ) or 468 bases from 3501–3968 nucleotides ( negative control for background Firefly activity , HEVcRNA ) were inserted downstream of Renilla , preceding the Firefly start site ( Fig 4B ) . Measurement of the Firefly and Renilla luciferase ratios revealed a significantly higher Firefly activity in HIRESl 315 sample ( Fig 4C ) . The core IRES-like element ( 2701–2787 bases , HIRESl 87 ) also displayed similar activity ( Fig 4C ) . Next , several mutant constructs were generated in which individual stem loops were destroyed by altering a few nucleotides at a time . Impairing stem loops A , B , C or bulge ( A* ) did not affect Firefly activity . A moderate and high reduction in Firefly activity was seen in samples containing dual mutations of both A , C and B , C stem loops , respectively ( Fig 4C ) . Dual mutations of both B and C were introduced into plasmids containing HEV ORF1 and HEV genome ( IRESl mut ORF1 and IRESl mut HEV , respectively ) . In TNT of IRESl mut ORF1 construct , ORF4-specific band disappeared without affecting that of ORF1 ( Fig 2A ) . No ORF4 was detected in cells transfected with IRESl mut HEV RNA upon tunicamycin treatment ( Fig 2B ) . Expectedly , IRESl mut HEV genome replication was significantly reduced irrespective of tunicamycin treatment in pCDNA5-Huh7 cells , which could be restored in ORF4-Huh7 cells , though in a tunicamycin insensitive manner ( Fig 3C ) . To explore the mechanism ( s ) by which ORF4 stimulated viral replication , we identified its interaction partners among viral proteins . ORF4 directly interacted with helicase , X and ORF3 proteins of g-1 HEV , evident from Yeast Two Hybrid ( Y2H ) assay ( Table 1 ) . X protein of g-3 HEV also interacted with ORF4 , however neither g-3 helicase nor g-3 ORF3 interacted with ORF4 ( Table 1 ) . Using overlapping deletions of ORF4 , the interaction domain was mapped to 54–122 amino acids for X and ORF3 and 1–124 amino acids for helicase protein of g-1 HEV ( Table 2 ) . Coimmunoprecipitation ( CoIP ) of Huh7 cells transfected with plasmids encoding ORF4 and various g-1 HEV proteins confirmed its interaction with X , helicase and ORF3 ( Fig 5A , 5B and 5C ) . Interestingly , CoIP also demonstrated that ORF4 interacted with g-1 RdRp in Huh7 cells ( Fig 5D ) . No other viral proteins interacted with ORF4 in CoIP ( S4 Fig ) . Since X and ORF3 interacted with the same region of ORF4 and helicase appeared to interact with a broader region/multiple domains of ORF4 , we next determined whether X and ORF3 competed or cooperated with helicase for binding to ORF4 . In ORF4-Huh7 cells coexpressing helicase and ORF3 , though ORF4 associated with helicase and ORF3 and vice versa , ORF3 was not coprecipitated with helicase , indicating that all three were not in the same complex ( Fig 6A ) . However , helicase , X and ORF4 could be coprecipitated , indicating cooperativity among them ( Fig 6B ) . We next tested whether viral RdRp associated with the X-helicase-ORF4 complex . CoIP in ORF4-Huh7 and its control cells demonstrated that RdRp coprecipitated with X and helicase only in the presence of ORF4 ( Fig 6C ) . Moreover , all four appeared to be part of one complex as helicase and RdRp antibody could coprecipitate X and ORF4 and vice versa ( Fig 6C ) . A pull down assay using purified proteins further confirmed that ORF4 indeed mediated the assembly of a complex consisting of RdRp , helicase , X and ORF4 ( Fig 6D ) . In contrast , ORF3 inhibited assembly of the above complex , probably by competing for binding to ORF4 ( Fig 6D , compare lane 2 with 3 ) . Since many of the g-1 HEV isolates encode a truncated ORF4 protein consisting of 139aa or 147aa ( from N-terminus ) and our Y2H based mapping of the X , Helicase and ORF3 interaction region of ORF4 was found to be confined to N-terminal 124aa , a pull down assay was performed using a deletion mutant of full length ORF4 protein comprising of N-terminal 124aa ( 124 ORF4-Flag ) . As expected , 124 ORF4-Flag could assemble a complex consisting of RdRp , X and Helicase; similar to the full length ORF4 protein ( Fig 6E ) . These data suggest that ORF4 is functionally active in all g-1 isolates . Since ORF4 interacts with helicase and RdRp , we wondered whether it influenced their activities . A helicase assay using Huh7 purified Helicase-Flag ( Fig 7A ) and bacterial purified GST-ORF4 ( Fig 7B ) revealed that ORF4 had no effect on RNA unwinding activity of helicase ( Fig 7C ) . Huh7 purified ORF2-Flag ( Fig 7D ) was used as a negative control . Helicase assay in the presence of Huh7 purified ORF4 ( Fig 7E ) produced similar results ( Fig 7F ) , indicating that under our experimental conditions , ORF4 had no effect on dsRNA unwinding activity of viral helicase . Next , an RdRp assay was performed using Huh7 purified RdRp-Flag ( Fig 8A ) in the presence of increasing amount of bacterial purified GST-ORF4 or ORF4-Flag . An in vitro transcribed RNA containing 130 bases from 5’-end and 210 bases from 3’-end of g-1 HEV genome was used as template for the assay ( Fig 8B ) . Addition of ORF4 significantly increased double stranded RNA intermediate level ( 680 bases ) , reflecting enhanced RdRp activity ( Fig 8C ) . Observed activity was specific to viral RdRp because no signal was obtained in reactions containing ORF2-Flag or GST-ORF4 alone . RdRp assay in the presence of Huh7 cell purified full length ORF4-Flag or 124 ORF4-Flag ( 1–124 amino acids of ORF4 ) produced a similar effect ( Fig 8D and 8E ) confirming that both full length and 124 aa ORF4 indeed enhanced viral RdRp activity . ORF4 indirectly associated with g-1 RdRp . In an independent study carried out in our laboratory to isolate direct interacting partners of g-1 HEV RdRp by screening a human fetal brain cDNA library using Yeast Two Hybrid ( Y2H ) technique , 21 host proteins were identified ( Table 3 ) . We tested the ability of those proteins to associate with ORF4 . Only eukaryotic translation elongation factor 1 α isoform 1 ( eEF1α1 ) , Tubulin beta ( Tubβ ) and actin gamma isoform 1 were found to be common interaction partners of both RdRp and ORF4 ( Table 3 ) . Though eEF1α1 interacted with equal strength with both RdRp and ORF4 , Tubβ and Actin gamma 1 weakly interacted with ORF4 , compared to RdRp ( Table 3 , compare growth on 3-amino 1 , 2 , 4 triazole ) . CoIP of Huh7 cells expressing RdRp and ORF4 demonstrated that both eEF1α1 and Tubβ associated with ORF4 and RdRp ( Fig 9A ) . We could not detect actin gamma 1 association with ORF4 in CoIP ( S4 Fig , top panel ) . Fraction of both eEF1α1 and Tubβ appeared to associate with ORF4-RdRp complex because both of them could be detected in samples subjected to two rounds of sequential immunoprecipitation ( Fig 9A ) . Next , eEF1α1 and Tubβ proteins were ablated using shRNA to find out whether either or both bridged the interaction between ORF4 and RdRp . shRNAs were approximately 95% and 80% effective in reducing eEF1α1 and Tubβ protein , respectively ( Fig 9B and 9C ) . CoIP revealed the inability of ORF4 to associate with RdRp in the absence of eEF1α1 though association of ORF4 and RdRp with Tubβ remained unaffected ( Fig 9D ) . In contrast , ablation of Tubβ had no effect on the interaction between ORF4 , RdRp and eEF1α1 ( Fig 9D ) . Next , an RdRp assay was conducted using purified RdRp-Flag from respective shRNA expressing cells . Stimulatory effect of ORF4 on RdRp activity was absent in samples lacking eEF1α1 ( Fig 9E ) . Level of RdRp in Flag-affinity purified sample was verified by western and quantified to ensure that eEF1α1 or Tubβ ablation did not prevent RdRp translation ( Fig 9F ) . Finally , we measured the level of sense and antisense RNA of wild type ( W ) and GAA mutant ( G ) HEV in DMSO or tunicamycin treated Huh7 cells expressing EGFP , heEF1α1 or hTubβ shRNA . Lack of eEF1α1 significantly reduced the level of both RNAs in DMSO and tunicamycin treated samples , confirming its essential role in g-1 HEV replication ( Fig 9G ) whereas absence of Tubβ had no effect . Measurement of the level of ORF4 protein in the presence of proteasomal inhibitor MG132 and lysosomal acid protease inhibitor NH4Cl revealed its sensitivity to the former ( Fig 10A ) . Degradation of ORF4 by proteasome was further evident from its polyubiquitination status ( Fig 10B ) . ORF4 contains a lysine at 51st amino acid position flanked by two proline residues ( hydrophobic amino acids ) , which represents a putative ubiquitination site . This Lysine was mutated to Asparagine ( K51N mut ORF4 ) . Monitoring the level of wild type and K51N mut ORF4 in the presence of cycloheximide ( blocks de novo translation ) revealed significantly higher stability of the mutant ( Fig 10C ) , confirming that ORF4 is indeed a target of the proteasome . The K51N substitution was introduced into g-1 HEV genome , followed by transfection of mutant genome into Huh7 cells . Immunofluorescence analysis revealed an increase in the number of ORF4 positive cells in the K51N mutant ( K51N HEV , Fig 10D ) . Measurement of sense and antisense RNA in wild type and mutant genome transfected cells revealed higher level of both RNAs in the K51N mutant , indicative of enhanced replication of proteasome resistant ORF4 encoding genome , which was further increased upon tunicamycin treatment ( Fig 10E ) . Since a proteasome resistant ORF4 mutant genome could significantly enhance the viral replication , we wondered whether such mutations are prevalent in natural cases of g-1 HEV infection . Analysis of the ORF4 ubiquitination site in g-1 HEV sequences illustrated in Fig 1E revealed that 51st Lysine is absolutely conserved in all . However , in one case ( AY204877 . 1 ) , 50th and 52nd Proline residues were substituted with Serine and Leucine , respectively ( Fig 1E ) . In 6 cases ( JF443721 . 1- JF443726 . 1 ) , 50th Proline was substituted with Leucine ( Fig 1E ) . The above substitutions are supposed to prevent ubiquitination at 51st Lysine . Bioinformatics analysis of the above seven ORF4 protein sequences using “UbPred” software ( predicts potential ubiquitination sites in a protein [40] ) also indicated lack of ubiquitination at the 51st Lysine . Thus , viruses containing these sequences should produce a proteasome resistant ORF4 protein , similar to K51N mutation . Despite lacking ORF4 , g-3 HEV replicates better than g-1 virus in mammalian cell culture [12] . We hypothesised that some host protein ( s ) might be substituting the function of ORF4 in the g-3 virus , allowing it to bypass the dependency on ER stress dependent synthesis of ORF4 . Host proteins identified as g-1 RdRp interaction partners ( Table 3 ) were tested for their ability to associate with g-3 RdRp ( Table 4 ) . Only 14 out of 21 g-1 RdRp interaction partners associated with g-3 RdRp ( Table 4 ) , indicating that g-3 RdRp interaction profile is different from that of its g-1 counterpart . Therefore , it may interact with additional host proteins that did not interact with g-1 RdRp . We further tested the direct and indirect interactions of g-3 RdRp with other proteins of g-3 HEV by Y2H and CoIP assays . No intra-viral interaction partner of g-3 RdRp could be detected in Y2H assay , in agreement with the data obtained for g-1 RdRp ( S2 Table ) . However , CoIP of Huh7 cell extract expressing both g-3 RdRp and g-3 X revealed that both of them coprecipitated with each other , indicating an interaction between them ( Fig 11A ) . No interaction was observed between g-3 RdRp and helicase or g-3 X and helicase ( Fig 11B and 11C ) . Next , a CoIP assay was performed to assess whether g-3 RdRp , X and helicase could assemble a complex . Indeed , all three could be coprecipitated ( Fig 11D ) , indicating that they remain associated with each other . These findings also suggest that some host factor is essential for bridging the interaction between g-3 RdRp , X and helicase .
The current study attempts to address a long standing issue for researchers in HEV biology , pertaining to poor replication of g-1 HEV in cell culture . We show through multiple experiments that a previously unknown viral protein , which we have named ORF4 , is essential for proper functioning of RdRp of g-1 HEV . Because ORF4 is synthesized only under condition of ER stress and it is a short-lived protein , replication of viral genome is inefficient in normal cells . Thus , it appears that ER stress , which is probably initiated as an antiviral response by the host , turns out to be the ideal cellular condition for optimal replication of g-1 HEV . Such a mechanism seems to be remarkably suited to the life of the virus , given that individuals under stress such as pregnant women display enhanced sensitivity towards HEV infection . These findings also suggest that a diverse range of diseases which induce hepatic ER stress , may sensitize individuals towards g-1 HEV infection . A study involving clinical assessment of hepatic stress , quantitation of ORF4 expression and viral titre in liver biopsy of different categories of g-1 HEV patients may unravel the correlation between ER stress , degree of ORF4 expression and disease severity . Interestingly , ORF4 is encoded only by the g-1 HEV . Our bioinformatics analysis did not predict the presence of ORF4 in other HEV genotypes and experimental analysis of g-3 HEV replicon ruled out the possibility of ORF4 expression by g-3 HEV . Moreover , ER stress inducing compounds tunicamycin and thapsigargin did not have any effect on g-3 HEV replication . The above observations gave rise to two important questions: ( a ) Does ORF4 really play an important role during the natural course of g-1 HEV replication , ( b ) If ORF4 is indispensible for g-1 HEV replication , how do other genotypes of HEV replicate in its absence . To answer the first question , we analysed all available g-1 HEV genome sequences and not only observed the presence of ORF4 but also observed a very high level of conservation of ORF4 protein sequence among all g-1 HEV isolates ( see Fig 1E ) . Though the C-terminal 19 amino acids of ORF4 were absent in ~50% of the viral genomes , our experimental data demonstrate that these 19 amino acids are dispensible for known functions of ORF4 . The N-terminal 124 amino acids of ORF4 are sufficient for interacting with other viral and host proteins ( see Fig 6E ) and in vitro , it is able to stimulate RdRp activity just like the full length ORF4 ( see Fig 8E ) . Therefore , ORF4 seems to be indispensable for g-1 HEV life cycle . Since our study indicates that ORF4 most likely acts by interacting with multiple viral and host proteins to assemble a replication complex and promotes g-1 RdRp activity by interacting with host eEF1α1 , in order to understand how other genotypes of HEV replicate in the absence of ORF4 , we compared the protein interaction profile of g-1 RdRp with that of g-3 RdRp . Though g-1 and g-3 RdRp share ~85% homology at the amino acid level , only 14 out of 21 g-1 RdRp interacting host proteins could interact with g-3 RdRp . Moreover , g-3 RdRp interacted with g-3 X protein in Huh7 cells , indicating that some host protein ( s ) bridges that interaction . Further , g-3 RdRp , X and helicase assembled a complex , probably mediated by some host protein ( s ) . G-3 X and g-3 helicase may display differential interaction profile than their g-1 counterparts , as observed in the case of g-3 RdRp . Thus , it is worth speculating that host proteins substitute the function of ORF4 in the case of g-3 HEV . Host protein interaction profile of g-3 RdRp , X and helicase needs to be established to identify the host proteins involved in assembling g-3 RdRp , X and helicase complex . Nevertheless , our data provides evidence for the acquisition of a regulatory system that enhances replication in g-1 HEV as it is not seen in any other genotype . ER stress independent constitutive assembly of the viral replication complex might account for the observed better replication efficiency of g-3 virus in mammalian cell culture . Investigation of the mechanism ( s ) driving ORF4 synthesis revealed that it is independent of ORF1 translation , which is cap-dependent . Subsequent studies led to the discovery of a RNA regulatory element , which mediated cap-independent translation of ORF4 . In vitro as well as in dual luciferase reporter assays , the HEV regulatory element functioned efficiently irrespective of thapsigargin and tunicamycin treatment . However , it was active only under conditions of ER stress in its natural location in the HEV genome , probably because the regulatory element is inaccessible or remains bound to inhibitory factors in the absence of ER stress . Several other viral and cellular IRESs are known to be active only under specific conditions . Notably , an IRES within Human immunodeficiency Virus-1 mediates viral structural protein synthesis during G2/M phase of cell cycle [27] and under conditions of oxidative stress [41] . Human cytomegalovirus latency protein pUL138 is translated by an IRES like element during serum stress [28] . A subset of cellular mRNAs such as c-Myc , Bip , Apaf-1 , p53 and XIAP are translated through IRESs under conditions of stress , hypoxia and/or in a cell cycle dependent manner [42] . Though our data demonstrates the coexistence of both cap-dependent and cap-independent modes of translation in g-1 HEV , we have designated the RNA regulatory element as “IRES-like” element because it does not closely resemble other well known IRESs except for weak homology with the ERAV-1 IRES . Identification of IRES trans acting factors and detailed understanding of the mechanism of translation mediated by this element would confirm whether it is a bonafide IRES . Nonetheless , current data adds g-1 HEV to the list of RNA viruses where both cap-dependent and independent modes of translation coexist . eEF1α1 ablation inhibited basal RdRp activity and antisense RNA synthesis whereas Tubβ ablation had no effect . Stimulatory effect of ORF4 on RdRp activity was also dependent on the level of eEF1α1 , indicating crucial role of the latter in viral replication . Notably , eEF1α1 is important for replication and encapsidation of many plant and animal RNA viruses [43] . It binds to RdRp of Tobacco mosaic virus and silencing it inhibits infection [44] . eEF1α1 interacts with p33 protein of Tombus virus and this interaction is essential for viral antisense RNA synthesis [45] . eEF1α1 appears to have a similar role in g-1 HEV antisense RNA synthesis . Our data demonstrates that ORF4 is degraded by the host proteasome . Early in life , virus focuses on replication and later on switches towards release of progeny . As lack of ORF4 dampens RdRp activity , it is possible that ORF4 performs two important functions in the life of g-1 HEV . At early phase , it promotes viral replication and later on , being a short lived protein , it acts as a regulatory switch to shift from replication to release . On the contrary , proteasomal degradation of ORF4 might also be an anti-viral strategy evolved in the host to restrict virus spread . Our limited analysis involving sequence analysis of 19 g-1 HEV isolates demonstrated conservation of the 51st Lysine . However , the ubiquitination site was lost in 7 sequences owing to alteration of 50th Proline to Leucine , suggesting that viruses in those patients produced a proteasome resistant ORF4 . It is noteworthy that 5 out of the 7 sequences were isolated from fulminant hepatic failure ( FHF ) patients and 2 were acute viral hepatitis patients . However , considering the very limited number of samples , it might be a coincidence that majority of them represented FHF cases . Experimental analysis of replication efficiency of these 7 genomes will further substantiate the role of proteasome resistant ORF4 in g-1 HEV replication . Furthermore , an elaborate study involving more patient samples should be conducted to confirm the above observation . Analysis of correlation between disease severity and appearance of stabilizing ( proteasome resistant ) mutations in the ORF4 would further establish its role as a pro-viral factor . Identification of stabilizing mutations in ORF4 seems to have an important practical application in developing a more efficient model of HEV . HEV genome harbouring the ORF4 K51N mutation , which displays a high replication efficiency , might be expressed in cell lines stably expressing viral capsid and ORF3 protein to generate a robust system for producing HEV in the laboratory . In conclusion , the present study provides yet another example of an opportunistic pathogen , which transforms the adversities imposed by the host towards its own benefit . Identification of ORF4 as an essential proviral factor , which is expressed only under conditions of ER stress , likely explains the inability of g-1 HEV to replicate efficiently in mammalian cell culture under standard laboratory condition . A proteasome resistant ORF4 harbouring HEV genome will be useful for establishing an efficient model of g-1 HEV . Our study also suggests that different HEV genotypes may have evolved different molecular mechanisms to exploit the host and successfully complete their life cycles .
HEV ORFs were PCR amplified from pSKHEV2 ( genbank: AF444002 . 1 ) or pSK HEV p6 luc ( genbank: JQ679013 . 1 ) plasmids and cloned into the required vectors following standard protocols [46] . HEV genomic RNA was in vitro synthesised , as described [13]; size and integrity was monitored by formaldehyde agarose gel electrophoresis . Huh7 human hepatoma cells were as described in Surjit et al . [4] and it was originally obtained from the laboratory of C . M . Rice [47] . HEK 293T cells were obtained from ATCC ( USA ) . Cells were maintained in Dulbecco’s modified Eagle medium ( DMEM ) containing 10% Fetal Calf Serum ( FCS ) , 50 I . U . /mL Penicillin and Streptomycin , in 5% CO2 . Cells were transfected using Lipofectamine 2000 or 3000 , following manufacturer’s protocol ( Life Technologies , USA ) or electroporated . shRNAs were designed using Oligoengine 2 . 0 software for cloning into pSUPER puro vector , following manufacturer’s guidelines ( Oligoengine , USA ) . Additional details in supplementary methods . Antibodies against Flag , GAPDH , Myc , Ubiquitin and actin gamma were from Santa Cruz Biotechnology ( USA ) . Antibodies against HA , eEF1α1 , Tubulin β were from Sigma ( USA ) . MG132 , thapsigargin , tunicamycin , cycloheximide and NH4Cl were from Sigma ( USA ) . Rabbit polyclonal antibodies against HEV ORF2 , Helicase and ORF4 were synthesised at Genscript ( USA ) and validated in our lab ( See supplementary methods ) . All chemicals were added 24 hours post transfection and maintained for 16 hours , or as indicated . Effective concentrations: MG132-25μM; cycloheximide-100μg/ml; thapsigargin-1μM; tunicamycin-10μg/ml; NH4Cl-30μM . HEV infected , acute liver failure serum samples were obtained from patients registered in the liver clinic of Department of Gastroenterology , All India Institute of Medical Sciences , New Delhi , India . Serum was also collected from two healthy individuals with informed consent . A total 57 samples were tested for anti HEV IgM by ELISA and viral RNA with nested semi quantitative RT-PCR and quantitative real time PCR ( primer sequences in S3 Table ) , respectively . Five samples were HEV IgM and g-1 HEV RNA positive . ORF4 coding and flanking region of these samples were sequenced and data submitted to genbank ( ID: KU168733-KU168737 ) . For testing cross-reactivity with purified ORF4 protein in western , serum was diluted 1:5000 , followed by incubation with 1:5000 diluted goat anti-human IgG HRPO ( Southern Biotech , USA ) . Viral genomic sequence of patient samples were aligned to HEV sequence ( AF444002 . 1 ) for comparison . ORF4 coding region nucleotide sequence of different HEV isolates was obtained from Genbank and translated into protein sequence using MacVector software . ClustalW alignment was done using MacVector . AF444002 . 1 sequence was considered as reference . Total RNA was isolated using TRI reagent ( MRC , USA ) , followed by reverse transcription ( RT ) and QRT-PCR , as described [48] . Random hexamers and HEVAS RP oligo were used in RT for detecting sense and antisense strands , respectively . Primer sequences are provided in S3 Table . Done as described [4] . Goat anti rabbit alexa Fluor 488 ( Molecular probes , USA ) secondary antibody was used in IFA . Nucleus was stained with 4’ 6’- diamino-2-phenylindole ( Antifade gold , Molecular probes ) . Images were acquired using a 60X objective in a confocal microscope ( Olympus FV1000 ) and analyzed by Fluoview software . Details in supplementary methods . A T7 polymerase based TNT kit ( Promega , USA ) was used for in vitro synthesis of proteins , following manufacturer’s instructions . A GAL4 based system ( Clontech , USA ) was used following manufacturer’s instructions . Briefly , Y2H gold strain was transformed using lithium acetate with required BD and AD plasmids , followed by replica plating of 8 random transformants on different selection media to evaluate the activity of reporters . ORF3-TSG 101 interaction was used as a positive control [4] . For screening the Y2H cDNA library of human fetal brain , g-1 RdRp was cloned into pGBKT7 vector and its self activation potential was evaluated in Y2H gold strain ( S4A Table ) . A mate and plate human fetal brain cDNA library ( Clontech , USA ) was used to screen the interaction partners of g-1 RdRp , following the instructions of the manufacturer . Mating condition and efficiency is mentioned ( S4B Table ) . From evaluation of the diploids obtained after mating to identification of the bonafide interaction partners is summarized ( S4C Table ) . All interactions were confirmed by retransformation of the prey and bait plasmids in pair along with appropriate negative controls ( Table 3 ) . Dual luciferase reporter constructs ( Firefly and Renilla , 1μg/well ) were transfected into HEK 293T cells at 70% confluency in 48 well plate using lipofectamine 2000 . Compounds were added for 16 hours , followed by luciferase assay using Dual Luciferase reporter assay kit ( Promega , USA ) . Firefly luciferase values were divided by that of renilla and plotted . Gaussia luciferase was measured from culture medium using renilla luciferase assay kit ( Promega , USA ) . Viability of same cells were measured using stable tetrazolium salt WST-1 ( Roche , USA ) . Gaussia values were normalised to that of cell viability and plotted . Values are mean ± SEM of three independent experiments done in triplicate . GST-ORF4 was expressed in E . Coli C-41 ( DE3 ) strain ( 0 . 1 mM IPTG , 18°C , 16 hours ) . Soluble protein was bound to Glutathione Sepharose beads , washed and eluted using 20mM glutathione . Eluted protein was Flag-affinity purified following manufacturer’s instructions ( Sigma , USA ) . Final protein was recovered in PBS . Flag-tagged ORF2 , helicase , RdRp and ORF4 were purified from Huh7 cells transiently expressing respective proteins by Flag-affinity purification . Silver staining was done using Pierce silver stain kit ( Thermo Scientific , USA ) . Glutathione Sepharose bound GST-ORF4 was mixed with equal amount of purified RdRp , X , Helicase and ORF3 in CoIP buffer [20mM Tris ( pH 7 . 4 ) , 150mM NaCl , 1mM EDTA ( pH 8 . 0 ) , 1mM EGTA ( pH 8 . 0 ) , 1% Triton X 100 , 2 . 5mM Sodium Pyrophosphate , 1mM β glycerol phosphate , 1mM sodium orthovanadate , protease inhibitor cocktail] and rotated overnight at 4°C . Beads were washed thrice in same buffer , bound proteins eluted in 20mM glutathione , followed by western blotting using indicated antibodies . Done as described [49] with the modification that the 16 base RNA oligo was labelled with 6 FAM ( 6-carboxyfluorecein ) . Additional details in supplementary methods . Done as described [50] with the modification that DIG-II-UTP was used . Additional details in supplementary methods . Data are presented as mean ± SEM of at least three independent experiments , analyzed using ‘‘GraphPad Prism” by the Student t test . p < 0 . 05 was considered significant . Peripheral Blood samples were obtained from HEV infected and healthy adults with informed consent . Written consent was obtained from each individual . The study protocol was approved by the Ethics committee of All India Institute of Medical Sciences , New Delhi , India .
|
Hepatitis E virus ( HEV ) is one of the most common causes of acute and sporadic viral hepatitis . It is a small positive strand RNA virus , which is transmitted through the feco-oral route . Owing to lack of sanitation and unavailibility of safe drinking water , populations of developing and resource starved countries are prone towards HEV infection . Recent reports also indicate HEV induced acute and chronic hepatitis in organ transplant patients . Another peculiar characteristic of HEV is attributed to its ability to cause high mortality ( ~30% ) in infected pregnant women . Even after 30 years of discovery of the virus , little information exists regarding viral life cycle and replication machinery . HEV is divided into seven genotypes . Genotype-3 and 4 viruses infect humans and a few animals ( such as pigs , deer , mongeese ) and have been reported from industrialized countries . Genotype-3 and 4 viruses have been successfully propagated in the laboratory in mammalian cell culture . However , genotype-1 virus , which is known to infect human and is a major public health concern in south Asian countries , replicates poorly in mammalian cell culture and no other efficient model exists to investigate the viral life cycle . With the goal of developing an efficient laboratory model of genotype-1 HEV , we attempted to identify a permissive cellular condition that would allow efficient viral replication in human hepatoma cells . Here , we report that endoplasmic reticulum stress inducing agents promote genotype-1 HEV replication by initiating cap-independent , internal translation mediated synthesis of a novel viral factor , which we have named ORF4 . Further investigations revealed that ORF4 is expressed only in genotype-1 and it acts by interacting with multiple viral and host proteins and cooperates with host eEF1α1 ( eukaryotic elongation factor 1 isoform 1 ) to control the activity of viral RNA dependent RNA polymerase . Moreover , a proteasome resistant ORF4 mutant significantly enhanced viral replication . Thus , our study identifies an optimal condition required for efficient replication of genotype-1 HEV and dissects out the molecular mechanism governing that . Data presented here will be instrumental in developing an efficient model of the virus .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"nucleic",
"acid",
"synthesis",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"enzymes",
"pathogens",
"immunology",
"enzymology",
"microbiology",
"viruses",
"immunoprecipitation",
"rna",
"synthesis",
"antibodies",
"chemical",
"synthesis",
"research",
"and",
"analysis",
"methods",
"immune",
"system",
"proteins",
"antisense",
"rna",
"proteins",
"medical",
"microbiology",
"gene",
"expression",
"microbial",
"pathogens",
"hepatitis",
"viruses",
"viral",
"replication",
"precipitation",
"techniques",
"biosynthetic",
"techniques",
"biochemistry",
"rna",
"helicases",
"nucleic",
"acids",
"virology",
"viral",
"pathogens",
"physiology",
"genetics",
"protein",
"translation",
"biology",
"and",
"life",
"sciences",
"organisms",
"hepatitis",
"e",
"virus"
] |
2016
|
Endoplasmic Reticulum Stress Induced Synthesis of a Novel Viral Factor Mediates Efficient Replication of Genotype-1 Hepatitis E Virus
|
Highly phagocytic macrophages line the marginal zone ( MZ ) of the spleen and the lymph node subcapsular sinus . Although these macrophages have been attributed with a variety of functions , including the uptake and clearance of blood and lymph-borne pathogens , little is known about the effector mechanisms they employ after pathogen uptake . Here , we have combined gene expression profiling and RNAi using a stromal macrophage cell line with in situ analysis of the leishmanicidal activity of marginal zone macrophages ( MZM ) and marginal metallophilic macrophages ( MMM ) in wild type and gene targeted mice . Our data demonstrate a critical role for interferon regulatory factor-7 ( IRF-7 ) in regulating the killing of intracellular Leishmania donovani by these specialised splenic macrophage sub-populations . This study , therefore , identifies a new role for IRF-7 as a regulator of innate microbicidal activity against this , and perhaps other , non-viral intracellular pathogens . This study also highlights the importance of selecting appropriate macrophage populations when studying pathogen interactions with this functionally diverse lineage of cells .
Mononuclear phagocytes are widely distributed in all tissues , and provide a broad range of homeostatic and immune functions during development and throughout adult life . Nevertheless , the heterogeneity of mature tissue macrophages represents one of the most striking , yet under-studied , features of mononuclear cell differentiation . Expression of a range of transcription factors and cellular receptors has helped define membership of the mononuclear phagocyte system [1] , [2] , [3] , and whereas some tissue macrophages have a capacity for local self-renewal , others are derived from blood-borne monocytes undergoing tissue-specific differentiation [reviewed in [4]] . In lymphoid tissues , macrophage heterogeneity is most strikingly evident . For example in mice , readily distinguishable populations of macrophages are found in the splenic MZ , red pulp , B cell follicles and white pulp [5] . Moreover , within the MZ , SIGNR-1+MARCO+ MZM occupy the outer rim adjacent to the red pulp , and CD169hi MMM border the lymphocyte-rich white pulp [5] , [6] , [7] , [8] , [9] , [10] . Mice lacking various transcription factors ( e . g . relβ and NFκB2 ) , TNF superfamily cytokines ( e . g . LTα ) , and chemokines ( e . g . CCL19/21ser ) exhibit steady-state defects in MZ macrophage differentiation and/or positioning [11] , [12] , illustrating the complexity behind this micro-anatomical organisation . MZM and MMM are well-placed in the marginal sinus to encounter blood borne antigens and pathogens . MZM are avidly phagocytic , are well-characterised as facilitating clearance of Streptococcus pneumoniae from the blood stream [13] , [14] , [15] and are involved in the initiation of Type I T-independent immune responses [16] . MMM , whilst also phagocytic [17] are better recognised for playing a role in antigen transport into B cell follicles [4] and are also known to be robust producers of IFN-α following infection with viral pathogens [18] . MMM , and to a lesser extent MZM , also express CD169 , a sialoadhesin first described on bone marrow ‘stromal’ macrophages involved in the support of erythropoiesis [19] and these cells may provide this or other stromal functions in the spleen . Similarly , in lymph nodes , subcapsular sinus macrophages mark a point of entry for viruses and bacteria entering via the afferent lymphatics and are intimately involved in the initiation of antibody responses [20] , [21] , [22] . Thus , various specialised macrophages populations are ideally situated within secondary lymphoid tissues to first encounter pathogens that enter the blood and lymphatics . Nevertheless , despite their importance in providing a ‘gatekeeper’ function , little is known about how these tissue resident macrophages subsequently deal with the pathogens they encounter . Direct study of MZM and MMM has been problematic and only rarely reported [6] , largely because of the inherent technical difficulties in isolating these scarce and fragile cells for functional analysis in vitro . Hence , our understanding of the effector functions of MZM and MMM is largely extrapolated from that of other diverse macrophage populations , such as those derived from the peritoneal cavity or grown from bone marrow precursors under the influence of CSF-1 . In this study , therefore , we set out to directly identify the mechanism by which splenic MZM and MMM are able to kill intracellular Leishmania parasites . We first identified a stromal macrophage cell line ( 14M1 . 4; [23] ) that displayed similar innate capacity to kill L . donovani as observed in MZM and MMM . By gene expression profiling , we showed that Leishmania infection stimulated expression of the transcription factor Irf-7 , as well as redistribution of cellular IRF-7 from MyD88+ endosomes to the Leishmania-containing phagosome . Initial control of intracellular Leishmania by 14M1 . 4 cells involved NO and was minimally affected by silencing of Irf-7 . However , subsequent leishmanicidal activity strictly required Irf-7 and was NO-independent . Furthermore , Irf-7 loss-of-function could not be overcome by exogenous IFN-α , suggesting a downstream effector mechanism uncoupled from Irf-7-dependent amplification of IFN-α . By immunohistochemistry of spleens from infected mice , we showed that MZM and MMM also responded to infection with heightened expression and phagosomal recruitment of IRF-7 . Finally , we showed that MZM and , to a lesser extent , MMM from Irf-7-deficient mice were unable to kill intracellular Leishmania . By focusing on these tissue-resident stromal macrophages , we have thus identified a new role for IRF-7 as a regulator of macrophage anti-leishmanial activity , and demonstrated that this transcription factor plays a role in innate effector responses to a broader range of intra-phagosomal pathogens than previously recognised .
Previous studies in BALB/c mice showed that after i . v . injection , L . donovani amastigotes are rapidly and selectively taken up by MZM and MMM , and that over the following 24h , parasite load in the MZ was significantly reduced [17] . We first confirmed this selectivity of uptake in the MZ of C57BL/6 mice , observing greater than 90% of amastigotes in the MZ ( within cells staining either for CD169 or SIGNR1 ) and few parasites in either the deep white pulp or in the red pulp ( Figure 1A ) . To confirm that C57BL/6 mice also expressed rapid leishmanicidal activity and to determine whether this might be mediated by NK cell-derived IFN-γ [24] , we treated mice with either control IgG or anti-IFN-γ . Both control and anti-IFN-γ-treated mice reduced parasite load to a similar extent , indicating that this response was largely IFN-γ independent ( Figure 1B ) . We then screened a number of macrophage cell lines ( RAW264 . 7 , J774 . 1 , 14M1 . 4 ) and primary macrophage populations ( bone marrow-derived , peritoneal ) for their ability to kill L . donovani amastigotes in the absence of IFN-γ activation . Of the cells tested , 14M1 . 4 cells were striking in their ability to clear amastigotes over a 24–48h period . Infected 14M1 . 4 cells reduced parasite load by ∼50% from 6h–12h after infection and sustained this activity over the subsequent 36h , resulting in almost complete clearance of amastigotes by 48h ( Figure 1C ) . In contrast , although RAW264 . 7 reduced the number of intracellular amastigotes by ∼30–40% from 12h–24h , no further reduction in parasite load was detected after this time . Differences in amastigote numbers were not readily attributable to differential rates of cell division between 14M1 . 4 cells and other cells tested , with 14M1 . 4 and RAW264 . 7 cells showing similar levels of CFSE dilution over 48h ( Supplementary Figure S1A ) . Likewise , equivalent levels of CFSE dilution were observed in intracellular amastigotes isolated from both cell populations , indicating that parasite growth rate was similar in these two host cells ( Supplementary Figure S1B ) . Thus , these data collectively demonstrated that 14M1 . 4 cells were intrinsically very efficient at killing intracellular amastigotes . To understand further the mechanism ( s ) by which 14M1 . 4 cells killed intracellular amastigotes , we used global gene expression profiling to examine the transcriptional response of 14M1 . 4 cells to L . donovani infection . 977 and 851 genes were found to be differentially expressed in infected vs . control 14M1 . 4 cells at 12h or 48h p . i . respectively , using a cut-off of a 2-fold change in gene expression ( Supplementary Tables S1 and S2 ) . These differentially expressed genes were analysed by generating gene interaction networks using Ingenuity Pathways Analysis software . 10 networks of high interconnectivity ( ≥20 differentially regulated genes ) were constructed for genes differentially regulated at 12h p . i . and 12 hubs at 48h p . i ( Supplementary Figure S2 and S3 ) . In contrast , only 136 and 161 genes were differentially regulated in RAW264 . 7 cells , at 12h and 48h p . i . respectively ( Supplementary Tables S3 and S4 ) , and only one hub with high interconnectivity could be identified at each time point ( Supplementary Figure S2 and S3 ) . In 14M1 . 4 cells , the highest scoring networks reflected genes involved in interferon signalling ( notably Irf-7 , Irf-1 , Stat-2 ) , cytokines ( notably Il-6 ) and cardinal interferon response genes ( including mx-1 , Rsad and various MHC and MHC related genes: Figure 2A and B ) . At 48h , 31% of genes ( 9/29 ) listed in the Ingenuity Pathways ‘Interferon Signalling’ pathway were represented in the differentially induced gene set , all being upregulated . Within the ‘Antigen Presentation Pathway’ , 25% ( 10/39 ) were either up- or down-regulated , with a marked differential response between MHCI-and MHCII-associated genes ( Figure 2B ) . A range of genes within these networks and associated with pathogen recognition and handling were also down-regulated including Irf-8 , recently associated with the innate control of intra-phagosomal pathogens [25] , Cx3cr1 , a chemokine receptor involved in the NO-dependent bactericidal activity of macrophages [26] , Clec7a ( Dectin-1 ) and Clec4a ( Mincle ) , both associated with fungal recognition [27] , [28] and Msr1 , the macrophage scavenger receptor and its cytoplasmic partner Hook3 [29] , [30] . Sc5d , associated with cholesterol synthesis [31] was also down-regulated , in keeping with data indicating that Leishmania infection depletes this membrane lipid [32] , [33] , as was Emr1 ( F4/80; [34] ) . TGFβ and a network of interconnected genes were also specifically down-regulated at 48h in 14M1 . 4 cells ( Supplementary Figure S3 ) . We next performed clustering using the CLICK algorithm contained in EXPANDER [35] which identified 550 genes in a single cluster ( homogeneity value: 0 . 895 ) that were increased in their expression level at both 12h and 48h p . i . ( Figure 2C ) . GO functional classes representing ‘immune response’ ( GO:0006955; 5 . 45% ) and ‘defense response’ ( GO: 0006952; 0 . 54% ) were significantly enriched in this cluster . We next used PRIMA software [36] to identify transcription factor sequences enriched amongst these 550 induced genes . By this analysis ( Figure 2D ) , we identified Irf-7 and Irf-1 and ISRE , as potential regulators of the response of 14M1 . 4 cells to L . donovani . Similar analysis of genes down regulated in 14M1 . 4 cells at both time points did not reveal any significantly enriched transcription factors ( data not shown ) . Based on the high levels of expression in 14M1 . 4 cells , absence of expression of an interferon response signature in RAW264 . 7 cells , and the known role of Irf-7 as a central regulator of the Type I interferon response [37] , [38] , we focused further attention on Irf-7 . Confirmation of the induction of Irf-7 following Leishmania infection was obtained by quantitative RT-PCR ( Figure 2E ) , by intracellular flow cytometry ( Figure 2F ) and by confocal microscopy ( Figure 2G ) . Amastigotes induced a dose- and time-dependent accumulation of Irf-7 mRNA and IRF-7 protein in 14M1 . 4 cells that was similar in kinetics but of somewhat reduced magnitude than poly ( I:C ) a known Toll-like receptor ( TLR ) -3-dependent inducer of Irf-7 . In contrast , RAW264 . 7 cells and bone marrow-derived macrophages induced minimal levels of Irf-7 mRNA accumulation or IRF-7 protein expression ( Supplementary Figure S4 ) . Together , these data indicated that 14M1 . 4 stromal macrophages , but not RAW 264 . 7 macrophages responded to L . donovani infection by strong induction of IRF-7 . In IRF-7/MyD88 co-transfected RAW264 . 7 and HEK 293 cells and in plasmacytoid DC , IRF-7 and MyD88 co-localise in endosomes [39] . To establish whether such co-localisation also occurred in un-manipulated 14M . 1 . 4 cells and to gain further insight into IRF-7 activation in these cells , we examined the distribution of MyD88 and IRF-7 by confocal microscopy ( Figure 3 ) . In resting 14M1 . 4 cells , IRF7 and MyD88 were largely co-localised in small cytoplasmic vesicles , with little evidence of nuclear IRF-7 ( Figure 3A; Video S1 , shown as a snapshot in Figure 3E ) . Following poly ( I:C ) activation , nuclear translocation of IRF-7 occurred , with residual cytoplasmic IRF-7 remaining associated with MyD88 containing vesicles ( Figure 3B ) . In sharp contrast , although 14M1 . 4 cells infected for 48h with L . donovani also had nuclear IRF-7 , cytoplasmic IRF-7 was now primarily associated with phagosomes containing Leishmania , and rarely with MyD88+ vesicles ( Figure 3C ) . By 48h p . i . , almost all L . donovani phagosomes were IRF-7+ ( Figure 3D ) . IRF-7 was almost undetectable in RAW264 . 7 macrophages , with or without infection , but where low levels of IRF-7 could be seen in rare cells , an association with the L . donovani phagosome was not apparent ( Supplementary Figure S4 ) . Of interest , IRF-7 in infected 14M1 . 4 cells was almost always localised to one pole of the phagosome ( Figure 3C insert; Video S2 , shown as a snapshot in Figure 3F ) . The recruitment and polarisation of IRF-7 in the Leishmania phagosome was specific , based on three sets of observations . First , MyD88 was not similarly detected at this location ( Figure 3C and F ) . Second , in 14M1 . 4 macrophages treated with poly ( I:C ) and then exposed to latex beads for 48h , no phagosomal recruitment of IRF-7 was observed in spite of high levels of induction of IRF-7 ( Figure 3G ) . Third , TLR3 accumulating on latex and Leishmania phagosomes did not show the polar distribution seen with IRF-7 on Leishmania phagosomes ( Figure 3G and H ) . With TLR3 used to demarcate the phagosome membrane , IRF-7 appeared to have an intra-phagosomal localisation ( Figure 3H ) , an observation that was supported by using LAMP1 as an alternative phagosome membrane marker ( Figure 3I and Video S3 ) . Although some co-localisation with LAMP1 was observed , a significant amount of the observed IRF-7 was clearly intra-phagosomal , either on , within or tightly apposed to the amastigote itself . In summary , these data demonstrate that dissociation of endosomal MyD88 and IRF-7 is followed by nuclear and phagosomal recruitment of IRF-7 . To determine whether the IRF-7 response in 14M1 . 4 cells actually reflected the response of MZM and MMM to Leishmania infection in situ , we prepared tissue sections from control and L . donovani-infected mice , and analysed IRF-7 expression ( Figure 4 ) . IRF-7 expression was minimal in MMM ( Figure 4A ) and MZM ( Figure 4B ) of naïve mice , but readily observed at 5h ( data not shown ) and 24h in MMM ( Figure 4C ) and MZM ( Figure 4D ) of infected C57BL/6 mice . At 5h p . i . , ∼90% and at 24h p . i . ∼65% of parasite-containing phagosomes had clearly associated IRF-7 staining ( Figure 4E ) , and although it was difficult to always assess the subcellular distribution of IRF-7 , a polar distribution , similar to that seen in 14M1 . 4 cells in vitro , was often observed ( Video S4 , shown as a snapshot in Figure 4F ) . Collectively , these data support the contention that the early leishmanicidal activity of splenic MZ macrophage populations is also accompanied by IRF-7 induction and recruitment of IRF-7 to the amastigote-containing phagosome . To directly determine whether IRF-7 induction was involved in regulating the leishmanicidal activity of 14M1 . 4 cells , we performed functional gene knock down of IRF-7 using RNAi . Knock-down clones were screened by RT-PCR ( Figure 5A ) and immunofluorescence ( Figure 5B–E ) in the presence or absence of poly ( I:C ) , and two independent clones with strong inhibition of Irf-7 mRNA accumulation were selected for further study ( KD#1 and KD#2 ) . Analysis of the fate of L . donovani in Irf-7 knock-down compared to control cells indicated that control of parasite burden was biphasic with respect to the requirement for Irf-7 . Amastigote numbers were reduced to a similar extent over the first 12h of infection in all cells tested , suggesting an early phase of leishmanicidal activity that was Irf-7-independent . In contrast , sustained leishmanicidal activity , measured over 12–48h , was completely abrogated by knock-down of Irf-7 ( Figure 5F ) . Thus , the leishmanicidal response of 14M1 . 4 cells consisted of an immediate Irf-7-independent component and a late-acting Irf-7-dependent component ultimately responsible for amastigote clearance . IRF-7 plays a central role in the amplification of the Type I IFN response [37] and NO has been reported to have a major role as a leishmanicidal effector in macrophages exposed to either Type I IFNs [40] or IFN-γ [41] . We therefore determined whether NO and IFN-α were involved in the killing of Leishmania by 14M1 . 4 cells . Although only minimal amounts of NO could be detected in the supernatants from infected 14M1 . 4 cells ( Supplementary Figure S5 ) , treatment of 14M1 . 4 cells with the Nos2 inhibitor aminoguanidine ( AMG ) blocked their ability to rapidly kill L . donovani as measured over the first 12h p . i . ( Figure 5G ) . Likewise , AMG also blocked the early leishmanicidal activity of Irf-7-deficient 14M1 . 4 cells ( Figure 5H ) . In contrast , AMG had a minimal effect on the late phase of leishmanicidal activity in 14M1 . 4 cells ( with parasite burden continuing to decrease steadily from 12–48h ) or in Irf-7-deficient 14M1 . 4 cells ( with amastigote numbers remaining at similar levels or increasing slightly from 12–48h; Figure 5G , H ) . Thus , whereas NO played a role in immediate parasite killing , it had limited involvement in the subsequent Irf-7-dependent phase of leishmanicidal activity . Next , we exposed wild type and knock down 14M1 . 4 cells to exogenous IFN-α . At early time points ( <12h ) , exogenous IFN-α significantly enhanced the rate at which wild type , but not Irf-7-deficient , 14M1 . 4 cells killed L . donovani ( Figure 5G , H ) , suggesting that the early augmentation of effector function mediated by exogenous IFN-α required the presence of Irf-7 . Importantly , however , at later time points ( 12–48h ) exogenous IFN-α was unable to overcome the loss-of-function associated with Irf-7-deficiency ( Figure 5H ) . Thus , leishmanicidal activity in 14M1 . 4 macrophages involves i ) an early NO-dependent and Irf-7-independent component , which can nevertheless be augmented by IFN-α in an Irf-7-dependent manner and ii ) a late-acting NO-independent , Irf-7-dependent component which can not be compensated for by exogenous IFN-α . The existence of an Irf-7-dependent leishmanicidal pathway in 14M1 . 4 cells led us to investigate whether innate killing of Leishmania in the marginal zone was similarly controlled by IRF-7 . In contrast to what has been observed with other interferon regulatory factor-deficient mice , no obvious structural differences were observed in the organisation and cellular content of the marginal zone in Irf-7−/− mice before or after infection ( Figure 6A–F ) . However , in contrast to B6 mice , B6 . Irf-7−/− mice failed to eliminate intracellular amastigotes over the first 24h of infection ( Figure 6G ) . To determine whether Irf-7-dependent killing was a property of both MMM and MZM , we separately scored the change in amastigote number within these two populations ( Figure 6H ) . In B6 mice , MMM and MZM both significantly reduced their parasite load from 5h to 24h ( by 62% and 37% respectively; p<0 . 001 ) . In Irf-7−/− mice , although MMM were still able to reduce amastigote load by ∼33% ( P<0 . 001 ) , killing activity was clearly reduced in comparison to MMM in wild type mice ( p<0 . 001 ) . For MZM , the impact of Irf-7-deficiency was more striking , with Irf-7−/− MZM failing to display any killing activity . Thus , IRF-7 played an essential role in regulating the innate ability of splenic tissue macrophages to kill L . donovani in vivo . In the course of these studies , we also noted that in B6 . Irf-7−/− , but not in B6 mice , amastigotes could be found more readily within red pulp macrophages ( Figure 6E and F ) , suggesting that the barrier function of the marginal zone that prevents early amastigote dissemination may also be impaired by Irf-7 deficiency .
MZM and MMM have previously been shown to be involved in the uptake of various blood borne pathogens , and in this way provide an important barrier to systemic infection [7] , [10] , [13] , [14] , [15] , [17] , [18] . Although macrophages in the marginal zone have been shown to be capable of IFN-α production [9] , [18] , their broader effector function has been largely inferred from the in vitro study of other macrophage populations . Here , using a comparative approach involving direct in vitro and in vivo analysis of microbicidal activity , we provide evidence that IRF-7 plays a central role in regulating the ability of MZM and MMM to control the intra-phagosomal pathogen , L . donovani . Although IRF-7 has long been acknowledged to orchestrate anti-viral immune responses , a role in macrophage effector responses to non-viral pathogens has not been previously described . IRF-7 has been shown to co-localise with MyD88 in endosomes of IRF-7 and MyD88 transfected cell lines [39] , an observation we confirmed here using non-transfected 14M1 . 4 cells . Our studies extend these observations by showing that upon Leishmania infection , the spatiotemporal link between MyD88 and IRF-7 was lost . Whereas MyD88 remained associated with small cytoplasmic vesicles , IRF-7 was selectively recruited to Leishmania-containing phagosomes , both in vitro in 14M1 . 4 cells and in vivo in infected MZM and MMM . This recruitment of IRF-7 was specific and not observed with latex bead phagosomes , even in cells also containing Leishmania or in cells highly activated for IRF-7 expression using poly ( I:C ) . It remains to be formally demonstrated whether this change in IRF-7 localisation reflects i ) dissociation of IRF-7 from existing complexes with MyD88 and subsequent re-location to the phagosome , or ii ) whether such IRF-7 undergoes nuclear translocation , with only de novo synthesised IRF-7 being recruited to phagosomes . Based on the short half life of IRF-7 [42] and the kinetics of mRNA accumulation described here , we currently favour the latter possibility . Interestingly , the absence of detectable MyD88 on the Leishmania phagosome , along with the absence of phagosomal TLR 9 expression ( Phillips et al , unpublished observations ) suggests that MyD88-dependent phagosomal signaling may not be a significant feature of Leishmania infection in these cells , at least at the time points studied . Similarly , it remains to be determined whether phagosomal IRF-7 plays any functional role in the IRF-7-dependent leishmanicidal activity observed here . On the one hand , phagosomal recruitment of IRF-7 might lead to IRF-7 forming part of a signalosome complex at the phagosome membrane , as has been described for phagosome-recruited AP-1 [43] . On the other hand , our confocal analysis suggests that the bulk of phagosomal IRF7 is intra-phagosomal ( as defined by LAMP1 ) and likely to be associated with the amastigote itself . This observation is strikingly similar to that made by Antoine and colleagues who demonstrated acquisition of phagosomal MHCII ( and not LAMP1 ) by L . amazonensis amastigotes [44] and suggests that the amastigote might act as a ‘sink’ that accumulates phagosome-associated IRF-7 . These two possibilities are not mutually exclusive , and we do indeed see low levels of IRF-7 co-localized with LAMP1 apparently in the phagosome membrane . Nevertheless , given that the amastigotes of L . donovani also have tight attachment sites to the phagosomal membrane [45]; and Phillips and Kaye , unpublished ) , future studies employing immunogold EM , cell fractionation and biochemical approaches will be required to determine the exact localization of IRF-7 in the L . donovani phagosome and whether such IRF-7 plays a functional role . The mechanism of killing we observe in 14M1 . 4 cells , in MZM and , to a lesser extent , MMM appears distinct from that previously reported for Leishmania studied in other macrophage populations . Innate IFN-γ-independent control over L . donovani in hepatic and splenic macrophages is a function often attributed to Slc11a1 [46] . However , this phagosomal iron transporter exerts leishmanistatic control over amastigote growth and , in any case , 14M1 . 4 cells carry the mutant allele of Slc11a1 . Similarly , expression profiling suggests that a number of other recently identified anti-microbial pathways are down-regulated in 14M1 . 4 cells after Leishmania infection . Although a role for NO in IFN-γ-independent killing of Leishmania has also been established [41] , our data indicate that the major phase of Irf-7-dependent killing occurs in a largely NO-independent manner . That previous studies have failed to identify this pathway may reflect the stromal origin of 14M1 . 4 cells and their closer relationship to resident tissue macrophages such as MZM and MMM . For example , RAW264 . 7 cells , a commonly used cell line in the study of Leishmania-macrophage interactions , responds minimally to infection with IRF-7 induction , and hence this pathway would not be evident in studies using this host-pathogen combination . Indeed , our data emphasize the importance of studying appropriate tissue macrophages when aiming to identify physiologically relevant effector mechanisms . Importantly , our data do not argue against a role for IFN-α in the regulation of Irf-7 [42] . IFN-α could be produced either through an autocrine pathway ( as with 14M1 . 4 cells in vitro ) or through a combination of autocrine and paracrine pathways , as is more likely to affect MZM and MMM in vivo . Indeed , plasmacytoid DC ( pDC ) , a major source of IFN-α are readily activated after Leishmania infection [[47] and Sanos and Kaye , unpublished ) . Rather , our finding that exogenous IFN-α could not compensate for Irf-7 loss-of-function provides evidence for an alternative downstream effector mechanism that cannot be amplified by IFN-α alone . Further studies to identify potential phagosomal effectors regulated by Irf-7 are currently in progress . Our studies were not designed to directly address the question of recognition of L . donovani during the process of early phagocytic uptake . Nevertheless , recent work has suggested that TLR9 is involved in the recognition of L . infantum and L . major promastigote-derived genomic DNA by CD11c+ cDC and by pDC [48] . In these published studies , however , no attempt was made to determine the immediate consequences of this recognition for the subsequent control of Leishmania survival . Using RAW264 . 7 macrophages , Descoteaux and colleagues suggested a role for both TLR2 and TLR3 in the phagocytic uptake of L . donovani promastigotes and a role for TLR3 in the leishmanicidal activity of the IFN-gamma-primed macrophages . However , as reported here , RAW264 . 7 cells do not recapitulate the IRF-7 response observed in 14M1 . 4 cells , MZM and MMM and these cells show only limited leishmanicidal activity in the absence of IFN-γ activation . Nevertheless , TLR-3 was induced in L . donovani-infected , as well as poly ( I:C ) treated 14M1 . 4 cells , but unlike IRF-7 , TLR-3 appeared to be promiscuously recruited to phagosome membranes , at least after IRF-7 activation . CD8+ T cell responses are of notable importance for protection against L . donovani , and vaccines which target CD8+ T cells are particularly effective against this infection in experimental models [49] , [50] . Although the role of tissue resident macrophages as antigen presenting cells has recently returned to the fore , with a variety of studies demonstrating antigen capture and presentation to B cells [13] , the importance of these cells for the induction of T cell immunity remains controversial [10] . Early studies suggested that innate Type I IFN responses were associated with an increased capacity for cross-priming [51] , [52] and might therefore favour the induction of CD8+ T cell responses to intracellular pathogens . However , whilst a recent systems biology analysis of the response to Yellow Fever vaccine has confirmed the induction of a Type I IFN gene signature centred on IRF-7 , this signature was not directly predictive of the magnitude of the CD8+ T cell response [53] . Previous studies of MHC gene expression in Leishmania-infected macrophages have reported inhibition of basal and IFN-γ-induced MHCI and MHCII gene expression in BALB/c derived peritoneal macrophages [54] . However , in the presence of the strong Irf-7-mediated response associated with infection of 14M1 . 4 cells , we observed a reciprocal regulation of genes related to MHCI- and MHCII-restricted antigen presentation , with elevated levels of transcripts for MHCI , TAP1 and tapasin . A similar role for IRF-7 in regulating cross-presentation has also recently been confirmed following adenoviral gene delivery of Irf-7 into peritoneal macrophages [55] . Together , these data suggest that further examination of the role of stromal macrophages , as represented by 14M1 . 4 cells and macrophages of the splenic marginal zone , in CD8+ T cell priming is now warranted . In conclusion , we have shown that IRF-7 is indispensable for effective control of L . donovani amastigotes in stromal macrophages in vitro and in vivo . Our results suggest that IRF-7 may play a similarly important role in the containment of other intracellular pathogens .
C57BL/6 mice ( CRUK , Margate , UK ) were used at 6–10 weeks of age and housed under specific-pathogen free conditions . B6 . Irf-7−/− mice were obtained from the RIKEN BioResource Center ( Ibaraki , Japan ) with kind permission of T . Taniguchi , University of Tokoyo , Japan . Amastigotes of the Ethiopian LV9 strain of L . donovani were maintained by passage in B6 . RAG1−/− mice . Mice were infected with 2–5×107 amastigotes intravenously in 100µl RPMI 1640 ( Gibco , Paisley , UK ) and killed by cervical dislocation as required . All experiments were approved by the University of York Animal Procedures and Ethics Committee and performed under UK Home Office license ( ‘Immunity and Immunopathology of Leishmaniasis’ Ref # PPL 60/3708 ) . RAW 264 . 7 macrophages ( American Type Culture Collection , Rockville , MD ) and 14M1 . 4 stromal macrophages ( [23]; a gift from D . Zipori , Weizmann Institute of Science , Rehovot , Israel ) were maintained in complete Dulbecco's modified Eagle's medium ( DMEM; Gibco ) containing 5% FCS supplemented with 1 mM sodium pyruvate , 2mM l-glutamine , 0 . 05 mM 2-mercaptoethanol , 100U/ml penicillin and 100µg/ml streptomycin ( all Gibco , Paisley , UK ) . 14M1 . 4 cells express low levels of SIGNR1 but undetectable levels of MARCO and CD169 ( data not shown ) . 14M1 . 4 cells deficient in Irf-7 were generated by transfection with lentiviral particles encoding shRNA for Irf-7 mRNA ( SHVRS-NM_016850 , Sigma-Aldrich , St . Louis , MO ) . Transfected cells were positively selected in complete media containing 5µg/ml puromycin ( Sigma-Aldrich , St Louis , MO ) and clones were generated by limited dilution . Irf-7 mRNA knock-down was screened by RT-PCR and immuno-staining . Before use in infection experiments , cells were collected by scraping with cell scraper ( BD Falcon ) and 5×105 cells in complete DMEM were added to 24 well plates ( Corning Inc . Corning , NY ) containing sterilised 13mm diameter coverslips ( VWR International , Leics , UK ) . After 24h , amastigotes were added at a multiplicity of infection ( MOI ) of 10∶1 ( in 300µl complete DMEM ) . After 1h at 37°C , non-adherent cells and free amastigotes were removed and cells were cultured for various times post infection ( p . i . ) . In some instances , 2µm latex beads ( Bangs Laboratories , Fishers , IN ) were used in addition to or instead of amastigote infections . Cell lines were treated for 2h with IFN-α ( 1000U/ml; PBL Interferon Source , Piscataway , NJ ) at the time of infection . To quantify infection level , cells were fixed and permeabilised using 4% paraformaldehyde + 0 . 1% Triton-X 100 ( BDH ) and coverslips were mounted in Vectashield with DAPI ( Vector laboratories , Burlingame , CA ) . The proportion of cells infected and the number of amastigotes per 100 host cells was calculated from triplicate cultures per treatment group . 106 control or infected 14M1 . 4 and RAW264 . 7 cells were incubated for 12 and 48 hours in T25 tissue culture flasks . Samples were prepared in triplicate on three separate occasions , with an average infection level of 38±5% for 14M1 . 4 cells and 36±8% for RAW264 . 7 cells . RNA was extracted from the cell lines using Trizol according to the manufacturer's protocol ( Sigma-Aldrich ) . One of the triplicate samples from each of the three experiments were used for microarray analysis . RNA concentration and integrity was established using 210 Bioanalyser ( Agilent Technologies , Palo Alto , CA ) . Extracted RNA was reversed transcribed to cDNA using the Affymetrix GeneChip one-cycle target labelling kit ( Affymetrix , Santa Clara , CA ) according to the manufacturer's recommended protocols and hybridised to GeneChip® Mouse Genome 430 2 . 0 Genome Arrays . Raw data processing was performed by using the Affymetrix GCOS 1 . 2 software . After hybridization and scanning , probe cell intensities were calculated and summarized for the respective probe sets by means of the MAS5 algorithm . MAS5 normalised data were collected and analyzed by using the ArrayAssist Expression software , Version 5 . 5 ( Stratagene ) . Differentially expressed genes were identified by using a two-class t test where significance level was set at p > 0 . 05 . Genes that were >2 . 0 fold up- or down-regulated between groups were selected . Data from these studies have been deposited in the EBI ArrayExpress data base ( Accession #: E-MEXP-2554 ) . Gene lists were susbsequently analysed using Ingenuity Pathway Analysis ( Ingenuity Systems , Redwood City , CA ) and Expander ( Shamir et al . , 2005 ) software . Proliferation of macrophages and L . donovani amastigotes was assessed by analysis of the loss of CFSE ( Invitrogen , Paisley , UK ) fluorescence . Macrophages were incubated with 10µM CFSE in PBS for 10 minutes at 37°C , washed repeatedly , and then plated as above . At the appropriate times , macrophages were harvested and analysed for CFSE fluorescence . To assess amastigote proliferation , amastigotes labelled with CFSE as above prior to use for infection . At appropriate times p . i . , macrophages were lysed with 1 . 25mg/ml saponin ( BDH ) in PBS . Released amastigotes were isolated by centrifugation and labelled with 10µM cell tracker blue ( Invitrogen , Paisely , UK ) for 20 minutes at 37°C , followed by specific anti-L . donovani antibody ( 1∶50 dilution of heat inactivated serum from an infected hamster ) and then goat anti-hamster alexa fluor 647 antibody ( 10µg/ml; Invitrogen , Paisley , UK ) . CFSE dilution of macrophages and/or viable parasites was determined by flow cytometry using a cyan flow cytometer ( Dako ) . 20 , 000–50 , 000 events were analysed per sample at each time point ( n = 3 ) . Macrophage RNA was isolated using an RNeasy kit according to the manufacturer's instructions ( Qiagen , UK ) . RNA was reverse transcribed to cDNA using the first-strand cDNA synthesis kit according to manufacturer's instructions ( Invitrogen , Paisley , UK ) . Primers for the specific amplification of hypoxanthine phosphoribosyltransferase ( HPRT ) have been previously described ( 45 ) . Pre-designed primers were used for the specific amplification of Irf-7 and Mx1 ( PPM04696E and PPM05520A , respectively: Superarray , Frederick , MD ) . PCR conditions for all primers were 95°C ( 15s ) , 62°C ( 30s ) , and 60°C ( 30s ) for 40 cycles . Real-time PCR was performed with the SYBR green PCR kit in an ABI Prism 7000 sequence detection system ( Applied Biosystems ) according to manufacturer's instructions . Relative expression of target genes was normalized to HPRT and was quantified by the formula 2-ΔΔCT where ΔΔCT = ( CT Target − CT HPRT ) Treated − ( CT Target − CT HPRT ) Control . Values are expressed as the mean of triplicates ± SE . Broken line indicates the expression of the untreated control to which all samples were calibrated . For IRF-7 protein expression , cells were fixed in 4% PFA and permeabilised on ice for 30 min with 0 . 1% Triton-X 100 buffer . Cells were incubated with a polyclonal rabbit anti-IRF-7 antibody ( Zymed , San Francisco , CA ) or control ( rabbit IgG Calprologics Inc . Harwick , MA ) used at 5µg/ml . Primary antibody was detected using a goat anti-rabbit alexa fluor 546 secondary antibody ( Invitrogen , Paisley , UK ) . Flow cytometric analysis was performed with a CyAn flow cytometer and Summit v4 . 3 software ( Dako ) . Macrophages were stained for MyD88 and IRF-7 after fixation and permeabilised as above , but at room temperature . Cells were blocked in 2% goat serum ( Vector laboratories ) in 0 . 5% BSA ( Fluka , UK ) in PBS for 2h at room temperature or at 4°C overnight . Polyclonal rabbit anti-IRF-7 or control polyclonal sera was used at 5µg/ml . Monoclonal rat anti-MyD88 ( R&D Systems , UK ) or rat IgG2a ( Invitrogen , Paisley , UK ) was used at 5µg/ml . Primary antibodies were incubated with cells for 1h at RT . Primary antibodies were detected using the following secondary antibodies ( 10µg/ml; 1h at RT ) ; goat anti-rabbit alexa fluor 546 ( Invitrogen , Paisley , UK ) , and goat anti-rat alexa fluor 488 ( Invitrogen , Paisley , UK ) . Coverslips were mounted in ProLong gold ( Invitrogen , Paisley , UK ) mounting medium . Immunofluorescence staining of 8µm cryosections of spleen was performed under same conditions except fixation/permeabilisation was carried out using ice cold acetone for 10 min . Sections were blocked as described above and antibodies were used as follows ( 5µg/ml ) ; rat anti-CD169 FITC conjugated ( Serotec , UK ) , rat IgG2a FITC isotype ( Invitrogen ) , rat anti-SIGNR1 biotinylated ( Bachem , UK ) , rat IgM biotinylated ( Invitrogen ) . Streptavidin alexa fluor 633 ( Invitrogen ) was used to detect biotinylated primary antibodies and IRF-7/amastigote staining was carried out as above . Specificity of IRF-7 staining was determined by absence of staining in sections taken from naïve and infected B6 . Irf-7−/− mice . Similarly , no specific staining around L . donovani amastigotes was observed following infection of 14M1 . 4 knock-down clones or following in vivo infection of B6 . Irf7−/− mice ( Supplementary Figure S6 ) , confirming that the polyclonal anti-IRF-7 antibodies did not react to Leishmania components . Sections counterstained with 1µM DAPI ( Invitrogen ) in PBS for 10 minutes and mounted . Slides were analysed on a 4 laser ( 7-line ) Zeiss Meta invert motorised microscope ( Carl Zeiss , Germany ) and images analysed in LSM software ( Carl Zeiss , Germany ) . Total numbers of parasites within the MZ ( and their distribution within macrophage subsets ) were determined from >5 sections per mouse , scoring all MZ profiles observed in each section . Data were pooled across 3–5 mice per time point and each experiment was repeated independently at least twice . Statistical analysis was performed using two-tailed student t test or Mann Whitney U test ( as appropriate ) , with p <0 . 05 considered significant .
|
Macrophages are phagocytic cells that play a dual role in infection . They can kill ingested micro-organisms and thus help eliminate the threat from infection , but some pathogens have adapted to survive within macrophages and use the intracellular niche they provide as a means of immune evasion . Although it has long been recognised that macrophages in different tissues look and behave differently , this heterogeneity is rarely taken into account when examining macrophage-pathogen interactions . By comparing gene expression profiles of different types of macrophages , we show here the diversity of the macrophage response to Leishmania donovani infection . In one cell line that resembles the macrophages of the spleen that are targets of L . donovani infection , we identified a gene expression signature more similar to that expected for a viral infection , and by using RNAi , identified that the transcription factor IRF-7 played an essential role in the selective capacity of these macrophages to kill this intracellular parasite . The importance of this pathway of host resistance was directly confirmed by examining the response of splenic macrophages in Irf-7-deficient mice . This comparative approach has , therefore , identified a new molecular pathway involved in the innate tissue-specific control of L . donovani infection .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"immunology/immunity",
"to",
"infections",
"microbiology/parasitology",
"immunology/innate",
"immunity"
] |
2010
|
Innate Killing of Leishmania donovani by Macrophages of the Splenic Marginal Zone Requires IRF-7
|
Methicillin-resistant Staphylococcus aureus ( MRSA ) is an important human pathogen , which is cross-resistant to virtually all β-lactam antibiotics . MRSA strains are defined by the presence of mecA gene . The transcription of mecA can be regulated by a sensor-inducer ( MecR1 ) and a repressor ( MecI ) , involving a unique series of proteolytic steps . The induction of mecA by MecR1 has been described as very inefficient and , as such , it is believed that optimal expression of β-lactam resistance by MRSA requires a non-functional MecR1-MecI system . However , in a recent study , no correlation was found between the presence of functional MecR1-MecI and the level of β-lactam resistance in a representative collection of epidemic MRSA strains . Here , we demonstrate that the mecA regulatory locus consists , in fact , of an unusual three-component arrangement containing , in addition to mecR1-mecI , the up to now unrecognized mecR2 gene coding for an anti-repressor . The MecR2 function is essential for the full induction of mecA expression , compensating for the inefficient induction of mecA by MecR1 and enabling optimal expression of β-lactam resistance in MRSA strains with functional mecR1-mecI regulatory genes . Our data shows that MecR2 interacts directly with MecI , destabilizing its binding to the mecA promoter , which results in the repressor inactivation by proteolytic cleavage , presumably mediated by native cytoplasmatic proteases . These observations point to a revision of the current model for the transcriptional control of mecA and open new avenues for the design of alternative therapeutic strategies for the treatment of MRSA infections . Moreover , these findings also provide important insights into the complex evolutionary pathways of antibiotic resistance and molecular mechanisms of transcriptional regulation in bacteria .
Methicillin-resistant Staphylococcus aureus ( MRSA ) is a leading cause of infections in hospitals in many countries and has also become an important community- and livestock-associated pathogen [1]–[3] . Recently , a report from CDC has reassessed the burden of MRSA infections in the USA , putting the number of deaths attributable to MRSA in front of those related to HIV-AIDS , Alzheimer disease or homicide [4] . MRSA are resistant to virtually all β-lactam antibiotics , one of the most clinically relevant class of antimicrobial agents . In addition , contemporary MRSA strains are frequently resistant to many other antimicrobial classes leaving clinicians with few therapeutic options . The MRSA characteristic phenotype is due to an extra penicillin-binding protein ( PBP2a ) coded by the mecA gene [5] , which has a remarkable reduced affinity for many β-lactams [6] . In addition , >95% of MRSA strains have also a β-lactamase enzyme coded by blaZ that confers penicillin-resistance [7]–[9] . The mecA transcription can be controlled by the divergent mecR1-mecI regulatory genes , coding for a sensor-inducer and a repressor , respectively [10] . This genetic organization of the mecA locus is similar to that of the β-lactamase , which contains the structural gene blaZ and the homologous blaR1-blaI regulatory genes . In fact , there is a cross-talk between both systems [11]–[14] , and the signal-transduction mechanisms are believed to be identical [15] , [16] , involving two proteolytic steps , in contrast with the most common bacterial signal transduction mechanism that involves the phosphorylation of regulatory proteins [17] . Specifically , the currently accepted model of mecA regulation involves two main steps: ( i ) binding of the β-lactam antibiotic to the extracellular sensor domain of MecR1 leads to the autocatalytic cleavage of the sensor-inducer and activation of the cytoplasmatic inducer domain , which appears to be a prometalloprotease; ( ii ) the activated inducer domain of MecR1 either directly cleaves the promoter-bound MecI dimers or promotes the repressor cleavage , which disables the ability of the repressor protein to dimerize and bind to the mecA promoter , enabling the expression of the resistance gene . MecR1 once cleaved can no longer transmit signal but , since the expression of mecR1-mecI is also up-regulated , the mecA induction continues as long as the antibiotic is present in the environment . Some details of the signaling mechanism involved in the transcriptional control of mecA are still elusive . For instance , induction of mecA by MecR1 has been described as extremely slow [12] , [13] , so that cells with intact mecR1-mecI regulatory system appear phenotypically susceptible in spite of the presence of mecA – the so-called “pre-MRSA” phenotype [10] , [18] . Based on these observations , it has been postulated that high-level resistance to β-lactams , characteristic of many contemporary MRSA clinical strains , implies a non-functional mecR1-mecI regulatory system . In agreement with this hypothesis , several studies have described the accumulation of point mutations and/or gene deletions in the mecR1-mecI coding sequences [19]–[23] . Still , in some of these studies [19] , [21] , [23] , based on contradictory observations , the existence of alternative mecA regulatory mechanisms has also been proposed . In fact , in a recent study , we could not establish any correlation between the mecR1-mecI integrity and the β-lactam resistance phenotype in a representative collection of epidemic MRSA strains and , unexpectedly , overexpression in trans of a wild-type copy of MecI had no effect on the phenotypic expression of resistance in most strains [24] . Here , we identify the missing link that explains the puzzling observations described above . We demonstrate that the mecA regulatory locus is in fact a three-component system that contains , besides mecR1-mecI genes , the mecR2 gene , which is co-transcribed with mecR1-mecI . In vitro and in vivo assays show that MecR2 acts as an anti-repressor by interacting directly with MecI repressor , disturbing its binding to the mecA promoter , which results in its inactivation by proteolytic cleavage . In MRSA strains with functional mecR1-mecI genes , MecR2 is essential for the full induction of mecA transcription , compensating for the inefficient induction of mecA by mecR1 and enabling the optimal expression of β-lactam resistance . These findings suggest a need to revise the current model for the induction of mecA expression in clinical MRSA strains and open new avenues for the design of alternative therapeutic strategies targeting the regulatory pathway of mecA expression . In addition , this unusual combination of repressor , sensor-inducer and anti-repressor , together with the unique modulation of a series of proteolytic cleavage steps underlying the signal transduction mechanism , provides important insights into the evolution of antibiotic-resistance and transcriptional control of genes in bacteria .
Since the MRSA phenotype is not affected by the overexpression in trans of the mecA repressor [24] , we hypothesized that a third regulator might be involved in the mecA transcriptional control . Taking into account that mecA gene is part of a large polymorphic exogenous DNA fragment ( the so called SCCmec element ) , which has integrated in the chromosome [25] , we reasoned that the putative additional regulator should be located within this chromosomal cassette , most likely upstream to the mecA gene; i . e . genetically linked to the mecR1-mecI region . Upon analysis of prototype sequences of SCCmec types II and III , which are characterized by complete mecR1-mecI coding sequences [26] , we found a highly conserved region ( 99 . 9% homology ) immediately downstream of mecI . This region contains the divergent small coding sequence for a phenol-soluble modulin , psm-mec , involved in S . aureus virulence and colony spreading [27] , [28] , and a putative open-reading frame ( ORF ) that , due to a difference in four-tandem thymine residues , has a variable length ( Figure 1A ) : 870 bp in SCCmec type II prototype strain N315 ( accession no . D86934 , positions 41794-40925 ) and 1149 bp in SCCmec type III prototype strain HU25 ( accession no . AF422694 , positions 4729-3861 ) . Both variants are homologous to the repressor of the xylose operon of S . xylosus , XylR ( accession no . X57599 ) with an amino-acid identity of 60–64%; the four-thymine deletion in strain N315 eliminates the N-terminal DNA-binding helix-turn-helix domain of XylR ( Figure 1B ) . Available genomic and SCCmec sequence data demonstrate that both ORF variants are disseminated in S . aureus and in other staphylococcal species containing SCCmec sequences ( Figure S1 ) . We coined the name of mecR2 for this putative ORF . Analysis of the upstream sequences of the putative mecR2 gene revealed no obvious promoter sequences . A putative terminator region , consisting of two perfect inverted repeats of 19 base pairs , was identified downstream the stop codon of mecR2 . No terminator sequences were found in the mecI-mecR2 intergenic region , suggesting that the putative mecR2 gene might be co-transcribed together with mecR1-mecI from the mecR1 promoter . This hypothesis was confirmed by transcriptional analysis of mecR2 by reverse-transcriptase PCR ( RT-PCR ) in prototype strains N315 and HU25 . Using internal primers for the putative coding region of mecR2 a positive signal was detected in both prototype strains . Moreover , using pairs of primers spanning the mecI-mecR2 and mecR1-mecI regions , positive signals were detected suggesting that mecR1-mecI-mecR2 genes are co-transcribed from the mecR1 promoter ( Figure S2 ) . We next evaluated the role of mecR2 on the MRSA phenotype by constructing a series of recombinant strains using two parental strains with contrasting phenotypes . The first of these , strain COL is highly resistant to methicillin , has no mecI , has a partially deleted mecR1 , expresses mecA constitutively , is negative for β-lactamase [29] , and is mecR2 negative . The second , strain N315 has a low-level methicillin-resistance phenotype , carries wild-type mecR1-mecI sequences , has an inducible expression of mecA , carries a β-lactamase plasmid [10] , [18] , and is mecR2 positive . In previous studies , we have observed a sharp decrease in resistance to oxacillin in strain COL overexpressing in trans the repressor mecI ( COL+mecI ) , whereas the great majority of other MRSA strains tested , including strain N315 , did not show alterations in the oxacillin-resistance phenotype [24]; being oxacillin a methicillin analogue that has replaced methicillin in clinical use . In this study , we have cloned the mecI-mecR2 region from strain N315 in the same plasmid vector . When strain COL was transformed with this recombinant plasmid ( COL+mecI-mecR2 ) the resistant-phenotype was completely restored and so was the constitutive expression of mecA ( Figure 2 ) . To exclude possible artifacts due to the overexpression of genes from multi-copy plasmids , we reconstructed the mecR1-mecI-mecR2 regulatory locus of prototype strain N315 in the chromosome of strain COL , using an insertion-deletion strategy with a thermosensible plasmid ( Figure S3 ) . First , we inserted the wild-type sequences of mecR1-mecI ( strain COL::RI ) , which caused a decrease of oxacillin-resistance when compared to the parental strain COL , in agreement with the poor induction of mecA by MecR1 alone ( Figure 3 ) . Compared to COL+mecI ( Figure 2 ) , the decrease of oxacillin-resistance in COL::RI was less severe , most likely due to the presence of the inducer MecR1 . Upon introduction of the complete mecA regulatory locus; i . e . mecR1-mecI-mecR2 ( strain COL::RI-R2 ) , the phenotype of parental strain COL was fully restored . As control experiments , we re-introduced the recombinant plasmid over-expressing mecI in recombinant strains COL::RI and COL::RI-R2 , originating strains COL::RI+mecI and COL::RI-R2+mecI , respectively . In both cases no effect was detected on the phenotypic expression of resistance , suggesting that the functions of mecR1 and mecR2 are not affected when mecI is overexpressed in trans , in line with our previously reported observations [24] . Altogether , these experiments suggest that mecR2 compensates for the poor induction of mecA by mecR1 , enabling the optimal expression of resistance . To further clarify the mecR2 function , the chromosomal copy of mecR2 gene from strain N315 was replaced by an antibiotic-resistance marker ( N315::ΔmecR2 ) , using a similar insertion-deletion strategy ( Figure S3 ) . Deletion of mecR2 caused a sharp decrease in the phenotypic expression of oxacillin-resistance in strain N315 . We then cloned the mecR2 gene from strain N315 under the control of an inducible promoter ( spac::mecR2 ) and , in the presence of the inducer ( IPTG 100 µM ) , we succeeded in complementing the mecR2 null-mutant in strain N315 ( Figure 4A ) and restored the COL phenotype of recombinant strain COL::RI ( Figure 4B ) . Since in these experiments only the mecR2 gene was deleted from the chromosome of strain N315 and complemented in trans , it can be concluded that the intergenic mecI-mecR2 region has no role in the phenotypic expression of β-lactam resistance . Altogether , these observations demonstrated that mecR2 interferes with the mecI-mediated repression of mecA , compensates for the inefficient MecR1-mediated induction of mecA and enables the optimal expression of β-lactam resistance in the presence of functional mecR1-mecI sequences . Of note , first attempts to complement the mecR2 null-mutant by overexpression it in trans only succeeded if mecR2 was co-overexpressed together with mecI ( data not shown ) . This requirement for low MecR2 cellular amounts and/or equimolar cellular amounts of MecI and MecR2 , suggests that at high cellular concentrations MecR2 function may be lost , either due to oligomerization or ( non-specific ) interference with essential cellular targets . A classical example of the requirement for equimolar ratios between interacting proteins is the Escherichia coli helicase DnaB/replication factor DnaC complex , in which the replication is inhibited when DnaC is in excess [30] . We next analyzed the effect of mecR2 on the induction profile of mecA transcription in parental strain N315 , its null-mutant for mecR2 , and in the complemented mutant , by Northern blotting ( Figure 5A ) and quantitative Real-time RT-PCR ( qRT-PCR ) analysis ( Figure 5B ) . In relative terms , upon induction with sub-MIC oxacillin , a much stronger induction of mecA transcription was observed in the parental strain than in the mecR2 null-mutant ( N315::ΔmecR2 ) , in which the amount of mecA transcript seems to be not sustained during the last two time-points . In the complemented mutant ( N315::ΔmecR2+spac::mecR2 ) there was a sustained induction of mecA transcription throughout the time-course of the experiment . However , in the complemented strain , although the resistant phenotype of the parental strain was fully restored ( as illustrated in Figure 4A ) , the amount of mecA transcript was substantially lower and virtually identical to the mecR2 null-mutant . Although this discordance is in agreement with previous studies reporting on the lack of a correlation between the cellular amounts of mecA transcript or protein and the phenotypic level of resistance [31] , [32] , we cannot formally exclude other possible explanations , such as MecR2 having multiple targets that affect the resistance phenotype . Nevertheless , these data suggest that the mecR2 interferes with the induction of mecA transcription in response to β-lactams . We have also analyzed by qRT-PCR the induction profile of mecR2 in parental strain N315 and in the complemented mecR2 null-mutant ( N315::ΔmecR2+spac::mecR2 ) – Figure 6 . qRT-PCR data for parental strain N315 showed that mecR2 transcription was induced in the presence of sub-MIC oxacillin , in agreement with data from RT-PCR that showed that mecR2 was co-transcribed with mecR1-mecI from the inducible mecR1 promoter ( Figure S2 ) . In the complemented mutant , in the presence of the inducer ( IPTG ) , mecR2 transcription levels were 10 fold higher than those of parental strain and , as such , the low levels of mecA transcription observed for this strain ( Figure 5 ) cannot be attributable to an inefficient induction of mecR2 transcription from the Pspac promoter . Of note , in parental strain N315 , mecR2 transcription levels appear to be residual when compared to those of mecA ( 109 fold less ) . Although this might be explained by an experimental artifact , one can also speculate that this may be due to different promoter strengths and/or to promoter blockage by RNA polymerase , since mecA and mecR1 promoters are divergent and overlap partially . In terms of signal-transduction mechanism , once the expression of mecA is induced in response to β-lactams , there is no need for high cellular levels of inducer , repressor or anti-repressor . Actually , the basal transcription of mecR1-mecI-mecR2 is only necessary to assure that the repressor protein is still present when the antibiotic induction stops , so that the transcription of the resistance gene is shutdown . The apparent very low transcription level of mecR2 in parental strain N315 may also explain the lack of complementation when mecR2 was over-expressed in trans . In fact , this artificial system , when compared to wild-type strains , presumably generates extremely high cellular amounts of MecR2 , which may originate a loss of function by oligomerization or non-specific interactions with other cellular targets . Finally , the apparent residual mecR2 transcription levels may also explain our failed attempts to analyze by Northern blotting the transcription of mecR2 in prototype MRSA strains , even with large amounts of total RNA ( 10–30 µg ) and long autoradiograph expositions ( 72 h ) . To our knowledge , Northern blotting analysis of mecR1-mecI/blaR1-blaI transcripts was described in only two studies and , in both cases , clear signals were obtained only when regulatory genes were overexpressed from recombinant plasmids [33] , [34] . Among the five major SCCmec types , only SCCmec types II and III are characterized by complete mecR1-mecI regulatory locus [26] – Figure 1A . SCCmec type III strains appear to have a conserved point mutation within mecI coding sequence resulting in a truncated non-functional repressor protein [24] , [35] . Concerning SCCmec type II strains , the accumulation of deleterious mutations has also been described in some strains [19]–[23] . However , data from our MRSA collections [24] , as well as from available genomic and SCCmec type II sequences , suggest that many strains have wild-type sequences for mecI ( and mecR1 ) . For instance , in a BLAST analysis against the mecI sequence from strain N315 , c . a . 20 entries were found with 100% sequence identity , mostly from S . aureus strains but also from a few coagulase-negative staphylococci ( S . epidermidis , S . saprophyticus , S . fleurettii , S . cohinii , etc . ) . These observations suggest that the mecR2 function may be required for the optimal expression of β-lactam resistance in those SCCmec type II strains with wild-type sequences for mecI and mecR1 . In order to explore that hypothesis , we sought to test the role of mecR2 in the phenotypic expression of β-lactam resistance in prototype strains of epidemic MRSA clones characterized by SCCmec type II . The MRSA population has a very strong clonal structure and only a few epidemic clones are responsible for the majority of infections worldwide [36] , [37] . Three epidemic MRSA clones characterized by SCCmec type II have been described [38]: clone ST5-II , “New York/Japan” or USA100; clone ST36-II , EMRSA-16 or USA200; and clone ST45-II or USA600 . MRSA clones ST5-II and ST36-II are two of the most important nosocomial clones in the USA and UK , respectively . Prompted by this epidemiological data , we evaluated the role of mecR2 in three representative strains of those SCCmec type II clones selected from a large US collection of MRSA [39]: strains USA100 , USA200 and USA600 . For this purpose , the chromosomal mecR2 deletion of strain N315 ( N315::ΔmecR2 ) was transduced into those strains originating the recombinant strains: USA100::ΔmecR2 , USA200::ΔmecR2 and USA600::ΔmecR2 . In the three prototype strains , deletion of mecR2 caused a sharp decrease of the phenotypic expression of β-lactam resistance , which could be complemented by expressing mecR2 in trans under the control of an inducible promoter ( spac::mecR2 ) ( Figure 7A ) . We have also analyzed the effect of mecR2 on the induction of mecA transcription in strains USA100 , USA200 and USA600 by Northern blotting . Compared to N315 , the three parental strains expressed mecA at higher levels and , in agreement with what was observed for strain N315 ( Figure 5 ) , deletion of mecR2 caused a sharp decrease on the mecA induction and transcription levels ( Figure 7B ) . Since strains N315 , USA100 , USA200 , and USA60 have complete mecR1 genes and strain COL has a truncated mecR1 gene but with a complete N-terminal inducer domain , with the previous experiments we could not formally exclude that MecR2 function is dependent of at least the N-terminal inducer domain of MecR1 . Therefore , we sought to test the effect of mecR2 in a prototype SCCmec type V MRSA strain , characterized by an extensive deletion of mecR1 spanning both N- and C-terminal domains [40] – Figure 1A . Among our collections , we selected strain HT0350 [41] , since it was the only strain also negative for the β-lactamase locus [24] . Similar to what was observed for strain COL , overexpression of MecI in strain HT0350 ( HT0350+mecI ) caused a sharp decrease of resistance level , which was fully reverted with the co-overexpression of MecI and MecR2 ( HT0350+mecI-mecR2 ) – Figure 8A . These data suggests that the effect of mecR2 on the expression of β-lactam resistance in S . aureus is not dependent of mecR1 , and as such MecR2 may act as an anti-repressor . Since mecA transcription can be co-regulated by the regulators of the β-lactamase ( bla ) locus , blaR1-blaI , and parental strains N315 , USA100 , USA200 and USA600 are bla positive , we sought to evaluate the effect of bla genes on the observed mecR2-induced phenotypes . For this purpose , we took advantage of the fact that the experimental strategy used to construct the mecR2 knockout in prototype strain N315 generated an intermediate mutant strain which lost the β-lactamase plasmid , probably due to the multiple passages , many of which at 45°C . As in all other chromosomal manipulations , the mecR2 genetic deletion was transduced back to the parental β-lactamase positive strain N315 to generate the final deletion mutant ( N315::ΔmecR2 ) tested in all previous experiments . As illustrated in Figure 8B , in both variants of the mecR2 chromosomal deletion , there was a sharp decrease of the β-lactam resistance . Together with the experimental data for strains COL and HT0350 , both bla negative , this assay indicated that the mecR2 function on the phenotypic expression of β-lactam resistance is not dependent on the presence of the β-lactamase plasmid . In addition , in order to exclude an interaction of MecR2 with bla regulators , we sought to evaluate the phenotype of a mecR2 deletion mutant in prototype strain HU25 , a highly resistant MRSA strain which is positive for the bla locus and has a truncated non-functional MecI protein due to a premature stop codon [24] . Previous studies have shown that in the presence of oxacillin , the transcription of mecA is readily induced in strain HU25 , presumably by the bla system [24] . As illustrated in Figure 8C , the absence of mecR2 in strain HU25 ( strain HU25::ΔmecR2 ) had no effect on the phenotypic expression of oxacillin resistance , suggesting that MecR2 is not required for the mecA induction mediated by the BlaR1-BlaI system . The MecR2 is homologous to the transcriptional repressor of the xylose operon , XylR [42] , which has a N-terminal DNA-binding domain and a C-terminal dimerization domain . The mecR2 gene in the prototype strain N315 has no DNA binding domain due to a deletion of four tandem-thymine residues , which , together with the genetic experiments done with this variant ( Figures 4 , 7 and 8 ) , suggests that only the dimerization domain is involved in the MRSA phenotype . Therefore , we reasoned that the mode of action of MecR2 might involve a direct interaction with the MecI dimer , through its dimerization domain , which eventually would interfere with its binding to the mecA promoter . To test this hypothesis , we evaluated the MecR2-MecI interaction using a bacterial two-hybrid in vivo strategy [43] . In these experiments , we used the small mecR2 variant present in prototype strain N315 . As in-house controls , the MecI::MecI interaction , previously demonstrated using the yeast two-hybrid strategy [44] , as well as the MecR2::MecR2 interaction were also evaluated . Positive results were observed in 4 out of the 8 MecI::MecR2 combinations and in 1 of 4 MecI::MecI combinations ( Figure 9A ) . No MecR2::MecR2 interaction was detected in the four combinations tested ( data not shown ) and , as such , the assay was not conclusive in this case . Altogether , these observations provide evidence for a MecR2::MecI direct interaction . Next , we evaluated the interference of purified MecR2 protein with the binding of MecI to the promoter of mecA ( PmecA ) at several molar ratios by the electrophoretic mobility shift assay ( EMSA ) , a strategy previously used to study the binding of purified MecI protein to PmecA [14] , [24] . In these experiments , we expressed in E . coli the full MecR2 protein from prototype strain HU25 , since the shorter variant of strain N315 could not be expressed and purified in a soluble form at high concentrations . As illustrated in Figure 9B , MecR2 interferes with the binding of MecI to PmecA in a concentration-dependent manner: the heavier band presumably reflecting the binding of MecI dimers to PmecA decreases in intensity , whereas the intermediate band reflecting the binding of MecI monomers to PmecA and the lighter free DNA band increase in intensity . In line with the genetic experiments , this effect was optimal for a MecR2::MecI molecular ratio below one; in the presence of excess MecR2 the binding of MecI to PmecA was restored . This in vitro loss of effect at higher concentrations of MecR2 suggests that under these conditions MecR2 may be trapped in a non-active conformation; e . g . MecR2 may oligomerize in a concentration-dependent manner and stop interacting with MecI . It should be noted that in wild-type strains , mecI and mecR2 are co-transcribed from the mecR1 promoter and , as such , the cellular amounts of both proteins are likely to be similar . Since in these experiments we used the full MecR2 variant containing a putative N-terminal DNA binding , control EMSA experiments with MecR2 alone were performed to verify that purified MecR2 did not bind to PmecA alone ( Figure S4A ) . In addition , control experiments with mixtures of MecI and MBP ( maltose-binding protein , which has an identical molecular weight to MecR2 ) were performed to demonstrate that inhibition of MecI binding to PmecA is specific for MecR2 ( Figure S4B ) . Finally , in order to exclude the hypothesis that at higher concentrations MecR2 binds ( not specifically ) to secondary sites in PmecA DNA in a MecI-dependent manner , EMSA assays with MecI-MecR2 mixtures were also performed with a much smaller DNA fragment ( 39 bp instead of 212 bp ) containing the MecI protected sequences and the same results were obtained ( data not shown ) . Altogether , these assays demonstrate that MecR2 acts as an anti-repressor disturbing the binding of MecI to the mecA promoter . Based on structural data , it has been suggested that the proteolysis of MecI observed during mecA induction is not mediated by the activated MecR1 inducer domain and instead , is a secondary event not required for induction [45]–[47] . This is in agreement with what has been found for the β-lactamase system of Bacillus licheniformis [48] . We speculated that MecR2 , by interacting with MecI and disturbing its binding to the mecA promoter , could foster a local melting of MecI-dimers , making the scissile bonds more accessible to proteolytic inactivation . To test this hypothesis , we compared by Western blotting the MecI proteolysis in total protein extracts from prototype strain N315 versus its mecR2 null-mutant under induction conditions , and from strain HT0350 overexpressing MecI-MecR2 versus HT0350 overexpressing MecI only . As illustrated in Figure 9C , in the absence of MecR2 , intact MecI accumulates in both pairs of strains analyzed . Because parental strain HT0350 is negative for all mecA regulators and its derivatives used in these experiments overexpress constitutively MecI-MecR2 or MecI , the observed MecR2-induced proteolysis of MecI does not involve MecR1 ( neither BlaR1 ) and , most likely , is mediated by unspecific cytoplasmatic proteases . The findings described in this report clarify some critical aspects of the unique signal transduction mechanism underlying the induction of mecA gene . First , we demonstrated that the cognate mecA regulatory locus contains , besides MecR1-MecI , the anti-repressor MecR2 . MecR2 compensates for the inefficient MecR1-mediated induction of mecA , being essential for the optimal expression of β-lactam resistance ( Figures 3 , 4 and 7A ) , and enabling the full induction of mecA transcription ( Figures 5 and 7B ) . These findings explain the puzzling observation of the poor mecA induction by MecR1 , reported in studies analysing the effects of mecR1-mecI only ( without mecR2 ) on mecA expression in recombinant strains [12] , [13]; an experimental artefact also observed in this study with recombinant strain COL::RI ( artificially made positive for mecR1-mecI only ) and in the mecR2 null mutant strains ( Figures 4 and 7 ) . Because wild-type MRSA strains positive for mecR1-mecI are also positive for mecR2 , these strains are in fact able to express optimal levels of β-lactam resistance and , as such , mecA is efficiently induced upon exposure to β-lactams by its cognate three-component regulatory system . Second , the findings herein described also clarify the relevance and specificity of MecI proteolysis observed upon induction with β-lactams [16] , [45]–[48] . Our data demonstrates that MecI proteolysis is required for optimal expression of resistance and that MecR2 alone ( i . e . without MecR1 , Figure 8A ) interferes specifically with the MecI function and promotes its inactivation by proteolytic cleavage , presumably mediated by ( non-specific ) native cytoplasmatic proteases ( Figure 9 ) . Our findings contrast with published observations for the blaR1-blaI system of Bacillus licheniformis , demonstrating that the proteolysis of the repressor is a secondary event not required for induction of resistance [48] , and also for the blaR1-blaI system of S . aureus demonstrating that BlaR1 directly promotes the BlaI cleavage [15] . Altogether , these observations suggest the existence of subtle differences between the mecR1-mecI-mecR2 and the blaR1-blaI regulatory systems and that , in S . aureus , BlaR1 may accumulate the MecR1 and MecR2 functions . Our findings lead us to propose a revised model for the induction of mecA expression in wild-type MRSA strains ( Figure 10 ) : ( i ) in the presence of β-lactams , MecR1 is activated and induces the transcription of mecA and mecR1-mecI-mecR2; ( ii ) the anti-repressor MecR2 , destabilizes MecI-dimers , disturbing their binding to the mecA promoter and fostering their proteolytic inactivation , resulting in a sustained induction of mecA transcription; ( iii ) when depletion of β-lactam occurs , MecR1 is no longer activated and a steady state is established consisting of stable MecI-dimers bound to the mecA promoter ( and protected from proteolysis ) and residual copies of MecR1 at the cell membrane; the remaining free MecR2 molecules are most likely degraded by the cellular protein turnover pathway . This model implies that in the absence of MecR2 , functional MecI-dimers are more resilient to proteolytic inactivation and outcompete the MecR1 signalling , a hypothesis supported by the Western blotting experiments ( Figure 9C ) . This study demonstrates that the central element of methicillin-resistance in S . aureus , the mecA gene , can be regulated by a three-component system consisting of a transcriptional repressor , a sensor-inducer and an anti-repressor , a very unusual arrangement for the transcriptional control of genes in bacteria . In addition , the induction of the resistance gene expression involves a unique series of proteolytic steps , being the proteolytic cleavage of the repressor modulated by the anti-repressor . This study also sheds light on the evolution of antibiotic-resistance genes . The mecA gene itself is probably ancient and predates the use of antibiotics in clinical practice [49] , [50] . Before its recent acquisition by MRSA , mecA was assembled into a gene complex containing its transcriptional regulators and incorporated into a mobile genetic element . Tsubakishita et al . have proposed that the mecA gene complex found in MRSA has been assembled in the animal-related Staphylococcus fleurettii species [51] . Remarkably , in this species the mecA-mecR1-mecI locus was found immediately upstream to the complete and functional xylose operon , containing the XylR repressor homologous to MecR2 . This suggests that a specific selection acted on XylR , a transcriptional repressor of sugar metabolism , to originate the MecR2 function , an anti-repressor of an antibiotic-resistance gene , and that the three-component mecA regulatory locus was assembled in S . fleuretti before being transferred to S . aureus . In short , this study points to a revision of the model for the transcriptional control of mecA by its cognate regulatory locus , which may pave the way for the design of alternative therapeutic strategies targeting the induction mechanism of the resistance gene [52] , [53] . If successful , these strategies may extend the clinical utility of β-lactams for the treatment of MRSA infections . Recycling β-lactams is particularly relevant given that MRSA pose a substantial burden for the public health , are often multi-drug resistant and , in the past 40 years , very few new classes of antibiotics have reached the clinic .
The bacterial strains and plasmids used in this study are listed in Tables S1 and S2 , respectively . S . aureus strains were routinely grown at 37°C with aeration in tryptic soy broth ( TSB , Difco ) or on tryptic soy agar plates ( TSA , Difco ) . E . coli strains were grown with aeration at 37°C in Luria-Bertani broth ( LB , Difco ) or in Luria-Bertani agar ( LA , Difco ) . Recombinant E . coli strains were selected and maintained with ampicillin at 100 µg/mL . Recombinant S . aureus strains were selected and maintained either with tetracycline at 5 or 40 µg/mL , chloramphenicol at 20 µg/mL , or erythromycin at 10 µg/mL , as appropriate . Phenotypic analysis of β-lactam resistance in S . aureus parental and recombinant strains was performed by diffusion-disks containing 1 mg of oxacillin , Oxacillin E-test ( AB Biodisk ) , or by population analysis profiles ( PAPS ) at 30°C for 24–48 h , as previously described [24] , [54] , [55] . Oxacillin is a methicillin analogue and has replaced methicillin in clinical use . DNA manipulations were performed by standard methods [56] , [57] . Total DNA from S . aureus was isolated from bacterial cultures with the Wizard Genomic DNA purification Kit ( Promega ) according to the manufacturer's recommendations and using lysostaphin ( 0 . 5 mg/mL ) and RNAse ( 0 . 3 mg/mL ) in the lysis step . Plasmid DNA was isolated from bacterial cultures with the High Pure Plasmid Isolation Kit ( Roche ) . For plasmid DNA isolation from S . aureus strains the culture pellets were ressuspended in “Suspension Buffer” supplemented with 0 . 1 mg/mL of lysostaphin and incubated at 37°C for 30–60 minutes . Restriction enzymes were used as recommended by the manufacturer ( New England Biolabs ) . Dephosphorilation of vector arms and insert ligation was performed with Rapid DNA Dephos & Ligation kit ( Roche ) , according to the manufacturer's recommendations . Routine PCR was performed with GoTaq Flexi DNA polymerase ( Promega ) . PCR amplification of cloning inserts was performed by high-fidelity PCR ( Pfu Turbo DNA polymerase , Strategene ) . DNA purification from PCR and digestion reactions was performed with High Pure PCR Product Purification Kit ( Roche ) . For ligation protocols , the inserts and linearized plasmids were resolved in a low melting agarose gel ( 1% ) ( Invitrogen ) and DNA bands were purified with Gene Clean Turbo kit ( MP Biomedicals ) , following the manufacturer's recommendations . DNA sequencing was performed by Macrogen ( www . macrogen . com ) or STAB Vida ( www . stabvida . com ) . All primers used in this study are listed in Table S3 . All recombinant plasmids used in this work were firstly constructed and stabilized in E . coli DH5α , electroporated into S . aureus restriction-deficient strain RN4220 and finally transduced by the 80α phage to the desired parental strain , as previously described [58] , [59] . The Integrity of plasmid inserts was confirmed by restriction analysis , PCR and DNA sequencing . The integrity of chromosomal insertion-deletions was confirmed by PCR , DNA sequencing and Southern blotting of pulsed-field gel electrophoresis of chromosomal DNA . Chromosomal insertion-deletions were backcrossed by phage transduction to the original parental strains . To co-overexpress mecI and mecR2 in strain COL , a fragment containing the mecI-mecR2 region from strain N315 was amplified using primers MI-P1/MR2-P1 , double-digested with EcoR1/BamH1 and cloned into pGC2 , originating the recombinant plasmid pGC::mecI-mecR2 . To reconstruct the mecA regulatory locus in the chromosome of strain COL , we first construct pSPT::IS-erm , a pSPT181 derivative containing the terminal fragment from IS1272 located in the upstream vicinity of mecA in strain COL and the erythromycin ( erm ) resistance cassette gene from Tn551 . The 0 . 6 kb terminal fragment of IS1272 was amplified from strain COL using primers IS1272-P1/IS1272-P2 , double-digested with PstI/SalI and cloned into pSPT181 , originating pSPT::IS . The erm cassette was recovered from the pSP64E plasmid by BamHI/SalI double-digestion and was cloned into pSPT::IS , originating pSPT::IS::erm . To reconstruct the mecR1-mecI locus in strain COL ( strain COL::RI ) , we amplified by high-fidelity PCR a 1 . 9 kb DNA fragment from strain N315 chromossomal DNA , containing the wild-type coding sequences of mecR1 and mecI genes , using primers MI-P2/MR1-P1 . The fragment was double-digested with BamH1/AvaI and directionally cloned into pSPT::IS-erm , originating pSPT::IS-erm-mecI-mecR1 . To reconstruct the full mecA regulatory locus in strain COL ( strain COL::RI-R2 ) , we amplified by high-fidelity PCR a 3 . 5 kb DNA fragment from strain N315 chromossomal DNA , containing the mecR1-mecI-mecR2 locus , using primers MR2-P1/MR1-P1 . The fragment was double-digested with BamH1/AvaI and directionally cloned into pSPT::IS-erm , originating pSPT::IS-erm-mecR2-mecI-mecR1 . As control , we constructed a strain with a integrated erm gene in the mecA upstream vicinity ( strain COL::erm ) : a 0 . 5 kb DNA fragment containing the terminal fragment of the N-terminal cytoplasmatic domain of mecR1 , was amplified using primers MR1-P1/MR1-P2 , double-digested with BamH1/AvaI and cloned in pSPT::IS-erm , originating pSPT::IS-erm-ΔmecR1 . The integration into COL chromosome of the three recombinant plasmids ( pSPT::IS-erm-mecI-mecR1 , pSPT::IS-erm-mecR2-mecI-mecR1 and pST::IS-erm-ΔmecR1 ) was performed by an insertion-deletion strategy by homologous recombination ( Figure S3 ) . First , insertion into the chromosome was promoted by growing transductants in TSB at a non-permissive temperature ( 45°C ) without antibiotic selection for 2–3 days , with daily re-inoculum in fresh medium . Serial dilutions were plated onto TSA plates supplemented with erythromycin ( Ery ) and tetracycline ( Tc ) . Single-colonies Ermr-Tcr were screened for chromosomal insertion of the plasmids by PCR and the absence of plasmid DNA was confirmed . Resolution of integrates by homologous-recombination was promoted by growing selected single colonies in TSB supplemented with tetracycline at 40 µg/mL at the permissive temperature of 30°C for 4–5 days , with daily re-inoculum in fresh medium . Finally segregation of the excised plasmids was promoted by growing cultures at 45°C without antibiotic selection for 2–3 days , with daily re-inoculum in fresh medium . Culture aliquots were plated onto TSA plates supplemented with erythromycin and single colonies Ermr-Tcs were selected by replica plating onto TSA plates supplemented with erythromycin or tetracycline . To construct the mecR2 gene null mutant in strain N315 , two DNA fragments of 1000 bp corresponding to the 5′ and 3′ vicinities of the mecR2 gene were amplified by PCR from strain N315 DNA using primers MR2-P2/MR2-P3 and MR2-P4/MR2-P5 , respectively . The cat gene coding for chloramphenicol resistance was also amplified by PCR from pGC2 plasmid with primers CAT-P1/CAT-P2 . The three fragments were double-digested with SalI/PstI , BamH1/XhoI and XhoI/SalI , respectively , and then sequentially cloned into pSPT181 , originating the pSPT::cat-ΔmecR2 recombinant plasmid . Following the same insertion-deletion strategy described above , but selecting for chloramphenicol resistance instead of erythromycin resistance , we obtained the recombinant strain N315::ΔmecR2 in which the chromosomal copy of mecR2 was replaced by the cat gene ( Figure S3 ) . To complement the N315ΔmecR2 null-mutant three recombinant plasmids were constructed: ( i ) pSPT::mecR2 , pSPT181 vector containing at the XmaI site the mecR2 gene from strain N315 , obtained by PCR with primers MR2-P6/MR2-P7 ( the proper insert orientation was selected by restriction analysis using the HindIII site within mecR2 gene ) ; ( ii ) pSPT::mecImecR2 , pSPT181 containing the mecI-mecR2 genes of strain N315 , constructed by sequential cloning , first , at the BamH1/PstI , the mecI gene site obtained with primers MI-P3/MI-P4 and then , at the XmaI site , the mecR2 gene obtained with primers MR2-P6/MR2-P7; ( iii ) pSPT::spac-mecR2 , pSPT181 with the mecR2 gene under the control of Pspac promoter , constructed by sequential cloning the 1 . 6 kb EcoR1-BamH1 fragment from plasmid pDH88 containing the spac locus ( Pspac-polylinker-lacI repressor ) and then , at the XmaI site of the spac polylinker , the mecR2 gene from strain N315 obtained with primers MR2-P6/MR2-P7 . As control , N315ΔmecR2 was transformed with a pST181 derivative containing the spac locus only ( pSPT::spac ) . To generate the mecR2 gene null mutant in prototype strains USA100 , USA200 , USA600 , and HU25 the chromosomal deletion of strain N315::ΔmecR2 was transduced by bacteriophage infection with selection for chloramphenicol resistance , originating recombinant strains USA100::ΔmecR2 , USA200::ΔmecR2 , USA600::ΔmecR2 and HU25::ΔmecR2 . Mutant strains USA100::ΔmecR2 , USA200::ΔmecR2 and USA600::ΔmecR2 were then complemented with recombinant plasmid pSPT::spac-mecR2 . Total RNA extraction and purification was performed as previously described [60] . Briefly overnight cultures were grown in TSB , supplemented with antibiotics when appropriate , and then diluted 1∶50 in fresh TSB and grown to the mid-log phase ( OD620∼0 . 7 ) . Cultures were stabilized with two volumes of RNAprotect Bacteria Reagent ( Qiagen ) , according to the manufacturer's recommendations . The cells were centrifuged and pellets were ressuspended in 1 mL of Trizol reagent ( Invitrogen ) . The ressuspended cells were transferred to a new tube with silica beads ( Lysing Matrix B tubes , Bio101 ) and cell lysis was performed in the FastPrep FP120 apparatus ( Bio 101 ) . RNA was extracted with chloroform , precipitated with isopropanol , washed twice with ethanol at 80% and ressuspended in diethyl pyrocarbonate ( DEPC ) -treated water . For the analysis of the mecA and mecR2 induction profiles , after cultures were grown to OD620∼0 . 7 , oxacillin at 0 . 05 µg/mL was added and cultures were incubated for an additional 60 minutes . Samples were taken either at 0 , 5 , 15 , 30 , and 60 or at 0 , 10 and 60 minutes , stabilized , pelleted and kept on ice until being simultaneously processed . For RT-PCR and qReal-time RT-PCR experiments ( see below ) , total RNA preps were treated twice with DNAse ( RNase-Free DNase Set I , Qiagen ) and purified with RNeasy Mini Kit ( Qiagen ) , according to the manufacture's recommendations . Control PCR reactions were performed to test the absence of DNA contamination in total RNA preps . Transcription analysis of mecR1-mecI-mecR2 was performed by RT-PCR for mid-log phase induced cultures ( oxacillin at 0 . 05 µg/mL ) of strains N315 and HU25 with primer pairs MR2-RT1/MR2-RT2 ( mecR2 transcript ) , MI-P5/MR2-P8 ( mecI-mecR2 co-transcript ) , MR1-P3/MI-P6 ( mecR1-mecI co-transcript ) , and MA-P1/MA-P2 ( mecA transcript , inducible positive control ) . RT-PCR reactions were set-up using the One-Step RT-PCR kit ( Qiagen ) , according to the manufacture's recommendations . To control the absence of DNA contamination , all samples were tested in a parallel reaction without the reverse-transcription step . To control the size of the amplified transcripts , PCR reactions with chromosomal DNA were also performed in parallel . The mecA transcript was detected in both induced and non-induced samples suggesting that the RT-PCR assay was too sensitive to discriminate between basal and induced transcription levels . The induction profiles of mecA and mecR2 were determined by quantitative Real-time RT-PCR ( qRT-PCR ) and/or Northern blotting . For the qRT-PCR data analysis , relative gene expression was expressed as a ratio to the transcript of pta , a housekeeping gene with constitutive expression [61] . Standard curves were generated using serial dilutions ( 0 . 4–40 ng/reaction ) of genomic DNA and primers MR2-RT3/MR2-RT4 , MA-RT1/MA-RT2 and pta-RT1/pta-RT2 for amplification internal fragments of mecR2 , mecA , and pta , respectively . qRT-PCR reactions were performed with QuantiTect SYBR Green RT-PCR Kit ( Qiagen ) ; each 25 µl reaction containing 12 . 5 µl SybrGreen mix , 0 . 25 µl RT enzyme mix , 12 . 5 pmol of each primer and 40 ng of purified RNA . Amplification consisted of an initial RT step at 50°C for 30 min , followed by a denaturation step at 95°C for 15 min , then by 45 cycles of 30 s at 94°C , 30 s at 53°C and 30 s at 72°C . For each RNA sample three independent qReal-Time RT-PCR experiments were carried out . Fluorescence was measured at the end of the annealing-extension phase of each cycle . A threshold value for the fluorescence of all samples was set manually . The reaction cycle at which the PCR product exceeds this fluorescence threshold was identified as the threshold cycle ( CT ) . The CT was then converted to relative quantity of mRNA by using a standard curve . To verify the specificity of the PCR amplification products , melting curve analysis was performed between 60–95°C . For Northern blot analysis , total RNA ( 5 µg ) was resolved through a 1 . 2% agarose-0 . 66 M formaldehyde gel in MOPS ( morpholine propanesulfonic acid ) running buffer ( Sigma ) . Blotting of RNA onto Hybond N+ membranes ( Amersham ) was performed with Turboblotter alkaline transfer systems ( Schleicher & Schuell ) . For detection of mecA specific transcripts , a DNA probe was constructed by PCR amplification with primers MA-P1 and MA-P2 . After purification the probe was labeled with a Ready To Go labeling kit ( Amersham ) by using [a-32P]dCTP ( Amersham ) and was hybridized under high-stringency conditions . The blots were subsequently washed and autoradiographed . This strategy is based on the restoration of the adenylate cyclase ( CyaA ) activity by heterodimerization of protein fusions containing the T25 and T18 fragments , which form the catalytic domain of CyaA . CyaA is involved on cAMP synthesis , which binds to CAP forming the cAMP/CAP complex that activates a specific reporter gene , lacZ [43] . All strains and plasmids used in the bacterial two-hybrid studies are described in Table S4 . Both genes , mecI and mecR2 , were amplified from the chromosomal DNA of strain N315 by high-fidelity PCR , using primers MI-BTH1/MI-BTH2 for mecI and MR2-BTH1/MR2-BTH2 for mecR2 . PCR products were double-digested with KpnI/XbaI and fused to T25 or T18 fragments either at the N′ or C′ terminals , using plasmids pUT18 , pUT18c , pKNT25 and pKT25 , originating the following fusion proteins: T18-MecI , MecI-T18 , T25-MecI , MecI-T25 , T18-MecR2 , MecR2-T18 , T25-MecR2 and MecR2-T25 . The eight MecI::MecR2 recombinant plasmid combinations were co-transformed into the reporter strain Escherichia coli BTH101 and grown on Luria-Bertani ( LB ) and LA agar supplemented with 8 µg/mL 5-bromo-4-choro-3-indolyl-β-D-galactopyranoside ( X-gal ) , 50 µg/mL kanamycin , 100 µg/mL ampicillin , 100 µg/mL streptomycin , 500 µM ( IPTG ) and 2% glucose . As a positive control , plasmids p25Zip and p18Zip , containing two leucine zipper domains , were also co-transformed into E . coli BTH101 strain . Additionally , as in-house controls , the previously reported MecI::MecI interaction based on the yeast two-hybrid strategy [44] was evaluated , as well the MecR2-MecR2 interaction , by co-transforming the four combinations of mecI-containing plasmids and the four combinations of mecR2-containing plasmids , respectively . To overexpress and purify MecR2 protein , a mecR2 gene insert was obtained from the chromosome of strain HU25 by high-fidelity PCR amplification with primers MR2-cri1/MR2-cri2 and double-digestion with NcoI/XhoI . The mecR2 insert was then cloned in frame into the expression vector pCri8a , generating the recombinant plasmid pCri8a::mecR2 , expressing mecR2 with a N′ terminal His6 tag . pCri8a::mecR2 was stabilized in E . coli DH5α and then transformed to E . coli Bl21 ( DE3 ) . MecR2 protein overexpression was carried out in LB medium supplemented with 50 mg/mL kanamycin , at 18°C , and induced at an OD600∼0 . 5 with 1 µM Isopropyl β-D-1-thiogalactopyranoside ( IPTG ) for 5 h . MecI protein was overexpressed from recombinant strain DH5α+pPROEX::mecI [62] , using LB medium supplemented with 100 µg/mL ampicillin , at 37°C and induced at an OD600∼0 . 5 with 1 µM IPTG for 3 h . Protein extracts were purified as previously described [62] . The purity of the proteins was assessed by 10% tricine SDS-PAGE analysis and mass-spectroscopy . The concentrations of purified MecR2 and MecI were estimated using the Protein Assay Kit II ( BioRad ) , as recommended by the manufacturer . For the electrophoretic mobility shift assay we used the chemiluminescent-based DIG Gel Shift Kit ( Roche ) , following the manufacturer's recommendations . As DNA target we used a 212 bp fragment encompassing the mecA promoter and operator sequences from prototype strain N315 obtained by PCR amplification with primers MA-PF1/MA-PR1 . The binding of each purified protein to the mecA promoter ( PmecA ) was first evaluated and then MecI-MecR2 mixtures were tested . As control the binding to PmecA of MecI-MBP ( Maltose-binding protein , MBP2* , New England Biolabs ) mixtures were also evaluated . EMSA assays with MecI-MecR2 mixtures were also performed with a smaller 39 bp DNA fragment containing the MecI protected sequences , obtained by annealing primers MI-Box1/MI-Box2 . To prepare protein extracts of S . aureus , parental and recombinant strains were grown in TSB supplemented with oxacillin at sub-MIC concentration ( 0 . 05 µg/mL ) until mid-log phase ( OD620∼0 . 7 ) . Cell pellets were frozen in liquid nitrogen , thawed and resuspended in Buffer A ( 50 mM Tris-HCl; 10 mM MgCl2; 0 . 5 mM PMSF ) containing 10 µg/mL DNase I . Cells were broken mechanically in a French press followed by centrifugation ( 22 , 000× g , 20 min , 4°C ) to remove unbroken cells and cell debris . The supernatants containing the cytoplasmic proteins were recovered and filtered through 0 . 45-µm-pore-size membrane filters . Protein extracts ( 60–80 µg ) were resolved in a 18% Tris-Glycine SDS-PAGE . After electrophoresis , the proteins were transferred to a 0 , 45 µm nitrocellulose membrane ( Trans-Blot , Bio-Rad ) . The membranes were blocked at room temperature for 1 hour , in 20 mL of Blocking solution - Tween- Phosphate Buffered Saline ( 137 mM NaCl; 2 . 7 mM KCl; 4 . 3 mM Na2HPO4; 1 . 47 mM KH2PO4; 0 . 05% Tween-20 ) with 6% low-fat milk . MecI protein was detected by imunoblot analysis using a custom polyclonal antibody ( Eurogentec ) raised against purified MecI ( 1/1 . 000 dilution ) and a 1/50 . 000 dilution of secondary antibody ( Goat Anti-Rabbit IgG ( H+L ) Peroxidase Conjugated Antibody , Pierce ) in 10% Blocking solution . The immune complexes were detected using an enhanced chemiluminescence system ( SuperSignal West Pico Chemiluminescent; Pierce ) , according to the manufacturer's instruction . Membranes were exposed to Amersham Hyperfilm ECL film ( GE Healthcare ) .
|
Methicillin-resistance Staphylococcus aureus ( MRSA ) is an important human pathogen , causing a wide range of infections . MRSA strains are resistant to virtually all β-lactam antibiotics and often are also resistant to many other classes of antibiotics , leaving physicians with few therapeutic options . MRSA is defined by the presence of the mecA gene . The induction of mecA transcription in response to β-lactams involves a unique series of proteolytic steps and some critical details of this signal transduction mechanism are still illusive . For instance , it is not fully explained why the induction of mecA by its cognate regulatory genes mecR1-mecI appears to be very inefficient and it is not clear if the observed MecI repressor proteolysis is mediated directly by the activated MecR1 sensor-inducer . In this study , we demonstrate that the mecA regulatory locus is not a two-component system but instead it is a three-component system containing the previously unrecognized anti-repressor mecR2 gene . MecR2 disturbs the binding of the repressor MecI to the mecA promoter , which leads to its proteolytic inactivation independently from MecR1 . Moreover , our data shows that in the presence of functional mecR1-mecI genes , mecR2 is essential for a robust induction of mecA transcription and , as consequence , for the optimal expression of β-lactam resistance .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"bacteriology",
"medicine",
"cellular",
"stress",
"responses",
"mechanisms",
"of",
"signal",
"transduction",
"gene",
"regulation",
"microbiology",
"dna",
"transcription",
"bacterial",
"diseases",
"methicillin-resistant",
"staphylococcus",
"aureus",
"microbial",
"evolution",
"molecular",
"genetics",
"infectious",
"diseases",
"signaling",
"in",
"cellular",
"processes",
"transmembrane",
"signaling",
"medical",
"microbiology",
"gene",
"expression",
"biology",
"staphylococcal",
"infection",
"molecular",
"biology",
"signal",
"transduction",
"cell",
"biology",
"genetics",
"molecular",
"cell",
"biology",
"bacterial",
"evolution",
"genetics",
"and",
"genomics"
] |
2012
|
The Anti-Repressor MecR2 Promotes the Proteolysis of the mecA Repressor and Enables Optimal Expression of β-lactam Resistance in MRSA
|
An important objective in genome research is to relate genome structure to gene function . Sequence comparisons among orthologous and paralogous genes and their allelic variants can reveal sequences of functional significance . Here , we describe a 379-kb region on chromosome 1 of maize that enables us to reconstruct chromosome breakage , transposition , non-homologous end-joining , and homologous recombination events . Such a high-density composition of various mechanisms in a small chromosomal interval exemplifies the evolution of gene regulation and allelic diversity in general . It also illustrates the evolutionary pace of changes in plants , where many of the above mechanisms are of somatic origin . In contrast to animals , somatic alterations can easily be transmitted through meiosis because the germline in plants is contiguous to somatic tissue , permitting the recovery of such chromosomal rearrangements . The analyzed region contains the P1-wr allele , a variant of the genetically well-defined p1 gene , which encodes a Myb-like transcriptional activator in maize . The P1-wr allele consists of eleven nearly perfect P1-wr 12-kb repeats that are arranged in a tandem head-to-tail array . Although a technical challenge to sequence such a structure by shotgun sequencing , we overcame this problem by subcloning each repeat and ordering them based on nucleotide variations . These polymorphisms were also critical for recombination and expression analysis in presence and absence of the trans-acting epigenetic factor Ufo1 . Interestingly , chimeras of the p1 and p2 genes , p2/p1 and p1/p2 , are framing the P1-wr cluster . Reconstruction of sequence amplification steps at the p locus showed the evolution from a single Myb-homolog to the multi-gene P1-wr cluster . It also demonstrates how non-homologous end-joining can create novel gene fusions . Comparisons to orthologous regions in sorghum and rice also indicate a greater instability of the maize genome , probably due to diploidization following allotetraploidization .
Evolution is based on genome instability . Because genome instability can be detrimental to an individual organism , highly sophisticated mechanisms evolved to maintain genome integrity . Processes to prevent instability , such as DNA damage repair systems , however , are error-prone . Consequently , chromosomal changes are passed onto the next generation and will be tested in evolution on the individual and population level . Species as well as inter-species sequence comparisons reveal the dynamic structure of plant genomes as a consequence of genomic instability . It appears that just a few mechanisms are required to explain genomic instability . Minor or local changes that can cause mutations are associated with inaccurate DNA replication , or DNA repair , or recombination [1] . Replication errors , impairment of base excision and mismatch repair , or error-prone translesion synthesis can lead to base substitutions , micro-insertions and micro-deletions . Micro- and minisatellite instability that results in expansion or contraction of short , repetitive sequences is caused by unequal homologous recombination , replication slippage , or by repair impairment . More dramatic or global changes in chromosome structure occur when two DNA fragments that were previously unlinked are being joined . Such chromosomal rearrangements include deletions , insertions , duplications , inversions , and translocations , and they can occur by transposition , unequal homologous recombination , or illegitimate recombination [2] . All of these processes involve DNA Double-Strand-Breaks ( DSBs ) and ligations . Already McClintock demonstrated that chromosomal rearrangements such as translocations , deficiencies , ring chromosomes and end fusions could be consequences of chromosome breaks [3] . DSBs can arise in all tissues at all stages of development and are induced by excision of transposable elements , endonucleases , ionizing irradiation ( UV , decay of naturally occurring radioisotopes ) , reactive oxygen species , and mechanical pulling of dicentric chromosomes . DSBs result in cell-cycle arrest and the recruitment of the DSB-repair machinery . An unrepaired DSB leads ultimately to cell death . Dependent on the phase of the cell cycle , availability of homologous sequences close to the break site , DSBs are repaired by illegitimate recombination ( also known as non-homologous endjoining ( NHEJ ) ) , homologous recombination ( HR ) , or even a combination of both mechanisms ( reviewed in [4]–[6] ) . During meiosis DSBs are probably exclusively repaired by HR . NHEJ is a major pathway for DSB repair in somatic tissue . The rejoining of the broken ends via NHEJ is associated with deletions of various sizes , but also insertions of sequences ( filler DNA ) that are often copied from sites close to the DSB . NHEJ does not require sequence similarities for the incorporation of filler DNA into the break . Taken together , NHEJ does not preserve genetic information and genomic integrity at the break site . Only few cases of filler DNA suggesting a DSB break repair have been reported . HR seems to play a minor role in DSB repair in somatic tissues . Homologous sequences used as template for the repair can be in allelic position ( sister chromatid ) , ectopic , or intrachromosomal . Intrachromosomal homologous recombination is often used to repair a DSB that is caused by the excision of a transposon located between two repeats . The DSB repair results in the deletion of the intervening sequence and one of the repeats [7] , [8] . Similarly , this mechanism also generates solo LTRs that are derived from LTR retrotransposons . Transposable elements contributed tremendously to genome evolution ( reviewed in [2] , [9] , [10] ) . They have modified single genes by transposing within or adjacent to them . Dependent on the insertion site , mobile elements affect genes in various ways . Despite the fact that transposons were discovered because of their chromosome-breaking features , they are mostly recognized for their mutagenic ability to disrupt gene functions . Transposon insertion as well as excision from a coding region with footprint formation can result in a nonfunctional gene product . They can change transcript processing , for example by providing cryptic donor and acceptor sites . Transposons are a source for cis-regulatory elements that can change expression of genes nearby . Insertions in promoter regions , for instance , can add or replace regulatory sequences such that the element gains transcriptional control over the affected gene . While transposons can activate gene expression , they also can cause silencing of adjacent genes . To avoid genomic instability , their hosts epigenetically silence most transposable elements . Silencing by heterochromatin formation is often not limited to the transposon but spreads to neighboring genes as observed in position-effect variegation . Despite transposon silencing mechanisms imposed by the host , transposons play an important role in plant genome expansion . Class I transposons transpose via a copy-and-paste mechanism , thereby generating additional elements . But also DNA transposons that employ a cut-and-paste mechanism for transposition increase their copy numbers . They amplify by transposition from a replicated donor site on one of the sister chromatids to a yet unreplicated insertion sequence . Amplification also occurs when a transposon is copied from a template into the empty excision site via homology-based gap repair [11] . To study genomic instability , we favored a gene that is ( 1 ) genetically well defined , ( 2 ) offers allelic variability , including epigenetically regulated alleles , and ( 3 ) has paralogous gene copies . The p1 gene of maize at the short arm of chromosome 1 ( bin 1 . 03 ) , which encodes an R2R3 Myb-like transcriptional activator meets such properties . p1 controls the structural genes c2 , chi1 , and a1 of the phlobaphene biosynthesis pathway [12] . Phlobaphenes are reddish flavenoid pigments that are frequently found in male and female maize floral organs . Genes involved in flavonoid pigment biosynthesis are well suited to study numerous aspects of genomic instability because they are dispensable for the organism and generate a visible , quantitative phenotype . Various alleles of p1 with distinct tissue-specific expression have been collected and investigated . Traditionally , p1 alleles have been classified phenotypically according to pericarp and glume pigmentation [13] . The p1 alleles are designated with a two-letter suffix that refers to pericarp and cob coloration , respectively . For example , the P1-rr allele displays red pericarp and red glumes , whereas P1-wr has white , or more precisely , colorless pericarp and exhibits red glume pigmentation ( Figure 1 ) . Although numerous alleles are genetically well described , few of them are sequenced , and often not to completion [12] , [14] , [15] . The best-studied alleles are P1-rr [12] ( and its derivatives P1-vv , P1-ovov , p1-ww1112 ) , followed by P1-wr [14] , [16] and most recently P1-rw [15] . The p1 gene is thought to have arisen by a duplication event from an ancestral p gene that is closely related to the recently isolated p2 gene from a teosinte parviglumis accession [17] . The p2 gene in maize , which is tightly linked to P1-rr , is located proximal of p1 in the same transcriptional orientation . The teosinte p2 gene will be referred to as p2-t and the maize p2 gene , which was isolated from a line carrying the p1 null allele p1ww1112 , will be designated as p2-m throughout the remaining text . In contrast to p1 , p2 does not induce visible phlobaphene synthesis in maize tissues [17] . p2 in teosinte , however , confers pigmentation to tassel glume margins . The gene duplication occurred approximately 2 . 75 million years ago [17] . Although the name p2-t implies the existence of a p1-t gene it remains to be seen whether the p2-t gene in teosinte is duplicated . From all alleles of the p1 gene the P1-wr allele is the technically most difficult allele to characterize because of its expanded size and repeat structure . Comparison of different p1 alleles in maize and its organization in related species of the same family illustrates various molecular mechanisms that have changed entire plant genomes . To distinguish these from lineage-specific events we also compared orthologous regions containing the p1 locus from rice , sorghum , and the homoeologous region in maize that resulted from an ancient allotetraploidization event .
Molecular analysis of the multi-copy P1-wr allele ( Figure 1 ) required the isolation of three overlapping BAC clones from inbred line B73 ( Methods ) . The large size of this locus is due to nearly perfect tandem genic repeats . The structural analysis of such repetitive genomic sequences requires a different strategy than the standard shotgun sequencing approach . Indeed , the P1-wr structure is not available from the current Maize Genome Sequencing Project . The P1-wr repeats created a gap in the minimum tiling path that was closed with our contiguous P1-wr sequence permitting the merger of finger-printed contigs ( FPCs ) 11 and 12 ( Figure 2 ) [18] . Sequencing required subcloning of individual repeats using conserved restriction sites and then ordering repeats based on nucleotide polymorphisms [19] . Using such an approach , we identified eleven P1-wr copies ranging from 12 , 602 bp to 13 , 026 bp in size in a contiguous sequence of 379 kb ( Figure 3 ) . The P1-wr repeat is defined as the sequence starting from the first nucleotide downstream of exon 3 of the previous copy to the last nucleotide of exon 3 . The P1-wr repeats , which are highly similar ( see below ) are named according to their order in the array , starting with P1-wr-1 for the most 5′ repeat . All numbers here refer to P1-wr-1 , which was analyzed as a prototype P1-wr repeat . P1-wr-1 , which is 12 , 648 bp in size , comprises 6 , 314 bp of transcribed sequence and a 6 , 334-bp region upstream of the transcription start site . Unless otherwise noted , nucleotide positions given for the P1-wr-1 sequence refer to the transcription start site . The P1-wr repeats are flanked by two genes that resemble p2 ( Figure 3 ) . Immediately upstream of the first P1-wr repeat is a chimeric p sequence that consists of a p2-like 5′ end , followed by a P1-wr 3′ end . Therefore , this sequence will be designated as p2/p1 . The sequence located downstream of the most distal P1-wr repeat resembles P1-wr in the regulatory regions and p2-m in the majority of the assumed transcribed part . Accordingly , this potential gene will be named p1/p2 . The 3′ end of p1/p2 is displaced by numerous retroelement insertions . Besides p-related genes , the analyzed sequence includes only two more predicted genes , which are located downstream of p1/p2 , one potentially encoding a calmodulin binding protein , the other an expressed protein , based on EST data and their syntenic positions in rice and sorghum ( see below ) . In addition , several pseudogenes are present in the P1-wr cluster region . The gene encoding an expressed protein is upstream of a Helitron terminal sequence , which can form a hairpin structure . The pseudogenes downstream of this Helitron terminus are potentially fragmented genes incorporated in the Helitron transposon . The cluster contains various transposable elements such as multiple LTR retrotransposons inserted mostly in a nested fashion , two CACTA elements ( misfit and doppia ) , one hAT element , one LINE element and several MITEs . p2/p1 is proximal , p1/p2 is distal to the centromere ( Figure 3 ) . As shown for P1-wr [W23] in transgenic studies , the upstream sequence contains all regulatory elements necessary for expression in pericarp and cob [20] . In P1-rr , a 235-bp fragment immediately upstream of the transcription start site has basal promoter functions ( Figure 4 ) [21] . The corresponding fragment in P1-wr-1 is identical to P1-rr excluding a 19-bp and 36-bp insertion in P1-wr-1 . A 1-kb HindIII-SalI fragment upstream of the basal promoter was previously identified as a P1-rr enhancer . This sequence is well conserved in P1-wr-1 varying only in four SNPs and six 1-bp indels . Structural analysis of this potential P1-wr-1 enhancer revealed a complex sequence composition that includes two almost perfect inverted repeats ( IR ) of 199 and 186 bp separated by 362 bp , which originated from a previously uncharacterized Mu-like transposable element that became truncated after insertion in the p1 gene . Further details about this new element family are provided in the supplemental material ( Text S1 , Figure S1 ) . This Mu-like transposon ( position -102 to -1072 ) occupies the greatest part of sequences defined as the proximal enhancer ( position -291 to -1301 ) and promoter ( position -1 to -290 ) ( Figure 4 ) . The element in the proximal enhancer region of p1 contains a sequence inserted between both TIRs , derived from the first intron of a calcium-dependent protein kinase gene on chromosome 10 . This structure and various truncated derivatives are present seven more times in the B73 genome . Compared to the transposon that is closest to the capturing event of the intron , the element in p1 lacks 226 bp of the 5′ TIR including the TSD and 200 bp from the center of the gene fragment . The TIR deletion break point is adjacent to a 15-bp direct repeat . A Heartbreaker MITE can be found 122 bp 5′ of the complex IR structure or 1 , 195 bp upstream of the transcription start site . Therefore , it appears that multiple double-strand breaks and repairs have to be invoked to compose a regulatory sequence consisting of a truncated Mu-like transposon and part of a MITE that are separated by 122 bp . Though nearly identical , P1-wr-1 differs from P1-rr in a region that contains the distal enhancer of P1-rr because it shares only a 408-bp sequence corresponding to the 3′ end of the distal enhancer of P1-rr . This sequence is located 4 , 886 bp upstream of the transcription start site ( position -4 , 886 to -5 , 293 ) . Despite this truncation , P1-wr regulatory sequences contain all elements necessary for expression in pericarp tissue [20] . The exon-intron boundaries of P1-wr repeats in B73 can easily be defined using the P1-wr cDNA from inbred W23 , as well as sequence alignments with the P1-rr allele . The transcript of P1-wr-1 is 1 , 610 nucleotides in size . P1-wr-1 [B73] is identical to P1-rr in the first exon , except for one non-synonymous substitution that converts the fourth amino acid residue alanine in P1-wr to threonine in P1-rr . This change does not affect the Myb domain of the P1-wr protein , which starts at the 12th residue . The second exon , which is only 130 bp in size , is identical in P1-wr-1 [B73] and P1-rr . P1-wr-1 [B73] and P1-rr do not differ in the coding sequence of exon 3 . Besides the substitution of the fourth amino acid residue , the deduced P1-wr-1 [B73] and P1-rr proteins are identical . The proteins , which are 335 amino acids in length , contain a conserved R2R3 Myb domain and a P protein specific activation domain . The Myb domain is located at the N-terminus including residues 12–115 . The acidic activation domain is 44 amino acids in size starting with residue 201 . P1-wr intron sequences are described in the supplemental material ( Text S1 ) . The alignment of P1-wr repeats reveals 103 polymorphic sites , spread over regulatory and transcribed sequences . Compared to the P1-wr consensus sequence of 13 , 172 bp , P1-wr repeats differ in 67 SNPs , 20 insertions and 16 deletions ( Figure 5 ) . Two indels have features of a remnant hAT-like transposable element . Polymorphisms that are shared by at least two P1-wr repeats indicate amplification of the P1-wr cluster by recombination ( unequal crossovers ) and/or gene conversion . 13 SNPs , 7 insertions , and 2 deletions are present in more than one P1-wr repeat . Still , long stretches of complete identity cannot be detected , demonstrating extensive reshuffling of the P1-wr repeats by numerous recombination events . Polymorphic sites are not evenly distributed across the P1-wr repeats . Interestingly , the most frequent mutations among the P1-wr copies can be found in both TIRs of the Mu-like element that contributes to the proximal enhancer and promoter region , although preliminary results indicate no effect of these polymorphisms on transcription rate . The alignment of P1-wr sequences exposes in P1-wr copies 6 and 11 a large insertion of 382 bp in the 5′ UTR of the first exon , 115 bp after the transcription start site . The insertion sequence exhibits features of an Ins2 transposable element as described in the supplemental material ( Text S1 ) . Because two P1-wr repeats contain this insertion , the transposition event probably occurred during P1-wr amplification . With exception of the Ins2 insertion , all repeats are identical in the first and second exon . Four SNPs , three insertions and two deletions are located in the larger exon 3 that can be used to distinguish P1-wr transcripts . All SNPs and most indels have no effect on the deduced protein sequence . Only a TGC-insertion in P1-wr-5 and a TGC-deletion in P1-wr-8 adds and deletes one alanine from a seven-alanine repeat at the C-terminal end , respectively . Polymorphisms in P1-wr transcribed regions enable us to investigate which repeats are expressed . In addition , differences in coding sequences are most informative regarding gene products . Total RNAs were extracted from pericarp tissue and reversed transcribed . Several primer pairs were used in PCR reactions that amplify polymorphic sequences of all P1-wr repeats , p2/p1 , and p1/p2 . Genomic DNA served as a control for presence and ratio of individual P1-wr repeats . The amplified products were cloned , sequenced , and analyzed . Using primer pair p-ex3-3 that spans four P1-wr repeat polymorphisms in exon 3 , four out of eleven P1-wr repeats and p1/p2 can be distinguished based on SNPs and indels in the amplified region ( Figure 5 ) . By utilizing different primer pairs in the same approach , the transcript analysis revealed that at least eight P1-wr repeats and p1/p2 are expressed ( data not shown ) . Included are P1-wr repeats 6 and 11 that contain a hAT-like element in the 5′UTR of exon 1 . Interestingly , the hAT insertion interferes with the correct splicing of intron 1 and prevents synthesis of functional transcription factors . We also investigated expression of P1-wr repeats in the presence of Ufo1 , an epigenetic modifier of P1-wr . Consistent with increased pigmentation levels in p1-expressing tissues by several fold [22] , we found that p1 transcript levels in an UfoI mutant background are augmented ( data not shown ) . However , none of the individual P1-wr repeats analyzed here seems to be preferentially activated in the UfoI background , indicating a mechanism that affects all P1-wr repeats in a similar way . While the SNPs in the P1-wr repeats indicate homologous recombination events during meiosis , the chimeric p2/p1 gene upstream of the P1-wr repeats ( Figure 3 ) arose by an entirely different mechanism . Until the end of the first repeat , p2/p1 is most closely related to p2-t ( 84 . 5% nucleotide identity ) . Downstream at the 3′ end , p2/p1 closely resembles P1-wr ( Figure 4 ) . This gene copy is expressed in silk ( data not shown ) and encodes a protein identical to P1-wr . Based on the p2 regulatory sequences of p2/p1 , p2/p1 and p2 are transcribed together in the same tissues at the same time . Although its regulatory sequences have not yet been determined , the initial 92 bp ( counting from the transcriptional start site ) , p2-m and p2-t are 88% identical to the basal promoter of p1 alleles . Because the 92-bp fragment is the only sequence common in p2 and p1 and silk is the only tissue where both p1 and p2 are expressed , it is conceivable that the 92-bp sequence contains a regulatory element necessary for gene expression in silk . Although in exon 1 , intron 1 , and exon 2 p2/p1 resembles p2-t more than P1-wr or P1-rr , the structure of the second intron differs remarkably from previously investigated p alleles . Intron 2 with 7 , 000 bp is significantly longer than in other p1 and p2 genes . It contains large direct imperfect repeats , 2 , 377 bp and 2 , 701 bp in size , beginning at 1 , 759 bp and 4 , 186 bp after the transcription start site , respectively . Both repeats are separated by a 50-bp sequence of known origin ( Figure 6 ) . The first 30 bp ( CATATTACTACAGTGCATATATGTGAGAAA ) are identical to the initial sequence of the second repeat ( 4 , 186–4 , 215 bp after the transcription start site ) . This sequence is followed by 19 bp ( ACAATATGGCCATCTGGTC ) that are also derived from the second repeat few nucleotides downstream of the first duplicate ( 4 , 278–4 , 296 bp after transcription start site ) . The 50th nucleotide A is unaccounted for . The 50-bp sequence can be clearly assigned to the second repeat , due to SNPs between the repeats . This 50-bp sequence is suggestive of filler DNA , which is associated with NHEJ and , therefore , p2/p1 originated from somatic tissue . The chimeric p1/p2 gene downstream of the most distal P1-wr repeat is 98% identical at the nucleotide level to P1-wr and P1-rr ( Figure 4 ) and is described in more detail in the supplemental material ( Text S1 ) . While the upstream chimeric p2/p1 gene is unusual because of the presence of a filler DNA , this chimeric gene has an unusual 3′ end . 247 bp after the stop codon , the homology to P1-wr is completely lost . Like in p2-m and P1-rw , an Eninu retrotransposon LTR follows the point of divergence in p1/p2 ( Figure 7 ) . The 5′ end of the Eninu LTR is difficult to identify because it is not well conserved relative to other Eninu LTRs . The endpoints of p2 and P1-rw transcripts have not been determined yet . The Eninu insertion happened upstream of a putative polyadenylation signal sequence AATAAA used for transcript termination . Displacement of this hexamer sequence in p1/p2 and p2-m requires a sequence in the retroelement that substitutes for the original poly-A signal element . The p1/p2 sequence continues in the 3′ UTR 68 , 190 bp further downstream , precisely where the homology to P1-wr and p2-t stopped before , indicating that a large insertion split off the p1/p2 3′ end ( Figure 7 ) . This insertion is bordered downstream by an Eninu retrotransposon LTR , which lacks approximately 700 bp from the 5′ end . Both Eninu LTRs are flanked by identical 5-bp sequences AAGAC , which identifies them as target site duplication sequences ( TSD ) caused by LTR retrotransposon insertion . This suggests that these LTR pairs belong to the same Eninu retroelement ( see below ) . The remaining 61 bp of exon 3 are followed by 4 , 100 bp 99 . 2% identical to the initial sequences of a P1-wr repeat ( beginning with 6 , 300 bp upstream of the transcription start site ) . This displaced P1-wr-like sequence reveals similar features and structures to P1-wr , including the distal enhancer region . The homology is interrupted by the insertion of a single LTR retrotransposon Shadowspawn . The homology to P1-wr terminates 769 bp after the LTR retrotransposon insertion . Interestingly , the homology to P1-wr ends exactly 5 bp before a fragmented Mu-like transposable element that has been shown to be part of the proximal enhancer of p1 alleles ( see above ) . Accordingly , proximal enhancer , promoter , and coding sequences are missing compared to a full length P1-wr repeat . Nine LTR-retrotransposons have been identified in the P1-wr cluster , not taking into account retroelements in flanking regions ( Figure 7 ) . Six retrotransposons have complete and conserved pairs of LTRs , allowing the use of the total LTR lengths for the dating of their divergence . The initial 700 bp of the 3′ LTR of the Eninu element are missing . Thus , only the remaining 635 bp from both LTRs were considered for the calculation of K . One of the Opie elements contains only one LTR and almost 5 kb of internal sequence . One of the two Diguus retrotransposons consists of only one LTR . The insertion time cannot be determined for the latter incomplete elements . Eight out of nine retroelements are inserted in a nested fashion as described for the Adh1-F region previously [23] . Intriguingly , eight out of nine LTR retrotransposons are inserted in the same transcriptional direction suggesting a preferred , rather than random , orientation upon insertion . The structural organization of the P1-wr cluster in maize is rather complex , involving the p2 gene and multiple copies of the paralogous p1 gene in a head to tail array . However , single copy p1 alleles such as P1-rr [24] and P1-rw [15] have been described previously . To analyze the highly dynamic p locus in an evolutionary context , we searched for orthologous gene copies in the close relatives , rice and sorghum , and in the homoeologous chromosome segment of maize . In rice ( Oryza sativa , japonica cultivar Nipponbare ) , the sequence most similar to a single maize P1-wr gene can be discovered on the short arm of chromosome 3 ( position 10 , 763 , 678 bp–10 , 757 , 863 bp ) [25] . Sequence alignments identify a single coding region that is 68 . 8% identical at the nucleotide level with p1 . Orthologous sequences are also present in Oryza sativa indica 93-11 , and wild relatives of rice , Oryza glaberrima , Oryza punctata , Oryza minuta , and Oryza officinalis ( data not shown ) . However , no matching transcript could be found in the rice EST collections . The deduced protein sequence is 332 aa in length , and is 56 . 3% identical with P1-wr of maize . Compared to P1-wr , the first 150 aa containing the R2R3 Myb-domain are highly conserved , while the C-terminus is more variable . Synteny among flanking genes is well maintained . They are present in rice in the same order and orientation , but spread over a region of only 82 kb . As expected , fragmented or pseudo-genes present around the P1-wr cluster in maize are absent from the corresponding “p” region of rice . The syntenic region in Sorghum bicolor , inbred line BTx623 , is located on the long arm of chromosome 1 ( about 74 Mb total length ) between 61 . 203 Mb and 61 . 309 Mb [26] . The genomic arrangement in sorghum is more complex than in rice , as it contains three p1-homologous genes . However , the collinearity of flanking genes is well conserved . A functional p1 ortholog , designated y1 ( yellow seed 1 ) , was described in inbred line CS8110419 [27] , [28] . The y1 gene encodes a protein of 383 aa in length that is 68 . 5% identical at the amino acid level with P1-wr . In the inbred line BTx623 that was sequenced and used in our analysis y1 is partially deleted [29] . The 3 , 218 bp deletion includes 5′ non-transcribed sequences , exon 1 , intron 1 , exon 2 , and parts of intron 2 . The y1 gene in BTx623 is a null allele that is not transcribed [29] . A second p1-homologous pseudogene , y2 , is located 3′ of y1 in direct orientation [27] . The y2 gene , which is nearly identical in lines BTx623 and CS8110419 , shows multiple indels of various sizes compared to y1 . The largest deletion includes intron 1 , exon 2 , and part of intron 2 , whereas a smaller deletion removed the putative translation initiation codon . The y2 transcripts were not detected by RT-PCR [27] . The investigated region contains a third gene copy that is homologous to the maize p1 gene . This putative gene , which we named y3 ( yellow seed 3 ) in accordance with y1 and y2 , is located 29 , 877 bp upstream of y1 . Based on RT-PCR experiments , y3 is expressed in panicles at the time of anthesis ( data not shown ) . Due to the lack of full-length transcript data , we deduced the exon-intron structure of y3 from sequence alignments with y1 . The largest ORF of y3 potentially encodes a protein of 356 aa in length that is 58% and 61 . 2% identical at the amino acid level to Y1 and P1-wr-1 , respectively . While the R2/R3 Myb domain and the acidic activation domain of Y3 are intact , a duplication of 146 bp in exon 3 and a truncation of 192 bp compared to y1 at the 3′ end could render y3 also nonfunctional because based on the white seed phenotype of BTx623 y3 cannot substitute for y1 . Surprisingly , the sequence 280 bp upstream the transcription start site that potentially carries promoter elements is well conserved in blocks compared to y1 , p2 from parviglumis and maize , with an overall nucleotide identity of 70% . The 3′ non-transcribed region has no similarity to y1 , y2 , p2 and P1-wr , which could be a continuation of the deletion of the 3′ coding sequence . The sequence separating y1 and y3 mostly consists of repetitive elements , but also harbors a gene 4 , 552 bp 3′ of y3 that encodes a conserved hypothetical protein of 178 aa in size . This gene is duplicated 23 , 459 bp upstream of y3 showing 91 . 2% nucleotide identity in a sequence that corresponds to the coding sequence of the downstream copy . However , this duplicated sequence is most likely a pseudogene , because indels cause several reading frame shifts resulting in a premature stop codon . Because both genes of unknown function and two y genes are arranged in an alternating pattern , they were probably duplicated together . To establish the order of duplication events that led from an ancestral y gene to three y copies we constructed a phylogenetic tree based on the coding sequences of exon 3 . We were limited to exon 3 in our phylogenetic analysis , because exon 1 is partially deleted and exon 2 is completely deleted in y2 . In addition to sorghum y genes , we included the p orthologous sequence from rice , p1 , p2 and the homoeologous p gene from maize ( see below ) in our multiple sequence alignments . The phylogenetic tree ( not shown ) reveals that the duplication of the ancestral y gene generated y3 and the y copy that in a second duplication event gave rise to y1 and y2 . Possible recombination sites flanking the y genes could not be identified . Presumably , the intergenic spaces were completely reshuffled by transposable elements since the duplications occurred , leading to a loss or change beyond recognition of recombination sites . The phylogenetic analysis also reveals that the sorghum y1 , y2 and y3 genes , and the maize p1 ( namely P1-wr-1 and P1-rr alleles ) and p2 genes cluster together . This is consistent with the fact that paralogous sequences arose after maize and sorghum shared their last common ancestor . Therefore , the amplification of p genes in maize and y genes in sorghum represent independent , parallel events . This finding is supported by the computed duplication times of p1 and p2 ( 2 . 75 mya [17] ) and y1 and y2 ( 9 . 08 to 11 . 3 mya [27] ) that are both younger than the maize and sorghum divergence time ( 11 . 9 mya ) [30] . As maize originated from an allotetraploid event [18] , we wanted to see whether a p-like sequence is retained on a homoeologous chromosome segment . Indeed , we found a highly similar sequence on the long arm of chromosome 9 , which we named p3 ( pericarp 3 ) consistent with the homoeologous copies p1 and p2 . According to EST data ( GenBank accessions EB702996 , EB702997 ) derived from mixed tissue ( silks , husks , ears , pollen , shoot tips , leaf , root tips , whole seed , embryo ) , p3 is transcribed . We verified the expression of p3 in silk tissue using RT-PCR ( data not shown ) . The coding sequences of P1-wr and p3 share 87 . 3% nucleotide sequence identity . The deduced protein is 344 aa long and has an overall amino acid identity with P1-wr of 79 . 9% . The sequence from the transcription start site until position -312 bp that potentially contains promoter elements such as the TATA box is well conserved in p3 compared to y3 , y1 and p2 . Genes bordering p3 on the centromere side are collinear with rice , sorghum and maize chr1 . However , there is a break in synteny at the telomere end . Six genes are missing in this chromosome segment before synteny with rice , sorghum and maize chr1 resumes again ( data not shown ) . While rice appears to have a relatively stable p locus , sorghum exhibits some degree of genome instability at its y locus . Clearly , in respect to ancient and recent genome instability the duplicated maize p loci stand out . Moreover , among all the maize p1 alleles , the P1-wr allele underwent the most complex series of chromosomal changes .
The filler DNA in the second intron of p2/p1 is evidence for a DNA DSB and its repair . Presumably the DSB was repaired via the NHEJ pathway that resulted in the deletion of the 3′ end of p2 ( from part of the second intron and the complete third exon ) and the 5′ end of a P1-wr unit ( until part of the second intron ) , incorporation of filler DNA and ligation of the broken ends ( Figure 6 ) . The current configuration can be used to determine the size , origin and complexity of the inserted filler DNA . The deletion size may depend on the kind of break-inducing mutagenic agent , on exonuclease activity , and efficiency of DNA end protection . Previously reported NHEJ deletion sizes in maize genes range from 340 bp in bz1 [31] , 60 to 980 bp for wx [32] , and 400 to 4 , 300 bp in R-r deletion derivatives [33] . Although a deletion size of about 10 kb for p2/p1 would be the largest documented so far , the size seems to be in accordance with the examples above . Filler DNA is derived from sequences close to the deletion endpoint . It can be simple or complex , meaning a mosaic of different sequences of different origin . The filler DNA in p2/p1 is 50-bp in size and is copied from two sites . Filler DNA can be as small as 8 bp as observed in bz1 [31] , or 1–131 bp as in wx [32] , or 31 to 84 bp in R-r [33] . Only the filler DNA in R-r is of complex origin . DSB repair via NHEJ mostly results in a recombination event that links two previously separated sequences . Therefore , the mechanism of NHEJ can be a source for creating new genes , as demonstrated here . Dependent on the size of the repair-induced deletion and distance between neighboring genes , two adjacent genes with the same transcriptional orientation can be merged , thus forming a novel hybrid gene . Based on the joining site , the fusion event may add or replace regulatory sequences as well as add or delete exons and introns . Here , we describe the likely fusion of two paralogous sequences with high sequence similarity . The new hybrid gene p2/p1 was produced from almost identical , recently duplicated , genes that exhibit the same exon-intron structure but vary in their regulatory sequences . The deletion end points in both genes happened to be in the second intron , but at different positions . Therefore , the hybrid gene p2/p1 maintains the overall gene structure of the parental genes , but slightly differs from p2 in the third exon . Due to the high sequence similarity between p2 and P1-wr in exon 1 and 2 , the deduced gene product of p2/p1 is identical with P1-wr . However , because P2 and P1-wr are almost identical and interchangeable , a functional change at the p cluster did not occur or is minimal . Despite the fact that the described fusion event has most likely no impact on the p cluster and downstream genes , this is , to our knowledge , the first evidence that NHEJ is linked to the formation of new , and functional genes in plants . While DSBs in meiosis are solely repaired by homologous recombination , NHEJ is the repair pathway for DSBs in somatic tissue . All premeiotic events will only have an impact on evolution , when induced changes are transmitted to the offspring . Therefore , the importance of NHEJ-associated gene rearrangements depends on the amount of repaired DSBs in somatic tissues and the fraction of those events that are passed through meiosis . Each observed filler DNA in a maize population represents a somatic clonal sector in meristematic tissue that gave rise to the plant germline . Although filler DNAs as visible markers of a DSB repair event have been reported in only few maize loci , we hypothesize that they are rather widespread in the maize genome . For instance , filler DNAs were identified in various other p1 alleles ( unpublished ) . Successful searches for genomic rearrangements and filler DNAs depend on the comparison of the sequence of interest with a reference allele/locus of known lineage . However , in absence of such a reference , detection of filler DNA requires intragenomic sequence alignments , probably leading to an underestimation of the importance of NHEJ for plant evolution . Transposable elements have shaped plant genomes in various ways ( reviewed in [2] , [9] , [10] ) . Most obviously , amplification of mobile elements , especially LTR retrotransposons , led to genome enlargement , to differentiation among species and homoeologous chromosomes , and even to allelic variation . This extensive expansion and contraction due to LTR retrotransposon insertions can also be observed at the P1-wr cluster . Insertions of nine LTR retrotransposons into the p1/p2 gene , which occurred between 1 . 4 and 0 . 2 mya , resulted into a 68-kb expansion of the P1-wr cluster ( Figure 7 ) . Transposition in a mostly nested fashion fragmented four LTR retrotransposons while five remained intact . Among nine elements , Opie and Diguus are present twice . Retroelement insertions into the p locus resemble transposition events at the r/b loci of maize in terms of insertion times and structure [34] . Only one solo LTR has been detected within this cluster . Solo LTRs can be generated by ectopic and intra-element recombination . DSBs between two repeat sequences are frequently repaired by intrachromosomal/intra-element recombination due to homologous sequences in close proximity to the DNA break [4] . Intra-element recombination events result in the deletion of one LTR and sequences in between them . Therefore , the TSD flanking the solo LTR are identical . Unequal crossovers involving homologous LTRs from ectopic positions cause segmental duplication and deletion of sequences . Because the TSD are derived from two independent insertion events , the short flanking sequences of the remaining LTR are different . The Diguus solo LTR is the outcome of intra-element recombination , because the TSD delimiting the LTR are identical . In the past 1 . 4 million years , the expansion of the maize genome within the p cluster due to retroelement transposition prevails the contraction due to deletion by about ten fold ( assuming that only internal Diguus sequences were removed ) . In general , the size increase of the P1-wr cluster reflects the expansion of the maize genome . The displacement of the 3′ UTR of p1/p2 due to transposon insertions generates a distant site for homologous recombination separated by retroelements that may suppress recombination frequencies in surrounding regions . Transposon mutagenesis in sequences upstream of the coding region of P1-rr revealed promoter and enhancer regions important for P1-rr regulation [21] , [35] , [36] ( Figure 4 ) . The putative regulatory fragments have been further characterized with a GUS reporter gene and tested in transient expression assays and in stably transformed plants [21] . Most of the promoter region and the putative proximal enhancer in P1-wr as well as P1-rr are composed of two transposable elements , namely 971 bp of an unknown Mu-like element ( Mule , see also Text S1 ) and 107 bp of a MITE , separated by a unique 122-bp sequence . The question arises which of the three fractions contains the actual cis-acting element ? It is tempting to speculate that the Mule contributes to the enhancer function in the p1 gene . TIRs of Mutator elements include regulatory sequences that are required for the transcription of its mudrA and mudrB genes . Although this Mule has not been functionally characterized , it is likely that the 425 bp TIRs harbor regulatory sites in addition to the ones used by the tRNA lys gene . The Mule captured a gene fragment exactly between the TIRs , which is derived from the first intron of a calcium-dependent protein kinase gene located on chromosome 10 . It is also possible that this acquired sequence is the source of the enhancer function . P1-wr was shown to be posttranscriptionally silenced [22] possibly due to a tissue-specific repeat-induced gene-silencing ( TRIGS ) mechanism [14] . Is a transposable element conceivably involved in such a TRIGS mechanism ? Transposons are subject to epigenetic silencing and most often dormant . TIRs of inactive Mu elements are heavily methylated which prevents transcription from the embedded promoter sites . Assuming that the Mule carries the proximal enhancer sequence , would silencing also contribute to P1-wr suppression ? Is it possible that the same transposon sequence has activating and depressing functions dependent on the p1 allele ? While P1-rr may benefit from the transposon sequence by an increase in expression , epigenetically silenced alleles such as P1-wr and P1-pr [22] , [37] may fall victim to transposon silencing . Interestingly , compared to P1-rr , the promoter and upstream regions of P1-wr are extensively methylated at HpaII/MspI sites [14] . We located p orthologous sequences in the rice and sorghum genomes , and confirmed their homology by descent with their syntenic positions . The rice genome contains only a single p orthologous sequence on the short arm of chromosome three . In contrast to the single p gene in rice , the sorghum-inbred line BTx623 contains three p equivalents ( y genes ) in a tandem array with y1 and y2 as non-functional deletion derivatives . Y3 does not rescue the colorless mutant phenotype of BTx623 , indicating that Y3 is possibly non-functional as well . Alternatively , y3 could be expressed in different tissues similar to p1 and p2 . Or Y3 has simply a totally dissimilar function . Two features of the grass genomes become immediately apparent ( Figure 8 ) . Although rice and maize diverged about 50 million years ago , the gene order between rice , sorghum and maize chromosome 1 and 9 is well conserved at the p orthologous regions . The only break in synteny occurred at maize chromosome 9 due to the removal of six genes ( two missing genes in Figure 8 ) flanking the p homoeolog . The second characteristic feature is the overall genome expansion from rice to maize . The presence of two p-like genes in a syntenic arrangement of neighboring genes on different maize chromosomes is the result of allotetraploidization [18] . The intergenic space in maize is increased compared to sorghum and rice due to LTR retrotransposon insertions [38] . Tandem duplications as seen here at the P1-wr cluster in maize also contribute to the growth of the genome size . The amplification of p-homologous genes in maize occurred in three steps at different times . Because of allotetraploidization , the first duplication of p progenitors took place with the hybridization of both parental genomes . As shown here , both ancestral p-like genes are retained in the modern maize genome . While p is well characterized , the function of its homoeologous counterpart on chromosome 9 is not known . While P1 is the only transcription factor required in the phlobaphene biosynthesis pathway [39] , it together with P2 is also involved in maysin accumulation . Maysin is a C-glycosyl flavone found in silk that confers resistance to corn earworm ( Helicoverpa zea , Boddie ) [40] . So far , no QTL for maysin synthesis has been reported at the chromosomal position of p3 [40] , [41] . The second p amplification happened by an unknown event as the result of a single gene duplication in contrast to whole genome duplication ( WGD ) [17] . It could have been the result of recombination between small homologous sequences such as MITEs , frequently found in or near genes [26] . This recombination event almost duplicated the entire gene because a site in an intergenic region approximately 5 kb downstream of the last exon was linked to a sequence 100 bp upstream of the transcription start site . The recombination sites can be inferred by aligning current sequences in the P1-wr cluster , because flanking sequences are usually maintained in unequal crossover , whereas gene numbers can change . Alternatively , one can envision that the truncated Mule caused a DNA DSB upon excision and initiated a repair/recombination event resulting in the duplication of the ancestral p gene . The Mule , which is absent in the proposed structure of an ancestral p gene , can be found at the putative recombination junction supporting this possibility . For example , an aberrant transposition event could have destabilized the p1 and p2 progenitor . Typically , a defective copy of the transposable element that triggered the genomic instability stays at the restructured site [33] . Although a mechanism for such gene duplication is not known , this model for initial amplification is more convincing than unequal crossover , especially because homologous sequences at the recombination junctions have not been detected . Not only would have the transposable element initiated a gene copying event but also added enhancer functions for p1 alleles . Insertion of this transposon 100 bp upstream of the transcription start site could have activated and/or altered the expression pattern of the second copy now known as p1 . A DNA fragment within a 400-bp sequence 3 . 8 kb further upstream in the intergenic region was recruited as a distal enhancer element for p1 . This scenario would provide a model how transposable elements directly would contribute to the subfunctionalization of gene copies . Similar to the first amplification step by polyploidization , none of the duplicated genes were removed from the genome or turned non-functional . p1 and p2 are the principle regulatory genes of the flavone pathway necessary for maysin synthesis . Although under distinct tissue-specific regulation p1 and p2 encode exchangeable products [42] . Based on our model , which is enhanced by the p2-t sequence from teosinte , the progenitor of modern maize , the ancestral p gene might have had a similar expression pattern as p2 . Although p2 is expressed in silk and anthers , p2 does not confer phlobaphene pigmentation to any floral tissue , including pericarp and cob glumes . Therefore , it is feasible that the ancestral p gene primarily evolved in maize to protect silk and anthers from corn earworm damage . With p1 acquiring new regulatory sequences due to the tandem duplication , expression of p1 was extended to additional tissues such as pericarp and cob glumes where it is involved in pigmentation as secondary function . Accordingly , the tandem duplication appears to have resulted in subfunctionalization of the original function of the ancestral p gene . The third amplification of p genes is allele-specific . While single copy p1 alleles were characterized before [12] , [15] , multicopy alleles such as P1-wr are less well studied . In the B73 inbred line , amplification in tandem increased the P1-wr copy number to eleven , excluding the p1/p2 hybrid gene . Due to their high sequence identity , tandem arrays could misalign in meiosis , which adds or deletes copies in case of a crossover event . Again , recombination is initiated by DNA DSBs , and its mechanism can be explained with the double-strand break repair ( DSBR ) model [43] , [44] . The tandem repeats are not only subject to reciprocal unequal crossover , but also unidirectional gene conversion , where the acceptor sequence is replaced by a sequence copied from a donor . Gene conversion could homogenize paralogous gene copies , consistent with concerted evolution . While none of the P1-wr repeats are identical , stretches of several kb in size share polymorphisms between two or more copies ( Figure 5 ) . The length of potential conversion tracts at the P1-wr repeats is supported by data derived from the bz locus , where conversion tracts are between 1 and 1 . 5 kb long [45] . Tandem duplications generate diversity due to the process of unequal recombination . Because tandem arrays are common in plant genomes , their effects on evolution are quite significant . Tandem arrays giving rise to novel alleles were reported for numerous maize loci such as the 27-kDa zein [46] , kn1 [47] , pl1 [48] , a1 [49] , R-r [33] , R-st [50] , rp1 [51] and rp3 [52] . Any model for the evolution from a simple ancestral p gene to the complex multigenic P1-wr cluster has to explain ( 1 ) the creation of p1 and p2 , including their distinct expression pattern , ( 2 ) the formation of the p2/p1 hybrid gene , ( 3 ) the amplification of P1-wr copies , ( 4 ) the formation of the p1/p2 hybrid gene , and ( 5 ) the existence of a p-homologous sequence separated from the main complex . At least several DNA breaks were required to remodel the p locus . Mechanisms that involve DNA DSBs either as cause or consequence are homologous recombination ( crossovers as well as gene conversion ) , illegitimate recombination , and transposition . Here , we can expand on a previously published model [17] to elucidate the evolution of the P1-wr allele . The duplication of an ancestral p gene gave rise to the p2 and p1 genes that encode functionally interchangeable proteins [42] . However , both genes differ in their regulatory sequences and hence in their tissue-specific expression . Two p copies in direct orientation were subject to unequal crossover , creating a third gene . Additional unequal crossover events amplified the P1-wr genes until the cluster reached today's copy number . In addition , gene conversion took place resulting in homogenized P1-wr copies . Similarly , the p1/p2 hybrid gene was generated by unequal crossover or by gene conversion . A retroelement transposition into the 3′ UTR of the distal p1/p2 gene , followed by additional nested insertions of multiple LTR retrotransposons , pushed the p1/p2 end 68 kb apart ( Figure 7 ) . At the 5′ end of the cluster , a DNA DSB , potentially triggered by a transposon excision , occurred , which resulted in the deletion of the final 2 kb of p2 ( including the third exon ) and 8 kb of the neighboring P1-wr repeat ( or 8 kb plus increments of P1-wr-repeats ) . The DSB was repaired by NHEJ as evidenced by filler-DNA at the junction sequence ( Figure 6 ) . The chronological order of all events in the proposed model can only partially be reconstructed . The two maize progenitors hybridized 4 . 8 mya [30] , leading to p2 on chromosome 1 and p3 on chromosome 9 . The first paralogous copy on chromosome 1 was produced 2 . 75 mya [17] and the earliest transposition 1 . 4 mya in p1/p2 . Obviously , the p1/p2 hybrid gene or at least the p2 end had to be generated between these time points . The p1 and p2 genes had sufficient time to diverge and later on amplify . However , we cannot make any extrapolation on when the modification at the 5′ end occurred .
Our Ufo1 stock ( X03G ) and the inbred lines B73 and 4Co63 were obtained from the Maize Genetics Cooperation Stock Center ( http://maizecoop . cropsci . uiuc . edu/ ) collection . Ufo1 plants carry an undefined P1-wr allele . Therefore , Ufo1 plants were crossed to plants from the 4Co63 inbred line that contain the p1-ww null allele . F1 plants were selfed and F2 plants were selected that are homozygous for p1-ww and display a stunted Ufo1 plant phenotype . These plants were crossed to B73 , and the resulting F1 plants carrying the B73 P1-wr allele and the 4Co63 p1-ww allele were used for our transcript analysis . The inbred line B73 contains a P1-wr allele , according to the colorless pericarp and red cob phenotype of B73 ears . Southern blot hybridizations , using the p1-specific probe 15 [24] , reveal that the P1-wr allele from B73 is composed of a similar repeat structure as P1-wr from the previously studied inbred line W23 ( data not shown ) [14] , [16] . Two publicly available BAC libraries constructed from B73 [53] were screened by hybridizing filters with probes 15 and 8B [24] . Whereas probe 15 is derived from a distal enhancer fragment of P1-rr , which is unique to p1 alleles , probe 8B is obtained from the second intron of P1-rr and detects both p1 and p2 . To minimize the P1-wr repeats per BAC , the strategy has been to pick minimally overlapping BACs that include only a subset of P1-wr repeats . Twenty-one BACs from two genomic libraries were isolated and further characterized . The presence of p2 was evaluated by PCR using the primer pair p2-1 ( p2-1F: ttacgcggcggcaggaaaatcacc , p2-1R: gacgcccaggccgcaggacag ) , which amplifies a 500-bp fragment about 650 bp upstream of the putative transcription start site of p2-t . In addition , insert size and both end sequences of each clone were determined and analyzed . BAC end sequencing was performed with the ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction kit and an ABI 3730 capillary sequencer ( Applied BioSystems ) . BACs c0347H24 , b0232A12 , b0033H05 and b0086K13 were chosen for sequencing based on the above-mentioned criteria ( Figure 2 ) . Notice that the maize FPC maps ( www . genome . arizona . edu/fpc/maize ) were not available at the time of the library screenings . Currently , BAC c0347H24 is included in the HICF FPC map , and BACs b0232A12 and b0086K13 are integrated in the agarose FPC map ( data not shown ) . BAC DNA was isolated using the Large-Construct Kit ( QIAGEN ) . For shotgun library construction , the purified BAC DNA was physically sheared and then ligated into a pUC vector as previously described by [54] . Plasmid inserts were sequenced from both ends using universal primers [55] , ABI 3730 capillary sequencers , and the ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction kit ( Applied BioSystems ) . Base calling and assembly were carried out using phred/phrap programs [56] . About 10× sequence coverage was generated for each BAC . Sequence gaps were closed by primer walking or by transposon minilibraries ( Finnzymes ) , constructed according to the manufacturer's instructions . To assemble the P1-wr containing part of the BACs , the sequence of the entire plasmid insert from which the shotgun sequence originates was determined . However , the insert sequences could not unambiguously be assembled into just one large contig , because the amount of polymorphisms among the repeats is rather small and the average length of a plasmid insert is less than 4 kb , which is significantly shorter than the size of one P1-wr repeat . Therefore , each complete P1-wr repeat from BACs c0347H24 , b0033H05 and b0086K13 was subcloned into the plasmid vector pBluescript II SK ( + ) ( Stratagene ) using the restriction endonucleases EcoRV or XhoI . Both enzymes release a full copy of a P1-wr repeat as they cut only once within a repeat . The unique sites were detected by assembling short ( up to 800 bp ) shotgun sequences derived from all P1-wr repeats . No polymorphisms were detected at their recognition sites . The individual clones were completely sequenced using primers that are spanning the entire repeat length ( approximately one primer every 300 bp , primer sequences available upon request ) . The sequencing reactions were carried out with the ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction kit and analyzed on an ABI 3730 capillary sequencer ( Applied BioSystems ) . The sequences were assembled and evaluated with the Lasergene software ( DNAstar ) . The recognition sites of the endonucleases are about 5 kb apart . The repeat order was established based on polymorphisms within the overlapping fragments . The GenBank accession number for the entire 379-kb sequence is FJ614806 . Total RNA was extracted from pericarp tissue 20 days after pollination with the RNeasy Plant Mini Kit ( Qiagen ) . RNA was reverse-transcribed to cDNA using the SuperScript First-Strand Synthesis System ( Invitrogen ) with oligo ( dT ) or random hexamer primers . cDNA was then PCR-amplified with four primer pairs that flank polymorphic sequences of all P1-wr repeats , p2/p1 and p1/p2: p-hAT-ex2F GCGGGCGGGCTTGGACAGGAAACT , p-hAT-ex2R GGGTGGCGTGGAGCTTGATGATGA , p-ex1-3-1F TAACCGTGCGCAAGTAGTAGTG , p-ex1-3-1R GGCCCGGCGGTGTATTTC , p-ex3-1F CCACCTCCCCGGCCGAACAGACAA , p-ex3-1R GCTCCGGCCCGCCCCACAGATG , p-ex3-3F GGGGGAGGCCGACAGCGAGATG , p-ex3-3R ACCGGCGGGAGAACTACCTTTACA . As a control for presence and ratio of individual P1-wr repeats , genomic DNA corresponding to the RNA sample was PCR-amplified in parallel . 96 PCR and RT-PCR products per primer pair were cloned and also sequenced with universal primers . DNA sequences were analyzed with Lasergene ( DNAstar ) . The maize sequence was manually annotated using homology searches in various GenBank databases with multiple BLAST programs ( BLASTN , BLASTP , BLASTX , TBLASTX ) [57] . By the same approach , existing annotations of the rice [58] and sorghum [26] sequences were manually adjusted where necessary . All sequences were aligned using CLUSTALW ( as implemented in MEGA4 . 0 [59] ) , and the alignment was manually adjusted . Phylogenetic analyses were conducted in MEGA4 . 0 [59] . The phylogenetic tree was inferred by using the Minimum Evolution ( ME ) method . The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test ( 1000 replicates ) was calculated . The evolutionary distances were computed using the Maximum Composite Likelihood method and are in the units of the number of base substitutions per site . The ME tree was searched using the Close-Neighbor-Interchange algorithm at the default search level . The Neighbor-joining algorithm was used to generate the initial tree . All positions containing gaps and missing data were eliminated from the dataset ( Complete deletion option ) . There were a total of 449 positions in the final dataset . Similar trees were obtained by using the same alignments with the Neighbor-Joining ( NJ ) and Maximum Parsimony ( MP ) method . The TSD sequence was used to identify a pair of LTRs that belong to the same LTR-retrotransposon . LTRs were aligned using ClustalX [60] , and the resulting alignment was manually adjusted . Distance estimations between pairs of LTRs were based on Kimura's two-parameter model ( K2P ) as implemented in MEGA4 . 0 [59] . Using the formula T = K/2R , where T is time , K is the number of substitutions and R is the rate of substitutions per site per year , we calculated the time of LTR retrotransposon insertions within the P1-wr cluster ( Figure 7 ) . K was computed using the software MEGA4 . 0 [59] . We applied the substitution rate R of 1 . 3×10−8 mutations per site per year that is based on the average level of nucleotide substitutions in intergenic regions [61] .
|
Plant genomes analyzed to date contain 15% or more genes that are arranged in tandem arrays . Tandem duplications are a source for allelic variability since their homologous sequences can serve in recombination events . For example , unequal crossing over between amplified genes can result in contraction and expansion of the array . Tandem gene multiplications are also subject to repeat induced gene silencing ( RIGS ) . Most importantly , gene duplications create the evolutionary potential for genetic novelty ( neo- or subfunctionalization ) . In addition to homologous recombination during meiosis , illegitimate recombination in somatic tissues of plants can create events that potentially can be transmitted through reproductive tissue to further enrich genetic diversity . Here we illustrate the evolution from a single Myb homolog to a multigene cluster that exemplifies the evolution of the maize genome . We used the p locus to demonstrate how plant genomes expand by polyploidization , gene duplication , and transposition . We characterized in detail the structural changes at the p cluster that resulted from genomic instability . Because structure determines function , we linked genomic rearrangements at the P1-wr cluster to functional consequences . At the P1-wr locus , structural changes caused regulatory/transcriptional modifications that in turn give rise to phenotypic alterations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"evolutionary",
"biology/plant",
"genomes",
"and",
"evolution",
"evolutionary",
"biology/plant",
"genetics",
"and",
"gene",
"expression"
] |
2009
|
Change of Gene Structure and Function by Non-Homologous End-Joining, Homologous Recombination, and Transposition of DNA
|
The membrane scission event that separates nascent enveloped virions from host cell membranes often requires the ESCRT pathway , which can be engaged through the action of peptide motifs , termed late ( L- ) domains , in viral proteins . Viral PTAP and YPDL-like L-domains bind directly to the ESCRT-I and ALIX components of the ESCRT pathway , while PPxY motifs bind Nedd4-like , HECT-domain containing , ubiquitin ligases ( e . g . WWP1 ) . It has been unclear precisely how ubiquitin ligase recruitment ultimately leads to particle release . Here , using a lysine-free viral Gag protein derived from the prototypic foamy virus ( PFV ) , where attachment of ubiquitin to Gag can be controlled , we show that several different HECT domains can replace the WWP1 HECT domain in chimeric ubiquitin ligases and drive budding . Moreover , artificial recruitment of isolated HECT domains to Gag is sufficient to stimulate budding . Conversely , the HECT domain becomes dispensable if the other domains of WWP1 are directly fused to an ESCRT-1 protein . In each case where budding is driven by a HECT domain , its catalytic activity is essential , but Gag ubiquitination is dispensable , suggesting that ubiquitin ligation to trans-acting proteins drives budding . Paradoxically , however , we also demonstrate that direct fusion of a ubiquitin moiety to the C-terminus of PFV Gag can also promote budding , suggesting that ubiquitination of Gag can substitute for ubiquitination of trans-acting proteins . Depletion of Tsg101 and ALIX inhibits budding that is dependent on ubiquitin that is fused to Gag , or ligated to trans-acting proteins through the action of a PPxY motif . These studies underscore the flexibility in the ways that the ESCRT pathway can be engaged , and suggest a model in which the identity of the protein to which ubiquitin is attached is not critical for subsequent recruitment of ubiquitin-binding components of the ESCRT pathway and viral budding to proceed .
The membrane scission event that separates the lipid membrane of nascent enveloped virions from host cell membranes is , in many cases , an orchestrated event requiring the participation of the class E vacuolar protein sorting ( VPS ) , or endosomal sorting complex required for transport ( ESCRT ) pathway . Ordinarily , the ESCRT pathway induces topologically equivalent cellular membrane scission events including the biogenesis of multivesicular bodies ( MVBs ) [1] , [2] and the membrane abscission event at the conclusion of cell division [3] , [4] . Components of the pathway can be recruited , either directly or indirectly , through the action of short peptide motifs called late ( L- ) domains in viral structural proteins [5] , [6] . Three classes of viral L-domains and cognate cofactors have been defined thus far: PT/SAP motifs bind Tsg101 [7] , [8] , [9] , [10] , LxxLF or YPXL motifs bind ALIX [11] , [12] , [13] , and PPxY domains bind Nedd4-like HECT ubiquitin ligases [14] , [15] , [16] , [17] , [18] . Disruption of late domain function results in the failure of membrane scission and the accumulation of assembled virions that remain tethered to the surface of the host cell by a continuous membrane . The ESCRT machinery is composed of ∼25 proteins , many of which participate in the formation of several multiprotein complexes , known as ESCRT-0 , -I , -II , -III [19] , [20] , [21] . ESCRT-III components are thought to drive the membrane scission event [22] , [23] , [24] , [25] and appear to be generally required for L-domain-dependent viral budding [7] , [11] , [12] , [13] , [26] . In contrast , other components of the ESCRT-pathway appear to be required in an L-domain specific way . For example , PTAP-dependent budding is especially sensitive to ESCRT-I perturbation , while YPXL-dependent budding is especially sensitive to ALIX depletion . Since ALIX interacts directly with ESCRT-III via its Bro1 domain [11] , [12] , [13] , [27] , [28] and ESCRT-I indirectly interacts with ESCRT-III via ALIX and/or ESCRT-II , [11] , [12] , [13] these observations suggest that YPXL and PTAP motifs access the same core scission machinery via alternative routes . In contrast , it has remained somewhat unclear how PPxY motifs access the scission machinery . Overexpression of certain HECT ubiquitin ligases that bind directly to PPxY or other motifs can markedly stimulate budding , and the catalytic activity of the HECT domain is essential for this activity [17] , [29] , [30] , [31] . Indeed , overexpression of catalytically inactive or truncation mutants of the HECT ligase WWP1 inhibits PPxY-dependent budding [17] , [29] . Some components of the ESCRT pathway are also required for PPxY-induced budding [7] , [31] , [32] . However , the precise means by which HECT ligase recruitment subsequently results in the engagement of the ESCRT machinery is not completely defined . One model invokes direct ubiquitination of Gag as the key event . This notion derives from observations that several components of the ESCRT pathway are thought to recognize ubiquitinated cargo through various low affinity ubiquitin-binding domains [7] , [33] , [34] , [35] , [36] , [37] and that monoubiquitination of cellular cargos can serve as a signal for endosomal trafficking and delivery to the lysosome [21] , [38] , [39] . Indeed , several observations are consistent with the notion that ubiquitination of retroviral Gag promotes virus particle release . For example , studies have noted an enrichment of free ubiquitin in retrovirus particles , and ubiquitinated Gag species have also been detected therein [14] , [40] , [41] , [42] , [43] . Additionally , late budding defects have been observed in cells treated with proteasome inhibitors , perhaps due to the depletion of free ubiquitin [14] , [44] , [45] . Mutation of multiple ubiquitin acceptor lysine residues in Gag has been shown to inhibit particle production by retroviruses [46] , [47] . Finally , direct fusion of ubiquitin to the C-terminus of Gag proteins has been shown to alleviate inhibition of particle release imposed by proteasome inhibitors , or to obviate the requirement for an L-domain in particle release [44] , [48] . Other observations suggest that PPxY and ubiquitin ligase-dependent budding may involve mechanisms other than direct Gag ubiquitination . In particular , overexpression of wild-type WWP1 stimulates PPxY-dependent particle production by a lysine-free Gag protein [29] in the absence of detectable Gag ubiquitination . This finding suggests the possibility that HECT ligases may promote budding by catalyzing the ubiquitination of specific trans-acting host factors , rather than Gag . Additionally , a HECT-truncated WWP1 protein , lacking the entire HECT domain , inhibits murine leukemia virus ( MLV ) budding more potently than the full length WWP1 protein with a disrupted active site [17] , suggesting that HECT domains may possess activities other than ubiquitin conjugation that are important for their function in viral budding . Moreover , HECT domains localize to aberrant endosomal ( so called class E ) compartments induced by overexpression of catalytically inactive ATPase VPS4 [17] , which is required for the disassembly of ESCRT complexes after each round of budding [49] , [50] . Since many VPS factors are trapped on VPS4-induced compartments , HECT domains may be recruited to these compartments by interaction with VPS proteins , either directly or through unidentified bridging factors . It has also been reported that HECT ubiquitin ligases can bind to , and/or catalyze the ligation of ubiquitin to , certain class E VPS factors [31] , [32] , [51] . Thus , the ubiquitin ligases might act as recruitment factors rather than , or in addition to , conjugating ubiquitin to key target proteins . In this study we investigated the role of PPxY motifs , HECT ubiquitin ligase domains and ubiquitin in viral budding , using a lysine-free viral protein from the prototypic foamy virus ( PFV ) , in which the attachment of ubiquitin to Gag can be rather precisely controlled . We show that the catalytic activity of a variety of HECT domains , targeted to a PPxY motif in assembling particles via a common C2/WW domain fragment of WWP1 , is essential for their ability to promote PPxY-dependent VLP release . In each case , however , Gag ubiquitination is dispensable for their activity . Rather , the ability of the chimeric ubiquitin ligases to promote budding correlated broadly , albeit imperfectly , with their ability to catalyze autoubiquitination , Moreover , we show that artificial recruitment of an isolated HECT domain can also stimulate budding , while a HECT domain becomes dispensable for PPxY motif dependent budding if the C2/WW domains of WWP1 are directly linked to the C-terminal domain of Tsg101 , an ESCRT-I subunit . Finally , we demonstrate that direct fusion of a single ubiquitin moiety to the C-terminus of PFV Gag is also capable of promoting budding , in a manner that recapitulates the ESCRT protein requirement for budding induced by PPxY-dependent ubiquitin ligase recruitment in the absence of ubiquitin acceptors in Gag . These results support a model in which PPxY motif-induced HECT ubiquitin ligase recruitment leads to the deposition of ubiquitin at or near the site of viral budding . However , the identity of the protein to which ubiquitin is attached , be it Gag or a bystander protein , perhaps including the HECT ubiquitin ligase itself , does not appear to be critical in order for subsequent recruitment of ubiquitin-binding class E VPS proteins and viral budding to proceed .
To ascertain what properties of HECT domains are important for stimulation of virus particle release , we compared the properties of a panel of HECT domains . Nine members of the Nedd4-like HECT ubiquitin ligase family have been described in humans and these have the same domain organization as a single prototype member of this family in yeast , namely Rsp5 ( reviewed in [52] ) . Specifically , an N-terminal C2 domain directs the protein to membranes , a central cluster of ‘WW” domains binds ligands , such as PPxY motifs , and a C-terminal HECT domain harbors the E3 ubiquitin ligase activity . Some of the intact ubiquitin ligases have been shown to vary in their ability to promote PPxY-dependent MLV virion release , due at least in part to differences in the affinities of their WW domains for the MLV L-domain [17] , but whether the various the C-terminal HECT domains are equivalently able to induce particle release has not been investigated . We reasoned that variation in the ability of HECT domains to stimulate virus budding , correlated with a given property of the HECT domains , might suggest properties that are important for inducing virion release . Since WWP1 has been previously shown to be efficiently recruited by a number of PPxY-type L-domains , including that of MLV [17] , we constructed a panel of chimeric ubiquitin ligases , consisting of membrane targeting and PPxY motif-binding domains ( C2 and WW domains ) of human WWP1 , coupled to various catalytic HECT domains derived from human WWP2 , Nedd4 , Nedd4L , Itch , Smurf1 , Bul2 or yeast Rsp5 HECT ligases ( Fig . 1A ) . To determine whether these chimeric ubiquitin ligases could support viral budding , we co-expressed each of them with a plasmid expressing a modified PFV Gag protein . Importantly , PFV Gag offers the advantage that it is naturally almost devoid of lysine resides . While PFV Gag normally requires a cognate Env protein for particle release , we have previously shown that appending a myristoylated , palmitoylated peptide from Lck at its N-terminus can overcomes this requirement by directing PFV Gag to the plasma membrane and thereby allowing the generation of extracellular particles in the absence of any other viral protein [29] . Throughout this study we used this N-terminally modified Gag protein , termed Lck-Gag , bearing a K396R mutation that renders the PFV Gag completely lysine-free . Examples of engineered variants of this Gag protein are illustrated in Fig . 1B , and include those that is otherwise unmodified and encode the natural PSAP late domain ( Lck-Gag ( PSAP ) ) , a PSAP mutant that contains no known L-domain ( Lck-Gag ( L- ) ) or another variant that has a PPxY late domain derived from MLV Gag appended to its C-terminus ( Lck-Gag-PY , Fig . 1B ) . In addition , we used an Lck-Gag-PY derivative containing three lysine residues adjacent to a PPxY late domain ( Lck-Gag-PY-3K ) to assess HECT ligase-induced Gag ubiquitination ( [29] , illustrated in Fig . 1B ) . Overexpression of ubiquitin ligases encoding a variety of HECT domains ( WWP1 itself , WWP1/Nedd4 , WWP1/Nedd4L , WWP1/Itch , WWP1/Smurf1 , or WWP1/Bul2 ) stimulated PPxY-dependent budding of lysine-free Lck-Gag-PY ( Fig . 2A , B ) . Conversely , WWP1/WWP2 and WWP1/Rsp5 did not stimulate budding or had marginal activity . The strongest stimulation was observed using chimeric ligases containing the Nedd4L and Itch HECT domains . Importantly , overexpression of chimeric ligases in which the catalytic cysteine was mutated to serine , failed to stimulate PPxY-dependent particle release ( Fig . 2A ) , indicating that the catalytic activity of each HECT domains was required , even when the viral structural proteins lack ubiquitin acceptors . To assess the relative catalytic activities of the chimeric HECT ligases , and assess whether this correlated with their differential ability to promote budding , we compared their abilities to carry out autoubiquitination and to ubiquitinate a Gag substrate encoding three lysine residues in close proximity to a PPxY late domain ( Lck-Gag-PY-3K , see Fig . 1B ) . To accomplish this , we immunoprecipitated either Gag or HECT ubiquitin ligases from 293T cell lysates , prepared 36 hours after co-transfection with plasmids expressing Lck-Gag-PY-3K , HA-tagged ubiquitin , and each of the YFP-fused chimeric HECT ligases . Cell lysates were prepared using denaturing , detergent-rich buffer ( containing 0 . 5% SDS ) to ensure dissolution of protein complexes , and ubiquitinated species were detected by immunoprecipitation with either αPFV Gag or αGFP antibodies followed by immunoblot analysis of the precipitates with an αHA antibody ( Fig . 3 ) . Each of the chimeric HECT ubiquitin ligases was able to reasonably efficiently catalyze the addition of 1 to 3 ubiquitin moieties to the Lck-Gag-PY-3K substrate ( Fig . 3A , upper panels ) . There was some variation in the ability of the HECT domains to catalyze the ligation of ubiquitin to Lck-Gag-PY-3K , with WWP1/Rsp5 and WWP1/Bul2 catalyzing the highest and WWP1/Nedd4 the lowest levels of ubiquitin ligation to Lck-Gag-PY-3K ( Fig . 3A ) . However , there was no correlation between the extent to which each HECT domain stimulated Lck-Gag-PY-3K ubiquitination ( Fig . 3A ) and the degree to which it stimulated the release of VLPs assembled using Lck-Gag-PY or Lck-Gag-PY-3K ( Fig . 2A and data not shown ) . For example , WWP1/Bul2 and WWP1/Nedd4 , which induced the highest and lowest levels of Gag ubiquitination , respectively ( Fig . 3A ) , stimulated budding to a similar extent ( about 6-fold , Fig . 2A ) . Moreover , WWP1/Rsp5 , which efficiently catalyzed Gag ubiquitination ( Fig . 3A ) , enhanced particle release only marginally ( Fig . 2A ) , much less efficiently than the WWP1/Nedd4L that induced comparatively modest levels of Gag ubiquitination ( Fig . 3A ) . We observed a better , albeit imperfect , correlation between the ability of the chimeric HECT ligases to catalyze autoubiquitination and to stimulate VLP production ( Fig . 3B , Fig . 2B ) . Chimeric ligases that strongly promoted Lck-Gag-PY VLP release ( e . g . WWP1/Itch and WWP1/Nedd4L ) were more heavily autoubiquitinated , while those that failed or only marginally promoted VLP release ( WWP1/WWP2 and WWP1/Rsp5 , Fig . 2 ) exhibited the lowest levels of autoubiquitination ( Fig . 3B ) . The correlation was imperfect , however , since WWP1/Nedd4 , which moderately enhanced particle release ( Fig . 2 ) , was consistently highly auto-ubiquitinated ( Fig . 3B ) . Notably , there was no correlation between the ability of the HECT ubiquitin ligases to catalyze autoubiquitination , and their ability to catalyze ubiquitin ligation to Lck-Gag-PY-3K ( Fig . 3A , B ) . Overall , these data confirm our previous finding that direct Gag ubiquitination is dispensable for HECT ligase-dependent budding [29] and further indicates that intrinsic catalytic activity of the HECT ubiquitin ligases is critical for their ability to stimulate budding . We next asked whether the need to recruit a HECT domain in the context of PPxY/WWP1 interaction was necessary for particle release , or whether the HECT domain could be bypassed by direct recruitment of putative downstream effectors . Additionally , we asked whether recruitment of a HECT domain in the absence of the other domains ( C2 and WW ) found in the Nedd4-like family of proteins was sufficient to stimulate particle budding . To accomplish this , we constructed hybrid L-domain cofactors in which the essential domains were split and linked to putatively complementing domains in another L-domain cofactor ( Fig . 4A ) . Specifically , Tsg101 is a core component of ESCRT-I and contains two domains that are functionally important with respect to viral budding . The N-terminal ubiquitin E2 variant ( UEV ) domain interacts directly with P ( T/S ) AP peptide motifs and ubiquitin [7] , while the C-terminal portion of the protein is a key structural component of ESCRT-I , interacting with other components , e . g . VPS28 and VPS37 [53] , [54] , [55] and is essential to support Tsg101 dependent budding . We constructed an artificial putative chimeric L-domain cofactor in which the C2/WW domains of WWP1 were linked to the C-terminal portion of Tsg101 ( Tsg-C ) that constitutes the core structural component of ESCRT-I ( residues 157–390 , Fig . 4A ) . Notably , overexpression of this chimeric protein , termed WWP1-Tsg-C , stimulated Lck-Gag-PY particle release in a dose-dependent manner but had no effect on particle production by the L-domain-deficient Lck-Gag ( L- ) protein ( Fig . 4B , left and middle panels ) . This chimeric protein , therefore , appeared capable of recruiting a functional ESCRT-I complex to PPxY L-domains and thereby stimulating particle production . Conversely , WWP1-Tsg-C overexpression inhibited Lck-Gag ( PSAP ) budding in a dose-dependent manner ( Fig . 4B , right panel ) . We surmise that since this chimeric protein lacks the domains required for interaction with PT/SAP motifs , it acts as an inhibitor of PSAP-dependent budding by sequestering endogenous components ( e . g . VPS28 and VPS37 ) into retargeted ESCRT-I complexes that can be recruited to PPxY , but not PT/SAP , L-domains . Thus , these experiments demonstrate that the requirement for a HECT domain ( and , by inference , the requirement for ubiquitin ligation ) in PPxY/ubiquitin ligase dependent viral budding can be bypassed , if an alternative link to the ESCRT machinery is provided . In a reciprocal experiment , we asked whether the PPxY motif and the C2/WW domains of WWP1 could be functionally replaced in the context of HECT domain/ubiquitin dependent budding . In other words , we determined whether recruitment of a HECT domain is sufficient to stimulate particle release , in the absence of the other protein domains ( C2 and WW ) to which it would ordinarily be linked . Specifically , we attempted to redirect P ( T/S ) AP-dependent particle production through a HECT domain-dependent pathway , by constructing chimeric proteins , termed Tsg-WWP1 , Tsg-Itch and Tsg-Nedd4L , that contained the N-terminal UEV domain ( residues 1–157 ) of Tsg101 linked to one of the three respective HECT domains ( Fig . 4A ) . To test the function of these artificial putative L-domain cofactors , we also constructed an attenuated “leaky” mutant of the PT/SAP motif in the Lck-Gag ( PSAP ) protein , namely Lck-Gag ( ASAP ) , by mutating the first proline residue of the PSAP motif to alanine . In the context of the HIV-1 PTAP motif , such a mutation reduces the affinity for , but does not eliminate binding to the Tsg101 UEV domain [7] . Correspondingly , the budding of Lck-Gag ( ASAP ) , was attenuated as compared to Lck-Gag ( PSAP ) , but the ASAP motif clearly retained some weak residual ability to stimulate budding ( Fig . 4C , leftmost three lanes ) , suggesting that it retains some residual ability to recruit the Tsg101 UEV domain . Overexpression of Tsg-WWP1 , Tsg-Itch or Tsg-Nedd4L , respectively ) resulted in clear stimulation of Lck-Gag ( ASAP ) budding ( Fig . 4C ) . Tsg101-Itch was the most potent of the three Tsg101-HECT proteins tested by this approach , and its overexpression resulted in a particle yield that matched or even exceeded that observed in the presence of the intact PSAP motif ( Fig . 4C ) . In contrast , expression of catalytically inactive versions of Tsg-WWP1 , Tsg-Itch or Tsg-Nedd4L inhibited rather than enhanced Lck-Gag ( ASAP ) particle production ( Fig . 4C ) . Because the Tsg101 UEV domain contains ubiquitin-binding activity that might complicate the interpretation of these results , we repeated these experiments using a mutant Tsg101 UEV domain ( N45A ) that is defective for ubiquitin binding , linked to a WWP1 HECT domain . The mutant Tsg ( N45A ) -WWP1 fusion stimulated budding at least as efficiently as did the unmanipulated Tsg-WWP1 protein ( Fig . 4D ) . Overall , the experiments in Fig . 4 demonstrate that the domains of the PTAP and PPxY binding cofactors can be functionally split into modular , interchangeable domains that are ( i ) necessary for binding to the L-domain and ( ii ) interface with downstream effectors that are critical for budding . Most notably , these findings suggest that simple recruitment of a HECT domain to sites of particle budding , irrespective of its mode of recruitment , and in the absence of ubiquitin acceptors on the viral protein , is sufficient to stimulate particle release and that other HECT ubiquitin ligase domains are dispensable for budding . The aforementioned experiments demonstrated that the requirement for a catalytically active HECT domain could be obviated by direct recruitment of ESCRT-I to a viral protein ( Lck-Gag-PY ) whose budding would normally be dependent on such recruitment . We next asked whether the requirement for HECT domain recruitment could similarly be obviated , in the context of a nearly identical viral protein , by simply depositing ubiquitin at the site of particle assembly , in the absence of ubiquitin ligase recruitment . To mimic the deposition of ubiquitin at sites of virion assembly , in the absence of ubiquitin ligase recruitment , we expressed an Lck-Gag protein , lacking L-domains , with a single ubiquitin appended at its C-terminus ( Lck-Gag-Ub , Fig . 5A ) . Ubiquitin is normally conjugated to proteins by an isopeptide bond between the C-terminal glycine residue of ubiquitin and the ε-amino group of a lysine residue within the substrate protein . Therefore , to avoid aberrant conjugation of our Gag-ubiquitin chimeras to other proteins we deleted two glycine residues from the C-terminus of ubiquitin ( Fig . 5A ) . Cells expressing ubiquitin-fused , but L-domain-deficient Gag ( Lck-Gag-Ub ) generated extracellular particles while those expressing the unfused , L-domain deficient counterpart Lck-Gag ( L- ) protein did not ( Fig . 5B ) . Directly fused ubiquitin-dependent particle release was strongly inhibited , in a dose dependent manner , by expression of a catalytically inactive version of the ATPase VPS4 ( Fig . 5C ) , indicating that the ESCRT pathway was required for Lck-Gag-Ub particle release . Thus , in the context of Lck-Gag , direct ubiquitin fusion appeared capable of substituting for a PSAP or PPxY containing L-domain . These results are similar to findings made by Joshi et al . who showed that direct fusion of ubiquitin to EIAV Gag can functionally substitute for the ALIX-binding YPDL L-domain encoded therein [48] . Similarly , we also found that ubiquitin-dependent budding was dependent on the ubiquitin hydrophobic patch residues ( L8 and I44 ) and additionally , marginally dependent on residues ( Q62 and E64 ) that have been implicated in ubiquitin-Tsg101 UEV domain interaction ( Fig . 5D ) . However , lysine residues ( K48 and K63 ) that are often important for the conjugation of further ubiquitin molecules could be mutated without affecting fused ubiquitin-dependent particle release ( Fig . 5D ) . Next we analyzed the effect of combining L-domains and ubiquitin on VLP release . To accomplish this , Lck-Gag proteins containing various combinations of the L-domains and C-terminally fused ubiquitin ( Fig . 6A ) were expressed . Quantitative analyses revealed that directly fused ubiquitin-dependent ( Lck-Gag-Ub ) particle release was at least as efficient as that driven by PSAP ( Lck-Gag ( PSAP ) ) or PPxY ( Lck-Gag-PY ) L-domains ( Fig . 6B ) . Moreover , and in contrast to the previous report with EIAV Gag [48] , we found that the combined presence of fused ubiquitin and a PSAP L-domain ( in Lck-Gag ( PSAP ) -Ub ) resulted in strongly synergistic effects on particle release ( Fig . 6B ) . Specifically , Lck-Gag ( PSAP ) -Ub generated ∼20-fold and ∼6-fold more particles than Lck-Gag ( PSAP ) and Lck-Gag-Ub , respectively ( Fig . 6B ) . No such synergy was observed when a PPxY L-domain and ubiquitin were combined in the same Gag protein . In fact , the Lck-Gag-Ub and the Lck-Gag-PY-Ub generated extracellular particles with approximately the same efficiency ( Fig . 6B ) . Less dramatic , but nonetheless synergistic enhancement of particle release was evident when PPxY and PSAP motifs were both present ( in the absence of ubiquitin fusion , Fig . 6B ) . In this case , the presence of the PPxY motif ( in Lck-Gag ( PSAP ) -PY ) enhanced particle release approximately ∼5-fold as compared to the situation where the PSAP motif was the only L-domain ( in Lck-Gag ( PSAP ) , Fig . 6B ) . Overall these results are consistent with the notion that ubiquitin behaves essentially like an L-domain , and further suggests that it functions synergistically with a PT/SAP motif , and redundantly with a PPxY motif . We next attempted to mimic a situation that is somewhat typical of retroviruses , where only a fraction of Gag expressed in cells carries ubiquitin . This was done by co-expressing ubiquitin-fused and unfused Lck-Gag proteins in varying proportions . When this was done in the context of a Lck-Gag proteins lacking a PSAP motif ( by co-expressing Lck-Gag ( L- ) and Lck-Gag-Ub ) , particle production was most efficient when a large fraction of the total Lck-Gag protein carried ubiquitin , and no stimulation of particle production was detectable when less than 25% of the Gag protein carried fused ubiquitin ( Fig . 6C , left panel ) . When similar experiments were done in the presence of a PSAP late domain , by co-expressing Lck-Gag ( PSAP ) and Lck-Gag ( PSAP ) -Ub , stimulation of particle release was observed when smaller fractions of Gag , as little as a few percent , carried ubiquitin ( Fig . 6C , right panel ) . Nonetheless , larger fractions of ubiquitin fused Gag had larger stimulating effects on particle release . Thus , these experiments suggest that the greater the number of ubiquitin molecules that are present at sites of particle assembly , the more efficient is particle release; however , relatively modest amounts of ubiquitin can significantly enhance particle budding in the presence of a PSAP motif . Several class E vacuolar protein-sorting factors have been reported to possess ubiquitin binding activity ( Table 1 ) . Although the affinity of such domains for monoubiquitin is generally quite weak ( Kd>100µM ) , several class E factors form multiprotein complexes with several ubiquitin-binding surfaces , which could provide sufficient avidity for their retention at sites of virion assembly . Under such a scenario , efficient recruitment of ESCRT complexes might require deposition of relatively large numbers of ubiquitin molecules in the vicinity of the assembling particle , a notion that is consistent with the finding that a large fraction of Gag must carry ubiquitin to compensate for the absence of a late domain ( Fig . 6C ) . To determine which of the mammalian ESCRT complexes and associated proteins might be most important for ubiquitin dependent budding , we performed a directed yeast two-hybrid screen in which ubiquitin binding to a range of human class E VPS factors and associated proteins was surveyed . These included components of ESCRT-0 ( Hrs , HBP/STAM ) , ESCRT-I ( Tsg101 , VPS28 , VPS37A , B , C , Mvb12 ) , ESCRT-II ( Eap20 , Eap30 , Eap45 ) ESCRT-III ( CHMP1A , 1B , 2A , 2B , 3 , 4A , 4B , 4C , 5 , 6 ) , as well as several ESCRT-associated proteins or proteins that are known to bind to components of the class E VPS pathway ( ALIX , LIP5 , VPS4 , UBPY , CMS , CIN85 ) . Most of these proteins , including known ubiquitin binding factors ( Table 1 ) , gave either weak or non-specific signals . Since we were testing ubiquitin binding by each protein individually and outside of its natural context and in the absence of ESCRT complex partners , it was perhaps to be expected that this assay would fail to detect ubiquitin interactions in at least some instances . Nonetheless , HBP/STAM , ALIX , and UBPY binding gave robust signals in WT ubiquitin binding assays , and binding was abolished when the ubiquitin hydrophobic patch was mutated ( I44A ) , ( Fig . 7A ) . We next determined the effect of siRNA mediated disruption of known ubiquitin-binding complexes , as well the additional ESCRT-associated factors that were positive in our yeast 2-hybrid survey ( ALIX and UBPY ) , on PPxY-dependent and fused ubiquitin-dependent Lck-Gag budding . The core components of the known ubiquitin binding ESCRT complexes ( ESCRT-0 , ESCRT-I and ESCRT-II ) were targeted using pools of four siRNAs directed to Hrs , Tsg101 and Eap45 , respectively . The potency of the siRNA pools was estimated by cotransfecting them with plasmids expressing YFP-tagged target proteins , followed by quantitative western blotting . By these criteria the Tsg101 , Eap45 , ALIX and UBPY siRNAs appeared effective ( Fig . 7B ) . However , knockdown of Hrs was inefficient , so its effect on budding could not be reliably assessed . Because antibodies to Tsg101 and ALIX were available , the level of endogenous proteins could also be monitored in these siRNA experiments . Quantitative western blotting analyses ( examples are shown in Fig . 7C ) indicated that Tsg101 and ALIX proteins were reduced to 38±9% and 16±4% of endogenous levels , respectively . Notably , control experiments showed that Lck-Gag ( PSAP ) particle release was specifically inhibited ( ∼5-fold ) by Tsg101 siRNA , but only marginally affected by EAP45 , ALIX and UBPY depletion ( Fig . 7C , D ) , while EIAV Gag particle release was specifically inhibited ( ∼3-fold ) by ALIX depletion , but not by depletion of the other ESCRT-associated proteins ( Fig . 7C , D ) . Ubiquitin-dependent ( Lck-Gag-Ub ) budding was modestly inhibited ( ∼3-fold ) by depletion of either Tsg101 or ALIX but was barely affected by UBPY or Eap45 siRNAs ( Fig . 7C , D ) , suggesting that ubiquitin binding to ESCRT-I and ALIX contributes to its ability to mediate particle release . This finding mirrors a previous report using ubiquitin fused to EIAV Gag [48] . Additionally , however , we further found that Lck-Gag-PY exhibited a similar pattern sensitivity to class E factor-targeting siRNAs , in that it was modestly sensitive to Tsg101 and ALIX but not Eap45 or UBPY siRNAs ( Fig . 7C , D ) . Similarly , the budding of an MLV Gag protein , that carries the same PPxY L-domain was also modestly sensitive to depletion of Tsg101 and ALIX Fig . 7C , D ) . Because ESCRT-I and ALIX perturbation both affected ubiquitin and PPxY-dependent budding , we sought to determine whether their simultaneous depletion would exhibit a stronger inhibitory effect . Unfortunately , cotransfection of the two pools of siRNAs ( or each pool together with normalizing control RNA duplexes , ) rendered each somewhat less effective , perhaps due to dilution of the active siRNAs ( Fig . 8A ) . Specifically , Tsg101 protein levels were reduced to 42±2% and 50±2% of endogenous levels , while ALIX protein levels were reduced to 30±3% and 27±2% of endogenous levels , when the Tsg101 or ALIX targeted siRNAs were cotransfected together or with normalizing control siRNAs , respectively ( Fig . 8A ) . Thus , under these conditions , siRNAs targeting ALIX did not inhibit Lck-Gag-Ub or Lck-Gag-PY particle release ( Fig . 8A , B ) . Nevertheless , simultaneous ( albeit partial ) depletion of Tsg101 and ALIX had a significantly stronger inhibitory effect on Lck-Gag-Ub , Lck-Gag-PY and MLV Gag budding ( Fig . 8A , B ) than did the more effective individual depletion of either Tsg101 or ALIX alone ( Fig . 7C , D ) , suggesting that they both proteins contribute to optimal PPxY and ubiquitin-dependent budding .
The precise role of HECT ubiquitin ligases in promoting PPxY-dependent virion release has , heretofore , been somewhat unclear . Our previous studies suggest that their ubiquitin ligase activity is critical for their ability to stimulate budding [17] , [29] , but the functionally relevant substrate for ubiquitination has been difficult to define . Additionally , there is some evidence suggesting that HECT ubiquitin ligases may also function as adaptors for bridging factors that recruit ESCRT proteins to assembling virions [17] , [32] , [51] . We compared the activities of HECT domains from various Nedd4-like family HECT ubiquitin ligases by fusing them to the C2 and WW domains of WWP1 . While this strategy does not illuminate which ubiquitin ligases are responsible for viral budding in the natural context , it does allow an assessment of HECT domain function in a uniform background . We found that HECT domains varied significantly in their ability to stimulate PPxY-dependent particle release in this context . This variability was evident when there were no ubiquitin acceptors in the Gag protein and correlated better with the ability of the HECT domains to drive autoubiquitination than with their ability to ubiquitinate a modified Gag substrate that contained lysines proximal to a PPxY motif . The correlation between autoubiquitination and budding was imperfect , however , and it is possible that variation among the HECT domains in their ability to catalyze different lengths and types of ubiquitin chains ( e . g . K48 versus K63-linked chains ) , or their ability to ubiquitinate other bystander proteins , could influence their ability to stimulate viral budding . In this regard it was notable that there was no correlation between the ability of the HECT domains to undergo autoubiquitination versus their ability to catalyze ubiquitin ligation to Lck-Gag-PY-3K . It was nonetheless true that the ability of the HECT domains to stimulate budding was , in every case , absolutely dependent on their ability to catalyze the ligation of ubiquitin to a substrate . This suggests that the proposed role of HECT domains as adaptors that bind directly to downstream factors is of secondary importance in stimulating budding , or that this adaptor function requires catalytic activity . This latter scenario could , conceivably , be operative as a result of HECT autoubiquitination . These studies underscore the remarkable flexibility in the ways that the ESCRT pathway can be engaged to achieve viral budding ( Fig . 9 ) Using a single viral Gag protein as a model , particle budding could be achieved by: ( i ) conventional direct recruitment of the ESCRT pathway via PTAP binding to Tsg101 , ( ii ) direct recruitment of the ESCRT pathway via PPxY binding to a hybrid cofactor consisting of the C2/WW domains of WWP-1 linked to the C-terminal domain of Tsg101 , ( iii ) recruitment of a HECT ubiquitin ligase via a PPxY motif , ( iv ) recruitment of an isolated HECT domain to a PTAP motif using a hybrid L-domain cofactor consisting of the UEV domain of Tsg101 linked to a HECT domain or ( v ) direct fusion of ubiquitin to Gag . These results suggest that the cellular factors ( in this case Tsg101 , ubiquitin ligases and ubiquitin ) that are either directly recruited or deposited at the site of viral particle budding behave as modular entities , with domains that are necessary and sufficient for their own recruitment , and distinct domains that are necessary and sufficient for the subsequent recruitment of downstream effectors of particle release ( Fig . 9 ) . When HECT domains were used to promote budding , the requirement for catalytic activity was absolute , irrespective of how they were recruited to Gag and , importantly , in the absence of ubiquitin acceptors on the viral Gag protein . This finding suggests that ligation of ubiquitin to trans-acting factors , perhaps including the HECT domain itself ( i . e . autoubiquitination ) , rather than Gag is important for viral budding . It is superficially paradoxical , therefore , that ubiquitin could promote budding of the very same Gag protein even when ubiquitin was not ligated to a trans-acting factor , but rather was directly fused to Gag . These findings suggest that the identity of the protein ( s ) to which ubiquitin is attached is not of critical importance , and ubiquitination substrates can , in principle , include Gag , the ubiquitin ligase itself , or other trans-acting proteins . The mere presence of ubiquitin at the site of particle assembly appears sufficient to engage the ESCRT pathway and stimulate budding . The intrinsic manipulability of L-domains , the proteins that bind to them ( specifically ESCRT-I and HECT ubiquitin ligases ) and the apparent lack of importance of the identity of ubiquitination substrate suggests that each serve simply as recruitment factors to engage the downstream machinery that mediates membrane fission and particle release . Since ubiquitin binds to the very same factors ( ESCRT-I and ALIX ) that are bound by PT/SAP and YPXL type L-domains , and depends on them to stimulate budding , then ubiquitin itself can be conceptually viewed , in the context of viral budding , as a transferable L-domain that acts in a position-independent manner . In essence , this notion is a simple extension of the concept originally demonstrated by Parent et al , who showed that conventional L-domains function in a position independent , transferable manner [56] . A finding that is consistent with the aforementioned arguments , is that budding that was dependent either on a PPxY motif or a ubiquitin fused directly to Gag exhibited similar dependence on particular components of the ESCRT pathway . Notably , perturbation of individual segments of the pathway ( ESCRT-I and ALIX ) caused partial inhibition of ubiquitin-dependent Lck-Gag-PY , Lck-Gag-Ub and MLV Gag particle release . Previous work has shown that Mason-Pfizer monkey virus particle release , which is dependent on a PPxY motif , is quite strongly inhibited by depletion of Tsg101 [31] and that budding of a EIAV Gag-ubiquitin fusion protein is modestly inhibited by Tsg101 or ALIX depletion [48] . We found that simultaneous perturbation of ESCRT-I and ALIX resulted a stronger suppression of Lck-Gag-PY , Lck-Gag-Ub and MLV Gag particle release than did depletion of either protein alone , suggesting that both ESCRT-I and ALIX can contribute to optimal PPxY- and ubiquitin-dependent budding ( Fig . 9 ) . Indeed , the class E VPS pathway includes multiple ubiquitin-interacting factors , each of which could , in principle , provide parallel mechanisms for engaging the ESCRT machinery . While ESCRT-I and ALIX appeared to be most important for PPxY- and ubiquitin-dependent budding , these experiments do not exclude a contributory role for other ubiquitin binding complexes in the ESCRT pathway . A similar notion was recently demonstrated in yeast , where simultaneous disruption of ubiquitin binding by ESCRT-I , -II and Bro1 ( the yeast homologue of ALIX ) was necessary to block the sorting of ubiquitinated cargo to the lysosome [57] . Thus , ubiquitin has several potential entry points into the ESCRT pathway , and it appears that multiple interactions must be simultaneously inhibited in order to profoundly inhibit ubiquitin- or HECT ligase-dependent budding . Since ubiquitin-binding class E VPS factors generally have a low affinity for individual ubiquitin molecules ( Table 1 ) , the efficiency with which they are recruited to , and retained at , sites of particle assembly is likely related to the number of ubiquitin molecules that are locally present . Indeed , in the context of direct ubiquitin fusion to Lck-Gag , particle release efficiency increased as the proportion of Gag molecules that carried a ubiquitin was increased , and directly fused ubiquitin could effectively bypass the need for a conventional L-domain only when a large fraction ( >50% ) of the Gag molecules were fused to ubiquitin . This approximates to ∼1000 to 2500 ubiquitin molecules per assembling virion . Previous studies have shown that direct ubiquitin fusion to RSV or EIAV Gag can alleviate a late budding defect imposed by proteasome inhibitors or functionally replace a YPDL L-domain [44] , [48] . However , this study is the first to demonstrate that ubiquitin can act synergistically with a PTAP motif , resulting in dramatically enhanced particle release when both are present . Moreover , the ability of fused ubiquitin to stimulate budding became evident at significantly lower Gag-ubiquitin abundance ( 5% to 25% of total Gag ) when a PTAP motif was also present in Gag . Since ubiquitin could serve as an additional docking site for Tsg101 , it might synergize with PTAP motifs by increasing the overall affinity of the assembling Gag lattice for individual ESCRT-I complexes . In fact , this property was predicted by previous binding studies involving Tsg101 UEV domain , PTAP containing peptides and ubiquitin [7] . Ubiquitin might also synergize with PTAP motifs by providing binding sites for distinct class E VPS factors ( e . g . ALIX ) , thereby optimally utilizing all the available components of the ESCRT machinery . Consistent with these ideas , PTAP and PPxY L-domains behaved synergistically in driving particle release , as did PTAP and Gag-fused ubiquitin . However , a PPxY motif and Gag-fused ubiquitin behaved redundantly , consistent with the notion that that they ultimately function through the same mechanism .
pCAGGS-based expression plasmids encoding Lck-Gag ( PSAP ) , Lck-Gag ( L- ) , Lck-Gag-PY , and Lck-Gag-PY-3K plasmids have been described previously [29] . The Lck-Gag ( ASAP ) plasmid was derived from Lck-Gag ( PSAP ) by PCR-based site-directed mutagenesis . The Lck-Gag ( PSAP ) -PY plasmid was generated by replacement of a StuI/XhoI fragment from the Lck-Gag ( PSAP ) plasmid with the corresponding fragment from the Lck-Gag-PY plasmid . cDNAs expressing Lck-Gag-Ub ( ubiquitinΔGG ) fusion proteins were generated by overlap-extension PCR , using pCAGGs-Lck-Gag ( PSAP ) , Lck-Gag ( L- ) , and Lck-Gag ( PSAP ) -PY as templates for the N-terminal portions and pHA-ubiquitin as the template for the C-terminal portion . The K48R , K63R , F4A , L8A , I44A , and QE62 , 64AA point mutations were introduced into the Lck-Gag-Ub construct by PCR-based mutagenesis . Each cDNA was cloned into pCAGGs for expression in mammalian cells . DNAs encoding the HECT domains from WWP1 ( residues 543–922 ) , WWP2 ( 491–870 ) , Nedd4 ( 520–902 ) , Nedd4L ( 593–975 ) , Itch ( 483–862 ) , Smurf1 ( 374–757 ) , and Rsp5 ( 431–809 ) were amplified from plasmids encoding the full-length HECT ligases [17] , [58] . The Bul2 HECT domain ( encoding residues 1189–1572 ) was PCR amplified from 293T cell cDNA . The catalytically inactive WWP1 HECT domain ( C890S ) was amplified from a previously described full-length mutant WWP1 ligase [17] . Catalytic point mutants of the remaining HECT domains were made by PCR-based mutagenesis . Chimeric ubiquitin ligases , comprising the C2 and WW domains ( residues 1–542 ) of WWP1 and each of the HECT domains described above were generated by overlap PCR . Likewise , plasmids expressing Tsg-WWP1 , Tsg-Nedd4L and Tsg-Itch ( residues 1–157 of Tsg101 fused to HECT domains of WWP1 , Nedd4L , or Itch ) as well as WWP1-Tsg-C ( residues 1–542 of WWP1 fused to residues 157–390 of Tsg101 ) were constructed by overlap-extension PCR . All cDNAs encoding chimeric proteins were inserted into pCR3 . 1/YFP , to express proteins fused to the C-terminus of YFP , for in mammalian cells . The class E VPS factor yeast two-hybrid library and plasmids expressing Vps4 E228Q , Tsg101 , Hrs , ALIX , UBPY , and Eap45 fluorescent fusion proteins in mammalian cells have been described previously [8] , [12] , [54] . Yeast two-hybrid plasmids encoding wild type and I44A mutant ubiquitin were constructed by PCR amplification of ubiquitinΔGG from the pHA-ubiquitin plasmid using 5′ and 3′ primers appended with EcoRI restriction sites and cloning into the pGBKT7 ( Clontech ) and pVP16 vectors [8] . For Gag particle release assays , 5×105 293T cells in six-well plates were transfected using polyethylenimine ( Polysciences ) with 1 µg of pCAGGs/Gag-derived plasmids , alone or with 1 µg of pCR3 . 1/YFP , pCR3 . 1/YFP-WWP1/HECT , pCR3 . 1/YFP-Tsg-HECT , or pCR3 . 1/YFP-C2-WW-Tsg-C plasmids , or the indicated amounts of pCR3 . 1/YFP-Vps4 E228Q plasmid . For EIAV and MLV VLP release assays , 293T cells were transfected in the same format with 500ng of , pCR3 . 1/EIAVGag or pCR3 . 1/MLVGag-HA plasmids . VLPs were pelleted by ultracentrifugation of 2 ml of 0 . 22-µm-filtered culture supernatants , collected 48 hours after transfection , over a 2ml 20% sucrose cushion for 90 min at 115 , 000×g . VLP and cell lysates were analyzed by Western blotting . 293T cells ( 5×105 ) in six-well plates were cotransfected with 1 µg of pCAGGs/Lck-Gag-PY-3K , 500 ng of pHA-ubiquitin , and 1 µg of the indicated chimeric pCR3 . 1-WWP1-HECT ligase . At 36h after transfection , cells were thoroughly lysed at room temperature in detergent-rich RIPA buffer ( 50mM Tris pH 7 . 4 , 150mM NaCl , 1mM EDTA , 1 . 0% glycerol , 0 . 5% SDS , supplemented with protease inhibitor tablets ( Roche ) and 5mM N-ethylmaleimide to inhibit deubiquitination ) and cleared of cellular debris by microcentrifugation . The lysates were then diluted 5-fold in the same buffer containing NP-40 rather than SDS , to adjust the concentration of SDS to 0 . 1% and NP-40 to 1 . 0% , and split into two fractions . From one fraction , Gag proteins were immunoprecipitated with αPFV serum and protein G-Sepharose beads . From the other fraction , YFP-HECT ligase proteins were immunoprecipitated with αGFP monoclonal antibody and protein G-Sepharose beads . Immunoprecipitates and unfractionated cell lysates were analyzed by Western blotting . 293T cells ( 3×105 ) in six-well plates were transfected with siGENOME siRNAs targeting Luciferase , Tsg101 , Hrs , Alix , UBPY , or Eap45 ( Dharmacon ) using Lipofectamine 2000 ( Invitrogen ) . After 24h , cells were transfected with the same siRNAs and the indicated Gag expression plasmids . VLP and cell lysates were prepared 48 h after the second transfection . To assess knockdown efficiency , 293T cells were transfected once with YFP-Tsg101 , -Hrs , -ALIX , -UBPY , or -Eap45 expression plasmids and corresponding siRNAs . GFP expression in cell lysates harvested 48 h after transfection was assayed by quantitative Western blotting . Virion and cell lysates and immunoprecipitates were separated on polyacrylamide gels , transferred to nitrocellulose membranes , and probed with various antibodies: anti-PFV human serum , anti-HIV-1 p24CA ( 183-H12-5C ) , anti-EIAV equine serum ( VMRD , Inc . ) , anti-GFP ( Roche ) , and anti-HA ( HA . 11 , Covance ) anti-Tsg101 ( 4A10 , Abcam , Cambridge , MA ) or anti-ALIX rabbit serum ( a gift from Wes Sundquist ) . Subsequently , the blots were probed with species-specific peroxidase-conjugated secondary antibodies and chemiluminescent substrate reagents . Alternatively , for quantitative Western blotting , membranes were probed with species-specific antibodies conjugated to IRDye800CW , and fluorescent signals were detected and quantified using a LICOR Odyssey scanner . Yeast cells ( Y190 ) were transformed with the pGBKT7- and pVP16-derived plasmids described above . Transformants were selected and protein-protein interactions were assayed by β-galactosidase reporter activity as previously described [8] .
|
The release of an enveloped virus particle from an infected cell requires the separation of the viral and cell membranes . Many enveloped viruses accomplish this by parasitizing a set of cellular proteins , termed the ESCRT pathway , that normally separates cellular membranes from each other . In some cases , viral structural proteins encode peptides motifs that bind directly to , and thereby recruit , the ESCRT machinery . Alternatively , viruses can recruit enzymes , termed ubiquitin ligases , that bind to other proteins , and catalyze the addition of ubiquitin to them . It has , heretofore , been somewhat unclear precisely how the recruitment of ubiquitin ligases leads to the engagement of the ESCRT machinery . We show that the simple recruitment of a fragment of a ubiquitin ligase that is responsible for the addition of ubiquitin to other proteins is sufficient to drive virus particle release , even when it is not possible to attach ubiquitin to viral proteins . Paradoxically , we also found that simple attachment of ubiquitin to the same viral protein can also drive particle release . These results show that there is flexibility in the ways in which the ESCRT machinery can be recruited and how ubiquitin can be co-opted to enable this .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"cell",
"biology/membranes",
"and",
"sorting",
"virology/virion",
"structure,",
"assembly,",
"and",
"egress"
] |
2010
|
Functional Interchangeability of Late Domains, Late Domain Cofactors and Ubiquitin in Viral Budding
|
Recent studies have demonstrated that the DNA methylome changes with age . This epigenetic drift may have deep implications for cellular differentiation and disease development . However , it remains unclear how much of this drift is functional or caused by underlying changes in cell subtype composition . Moreover , no study has yet comprehensively explored epigenetic drift at different genomic length scales and in relation to regulatory elements . Here we conduct an in-depth analysis of epigenetic drift in blood tissue . We demonstrate that most of the age-associated drift is independent of the increase in the granulocyte to lymphocyte ratio that accompanies aging and that enrichment of age-hypermethylated CpG islands increases upon adjustment for cellular composition . We further find that drift has only a minimal impact on in-cis gene expression , acting primarily to stabilize pre-existing baseline expression levels . By studying epigenetic drift at different genomic length scales , we demonstrate the existence of mega-base scale age-associated hypomethylated blocks , covering approximately 14% of the human genome , and which exhibit preferential hypomethylation in age-matched cancer tissue . Importantly , we demonstrate the feasibility of integrating Illumina 450k DNA methylation with ENCODE data to identify transcription factors with key roles in cellular development and aging . Specifically , we identify REST and regulatory factors of the histone methyltransferase MLL complex , whose function may be disrupted in aging . In summary , most of the epigenetic drift seen in blood is independent of changes in blood cell type composition , and exhibits patterns at different genomic length scales reminiscent of those seen in cancer . Integration of Illumina 450k with appropriate ENCODE data may represent a fruitful approach to identify transcription factors with key roles in aging and disease .
Recent studies , using Illumina Infinium beadarrays , have demonstrated that genome-wide DNA methylation patterns change with age [1–6] . Further studies have indicated that this age-associated epigenetic drift may have deep implications for stem-cell biology [7] , disease development [8] and possibly also human evolution [9 , 10] . Thus , it has become of great interest and importance to study the detailed dynamics of the DNA methylation landscape in response to aging . In this regard however there are many pressing unanswered questions . First is the issue of cellular heterogeneity . Although a number of studies have already indicated that a significant component of the epigenetic drift is tissue-independent , and therefore unlikely to be caused by underlying changes in cell subtype composition [2 , 3 , 6] , this result remains unproven and contrasts with a number of Epigenome-Wide Association Studies ( EWAS ) for specific diseases , notably Rheumatoid Arthritis [11] and cancer [12 , 13] , which have shown that in the disease context , correction of intra-sample cellular heterogeneity can be critical [11 , 14] . Indeed , a number of statistical methods have emerged allowing correction for cell subtype compositional changes [15 , 16] . However , to date no study has applied these algorithms in the context of ageing to assess how much of the epigenetic drift is due to underlying changes in cell-type composition . This is particularly pertinent in the context of blood tissue , since in this tissue it is known that there is an age-associated increase in the granulocyte to lymphocyte ratio [12 , 15] . Another outstanding issue concerns the characteristic genomic length-scales of epigenetic drift . Cancer epigenome studies have demonstrated that cancer-associated DNA methylation changes exhibit differential patterns at different length scales , with local hypermethylation at CpG islands ( CGIs ) often immersed within large-megabase scale blocks of widespread hypomethylation [17 , 18] . However , so far no study has comprehensively explored whether hypomethylated blocks emerge in normal tissue as a function of age , and whether these bear any resemblance to those seen in cancer or early neoplastic lesions . Given that age is a major risk factor for many diseases , specially cancer , the existence of such blocks could provide an important indicator of future disease development . Finally , epigenetic drift is thought to be one mechanism underlying the decline of stem-cell function with age , thus compromising normal homeostasis [7 , 9] . That this might be the case is further supported by a recent study of dynamic DNA methylation changes during cellular development [19] , which showed that differentially methylated regions ( DMRs ) in development are strongly enriched for regulatory elements . Indeed , differential binding of transcription factors as a result of differential DNA methylation at transcription factor binding sites ( TFBSs ) may be an important mechanism of cellular development and lineage specification . It follows that age-associated epigenetic drift may compromise binding of key lineage-specifying transcription factors . Although there is already considerable evidence that developmental transcription factors represent targets of epigenetic drift [2 , 3] , no study has yet explored in detail which transcription factor binding profiles may be disrupted in aging as a result of drift . More generally , the ability to detect putative differential binding of transcription factors by studying DNA methylation patterns around their TFBSs is specially important because , unlike ChIP-Seq , DNA methylation can be reliably measured genome-wide from limited amounts of DNA [20] . Thus , DNA methylation can be measured in large numbers of clinical or non-clinical specimens , allowing in principle transcription factors disrupted in disease and ageing to be identified [21] . Moreover , differential binding caused by differential DNA methylation may represent a more accurate way of assessing differential activity of transcription factors . For instance , assessing transcription factor activity from its mRNA or protein expression level is problematic due to postranslational modifications [22] . Thus , DNA methylation may provide improved or complementary information about transcription factor activity . In this work we perform a comprehensive study of DNA methylation drift addressing the core issues mentioned above . We focus on blood because of its availability and the fact that the largest studies to date have been performed in this tissue [4 , 5] . By using a powerful algorithm to correct for cellular heterogeneity [15] , we demonstrate that most of the epigenetic drift , specially the age-hypermethylated component of it , is not caused by changes in blood cell subtype composition . We further demonstrate that age-associated DNA methylation patterns exhibit spatial patterns at different genomic length scales which are reminiscent of those seen in cancer . Moreover , by integrating large-scale 450k DNA methylation data with extensive TFBS information from the ENCODE project [23–25] , we demonstrate the feasibility of the Illumina 450k technology to identify important lineage-specific transcription factors . Having demonstrated the feasibility of this approach in the context of cellular development , we next apply the same method to aging , identifying novel transcription factors which may be implicated in the aging process .
Age-associated DNA methylation changes have so far only been studied comprehensively at the level of individual CpGs [1–6] , and in whole blood tissue , a tissue for which the largest sample collections are available [4 , 5] . However , the interpretation of age-associated changes in whole blood is problematic due to underlying age-associated changes in blood cell subtype proportions [12 , 15] . To address these challenges , recent studies have developed methods that allow more robust inference of differential DNA methylation at the level of genomic regions [26 , 27] and which is independent of underlying changes in cell subtype composition [15 , 16] . Here we decided to use one of these recent methods [15] , to re-analyze one of the largest Illumina 450k DNA methylation data sets available , encompassing whole blood samples from 656 individuals spanning a wide age range ( 19 to 101 year olds ) [5] . In order to call differential methylation more robustly and to avoid any statistical biases caused by neighboring probes on the beadarray , we decided to collapse neighboring probes in the genome into specific clusters [26] , dividing all probes up into 3 different regional classes: CpG islands , shelves & shores and open sea . Specifically , probes within a class and which were spatially close were grouped into regions with an upper bound of 1 . 5Kb set on the size of these regions ( Methods ) . This resulted in 239650 regional clusters , consisting of 109424 open sea , 90090 shelve/shore , and 40136 CpG island regions . Probe methylation values within regions were averaged . Supervised linear regression analysis was performed for each of the 239650 regions , adjusting for plate and gender ( plate was fully correlated with ethnicity in such a way that adjusting for plate also adjusts for ethnicity ( Methods ) ) , which identified 124352 age-DMRs at an FDR < 0 . 05 . Focusing on the top 5% age-DMRs ( a total of 11982 ) , the far majority ( 75% ) were hypomethylated with age , consistent with previous observations [4 , 5] . Although the majority of the top 5% DMRs fell into open sea and shore/shelf regions , CpG-islands were more strongly enriched given their overall lower numbers in the genome ( Fig . 1A ) . We observed a marked difference between open sea or shore/shelf regions and CpG islands , with approximately 90% of open sea/shore/shelf age-DMRs exhibiting hypomethylation , in contrast to age-DMRs mapping to CpG islands , which were overwhelmingly hypermethylated ( 87% ) with age ( Fig . 1B ) . We verified that all these results remained unchanged had we used a less stringent Bonferroni correction threshold ( 26019 age-DMRs ) ( S1_Fig in S1 Text ) . Age-DMRs mapping to CpG islands predominantly exhibited low levels of DNA methylation in the youngest individuals of our cohort ( age range 19–28 ) , whereas CpG-islands whose DNAm did not change with age exhibited both low as well as high levels of DNAm ( Fig . 1C ) . In order to assess the impact of changes in blood cell subtype composition , we applied the reference-based Houseman algorithm [15] to estimate the relative proportions of 6 blood cell subtypes ( CD4+ & CD8+ T-cells , NK-cells , B-cells , monocytes and granulocytes ) in the 656 whole blood samples . The algorithm predicted an age-associated decline in the relative numbers of T and B-cell lymphocytes , whilst the proportion of granulocytes and monocytes increased ( S2_Fig in S1 Text ) , in line with previous observations [7 , 12] . By using these sample-specific cell proportion estimates as covariates in the linear regressions , we rederived an adjusted set of age-DMRs , which resulted in 85299 regions at FDR < 0 . 05 , i . e 69% of the total number identified without adjustment . Focusing on the top 5% ( 11982 regions ) age-DMRs from the adjusted analysis , we observed that the enrichment of CpG-islands among the adjusted age-DMRs was increased relative to the unadjusted analysis and relative to open-sea and shore/shelf regions ( Fig . 1D–E ) . Among the top 5% age-DMRs from the adjusted analysis , 63% and 37% were hypo and hypermethylated , respectively , i . e . less hypomethylation was observed after adjustment for cell-type composition , consistent with previous data [7 , 12] . Importantly , we observed a strong correlation between the unadjusted and adjusted analysis , with approximately 50% of the unadjusted age-DMRs ( FDR < 0 . 05 ) retaining significance at the same FDR level in the adjusted analysis ( Fig . 1F ) . Using a more stringent threshold , i . e declaring only the top 5% of regions as age-DMRs in the unadjusted analysis , resulted in 83% of these passing an FDR < 0 . 05 in the adjusted analysis . Thus , these results indicate that although less hypomethylation is observed after adjustment , that most of the changes , and in particular those involving hypermethylation , are independent of blood cell subtype , in agreement with previous observations made with the older Illumina 27k technology [2 , 3 , 28] . It is of interest to study epigenetic drift on larger genomic length scales , since recent studies have demonstrated the existence of large mega-base scale blocks of hypomethylation in disease [17 , 18 , 26] . We sought to determine if such hypomethylated blocks are also seen in normal tissue as a function of age . To this end , we applied the same block-finding algorithm of Aryee et al [26] . We identified a total of 351 age-associated blocks ( FWER < 0 . 05 ) , with an overall genome coverage of 433Mb ( i . e . 14% of the human genome ) and with a median block size of 1 . 2Mb ( Table S1 in S1 Text ) . The overwhelming majority ( 98% ) of these blocks exhibited hypomethylation ( Table S1 in S1 Text , Fig . 2A ) . Next , we asked if age-hyperM DMRs ( FDR < 0 . 001 ) mapping to CGIs were enriched within these blocks . Most blocks ( 309/351 , 88% ) were either not enriched for age-hyperM CGI-DMRs or the numbers of age-hyperM CGIs were too small to reliable assess statistical significance ( Table S1 in S1 Text ) . Of the 42 blocks which were enriched ( Binomial P < 0 . 05 ) , 18 showed strong enrichment encompassing at least 10 age-hyperM CGI-DMRs ( Table 1 ) . We note that because blocks are not directly comparable , adjustment for multiple testing in this context is prone to substantial error . However , using Benjamini-Hochberg , 21 of these 42 blocks remained significant at an FDR < 0 . 05 , and all 42 were significant at a more relaxed adjusted threshold of FDR < 0 . 17 . Importantly , the bi-modality of enrichment of age-associated hypomethylated blocks indicates that the mechanisms leading to DNA methylation deregulation may be distinct for different genomic blocks . Many of the hypomethylated blocks contained many well separated CGIs exhibiting hypermethylation and targeting multiple genes ( Fig . 2B ) . Among these was a block on chromosome-19 containing multiple genes encoding for zinc-finger proteins , as well as a block on chromosome-13 containing MYCBP2 and POU4F1 . However , we also observed hypomethylated blocks where the enrichment was driven by hypermethylated CGIs which were all in close proximity to each other , for example this was the case for a block on chromosome-4 with all CGIs in the neighborhood of the HAND2 gene ( S3_Fig in S1 Text ) , a gene which has already been linked to aging [1] and which has also been causally implicated in endometrial carcinogenesis [29] . Thus , our analysis suggests that only a relatively small fraction of hypomethylated blocks are enriched for age-hypermethylated CGIs , with a few of these blocks representing hotspots of CGI hypermethylation . Next , we asked if age-associated hypomethylated blocks overlap significantly with those seen in cancer . To investigate this , we mapped the age-associated hypomethylated blocks onto cancer DNA methylation data from the TCGA [30] , comparing average DNA methylation levels of open sea probes within blocks between normal and age-matched cancer tissue . Blocks showed significantly lower DNAm levels in cancer tissue compared to age-matched normal samples , independently of tissue type ( Fig . 3 ) . Next , we randomly picked open sea blocks which did not show significant age-associated hypomethylation in blood and recomputed statistics of differential methylation for these blocks . Comparing the distribution of the statistics of differential DNAm for the age-associated blocks to these randomly picked blocks , revealed significantly larger negative statistics ( Kolmogorov-Smirnov P < 1e − 10 ) for the age-associated hypoM blocks ( Fig . 3 ) . Thus , this extends previous observations made at the level of CGIs to the mega-base scale block level , further supporting the view that much of the cancer-associated deregulation of DNA methylation may already be present in aged normal tissue [31] . Next , we decided to shed further light on the potential functional effect of epigenetic drift . Although a recent study , which performed matched Illumina 450k and gene expression profiling for the same blood samples , concluded that most drift does not cause gene expression changes , this study was significantly underpowered [32] . Thus , given that another unmatched study did report a weak association between age-DMRs and gene expression [5] , the functional significance of epigenetic drift remains unclear . To address this question , we analysed one of the largest whole blood gene expression datasets available , encompassing over 200 samples [33] ( Methods ) . We focused on DNAm levels in CpG island probe clusters with probes mapping to within 200bp of the transcription start site ( TSS ) or 1st exon , since it was shown previously that Illumina 450k probes mapping to these regions provide the best predictive power of a corresponding gene’s expression level [34] . However , we did not find any global statistical significance between the statistics of differential DNA methylation and those of differential expression ( S4_Fig in S1 Text ) . To understand why , we posited that age hyperM CGI-DMRs may preferentially target genes which are normally not expressed in blood , whilst age-hypoM CGI-DMRs may correspondingly target more highly expressed genes . To test this hypothesis it is important to first obtain estimates of the baseline levels of expression in blood , ideally at birth . To this end , we obtained an expression data set of cord blood and placenta samples [35] . Confirming our hypothesis , we observed a highly significant trend with genes undergoing age-hypermethylation around their TSSs exhibiting significantly lower levels of expression than genes exhibiting age-associated hypomethylation ( Fig . 4A ) . Thus , age-associated modulation of DNA methylation is unlikely to cause widespread in-cis expression alterations , because the direction of DNAm change may only act to stabilize pre-existing expression levels . To validate this result , we decided to repeat this analysis in the de Jong et al cohort [33] , separately on young and old age groups . Remarkably , we observed a similar trend as in the cord blood and placenta data ( Fig . 4B ) . Importantly , we can also observe no difference in the expression levels of age-hyperM or age-hypoM genes between the younger and older individuals ( Fig . 4B ) . In order to validate these results further , we collected another relatively large gene expression data set of 198 whole blood samples , albeit this cohort consisted of significantly older individuals [36] . Confirming our earlier result , we did not find that genes mapping to age-hyperM DMRs exhibited age-associated decreases in gene expression , and similarly that genes mapping to age-hypoM DMRs did not exhibit gene expression increases ( S4_Fig in S1 Text ) . Importantly , in this cohort we also observed a significantly higher level of expression of genes undergoing age-associated hypoM in their promoters compared to those undergoing age-associated hyperM , a result which was also independent of age-group ( Fig . 4C ) . A recent study used whole-genome bisulfite sequencing ( WGBS ) to demonstrate that differential DNA methylation can be a powerful means of identifying regulatory elements , including transcription factors , which play key roles in cellular development [19] . Similarly , we posited that differential DNA methylation changes associated with age , if enriched for transcription factor binding sites ( TFBS ) , may allows us to identify transcription factors whose differential binding and activity patterns become disrupted with age . In order to assess the potential of the Illumina 450k platform to identify relevant transcription factors through differential DNAm patterns , we first considered the case of cellular development and lineage specification , where key transcription factors are already known . Specifically , we collected an Illumina 450k DNA methylation data set of 153 samples , encompassing human embryonic stem cell ( hESCs ) and induced pluripotent stem cell ( iPSCs ) lines , as well as somatic ( differentiated ) tissue specimens ( Methods ) [37] . We derived DMRs between the hESCs and the differentiated cell types from a total of 244347 regional clusters ( Fig . 5A ) . Focusing on the top 5% of DMRs ( all passed FDR < 0 . 001 ) , we observed that the majority were hypomethylated in differentiated cells , with open sea regions generally exhibiting hypomethylation , in contrast to shore/shelf and CGIs which were mostly hypermethylated ( Fig . 5B ) . To assess enrichment of TFBS among the DMRs , we first mapped 450k probes onto binding sites of 58 transcription factors ( TFs ) , as assessed by the ENCODE consortium [23] in the H1 hESC line ( Methods ) . The average fraction of 450k probe CpGs mapping within a binding site of one of the 58 TFs was 3% ( ∼ 16316 probes ) , with Pol2 exhibiting the largest overlap ( 12% ) and BCL11A the lowest ( < 0 . 01% i . e . less than 454 probes ) ( S5_Fig in S1 Text ) . Enrichment analysis was then performed for each one of the 58 ENCODE transcription factors , and separately , on the hypermethylated and hypomethylated DMRs ( Table S2 in S1 Text ) . Confirming the results of Ziller et al [19] , we observed that DMRs which exhibited lower levels of methylation in hESCs were massively enriched for binding sites of well-known pluripotency factors such as POU5F1 and NANOG ( Fisher-test P < 10−10 , Fig . 5C ) . Out of the 58 TFs considered , besides OCT4 and NANOG , only 7 others ( BCL11A , HDAC2 , SP1 , MAFK , c-JUN , CtBP2 and RXRA ) showed specific enrichment ( all significant with Fisher-test P < 10−5 and with Benjamini-Hochberg ( BH ) adjusted P < 0 . 05 ) among DMRs hypomethylated in hESCs . A total of 5 TFs ( TCF12 , p300 , TEAD4 , ATF2 , JUN-D ) showed bivalent enrichment in both hypermethylated and hypomethylated DMRs . The rest of TFs showed specific enrichment among the DMRs hypomethylated in differentiated cells ( Fig . 5C ) . Notably , CTCF was the most strongly enriched TF among DMRs losing methylation in differentiated cells . It is also noteworthy that components of the polycomb complex ( e . g . SUZ12 or EZH2 ) were not enriched in either the hypermethylated or hypomethylated DMRs ( Fig . 5C ) , suggesting that polycomb binding does not play a major role in cellular differentiation . Confirming the robustness of the results and reliability of the ENCODE data , we observed that ChIP-Seq binding profiles of the same TF but generated by different laboratories had very similar enrichment values ( Fig . 5C ) . Among the 60 somatic tissue samples , some tissues were represented in sufficient numbers to also allow for tissue-specific analyses . To see whether differential DNA methylation would allow us to infer lineage-specific transcription factors , we considered the case of liver-tissue [19] . Because ChIP-Seq for a reasonable number of TFs ( more than 50 ) has only been performed for a liver cancer cell line ( HepG2 ) , we used the binding site profiles as determined in this cell-line . We identified a total of 1547 DMRs hypomethylated in hESCs compared to normal liver samples ( n = 4 ) , compared to as many as 10670 DMRs which were hypomethylated in the normal liver cells . TFBS enrichment analysis revealed massive enrichment of FOXA1 , FOXA2 , CEBPB , HNF4A and HNF4G binding sites ( Fisher test P < 10−10 , BH-adjusted P < 0 . 05 ) among the liver-specific hypomethylated regions ( Fig . 5D & Table S3 in S1 Text ) , once again consistent with previous observations [19] and with the previously documented role of these TFs in liver specification [38 , 39] . Having validated our algorithm in the context of cellular development , we next asked if specific TFBSs may be enriched among age-associated DMRs . Using ENCODE TFBSs as assessed in the H1-hESC line , we found that even though only the minority ( 3036 DMRs , 25% ) of age-DMRs were hypermethylated , that these were more strongly enriched ( Table S2 in S1 Text ) . The transcription factors most strongly enriched among age-hypermethylated DMRs were two polycomb components ( EZH2 , SUZ12 ) , as well as RBBP5 and NRSF/REST ( Fig . 6A ) . Far fewer TFs were enriched among age-hypomethylated DMRs , but most of these also showed enrichment of binding sites among age-hypermethylated DMRs ( e . g . RAD21 , ZNF143 , CTCF ) ( Fig . 6A , Table S4 in S1 Text ) . We note that most of the enrichments were highly significant ( Fisher-test P < 10−10 , BH-adjusted P < 0 . 05 , Fig . 6A , Table S4 in S1 Text ) . Because many transcription factors co-bind at specific sites , we next performed a multivariate regression analysis to assess if the association of a given TF binding profile with age-associated DNA methylation is independent of co-binding by other TFs ( Methods ) . We note that no pair of distinct TFs exhibited a Jaccard Coefficient overlap higher than 0 . 7 , thus allowing us to use all distinct TF binding profiles in the multivariate analysis , which largely confirmed the previous enrichment analysis ( Fig . 6B ) . Specifically , the polycomb factors ( EZH2 , SUZ12 ) were once again highly enriched among age-hyperM DMRs , as well as RBBP5 , TAF1 , NRSF/REST , CTBP2 and EGR1 . TFs whose binding sites were enriched among age-hypoM DMRs included as before RAD21 and ZNF143 , but now also ATF2 , CEBPB and JARID1A . We verified that results were largely unchanged had we used the age-DMRs from the cellular heterogeneity adjusted analysis ( SI , S6_Fig in S1 Text ) . Next , we asked if particular TFs have binding sites enriched within the previously identified age-associated hypomethylated blocks . Not unsurprisingly , this revealed TFs whose binding sites we previously found to be enriched within age-hypoM DMRs ( e . g . RAD21 , ZNF143 , CTCF , S7_Fig in S1 Text ) . However , it also identified a number of TFs whose binding sites were enriched only among age-hyperM DMRs , specifically this was the case for SIN3A , TBP and TAF1 . Interestingly , these specific TFs also had substantially more binding sites within open-sea regions , in comparison to say the polycomb factors ( EZH2 , SUZ12 ) which exhibited the least enrichment of binding sites within blocks ( S7_Fig in S1 Text ) . In summary , our integrative DNA methylation ENCODE analysis not only points towards an age-associated disruption of PRC2 binding , but also of that of important transcription factors like REST , which has recently been strongly implicated in Alzheimer’s [40] , and RBBP5 , which interacts with members of the histone methyltransferase MLL complex . In addition , the data points towards a potential binding site redistribution of other transcription factors like RAD21 , ZNF143 and the chromatin organization factor CTCF . Of note , pluripotency factors , which were strongly enriched in the previous cellular development/differentiation analysis , were not so in the aging analysis ( Tables S2 & S4 in S1 Text & Fig . 6 ) .
Given the emerging importance of age-associated epigenetic drift , we decided to conduct an in-depth novel integrative multi-scale analysis of this epigenetic phenomenon in whole blood tissue , the tissue for which the largest data set is available . Our analysis makes a number of important novel observations , whilst also confirming some earlier findings made with the less comprehensive Illumina 27k arrays [41] . First , we have demonstrated that most of the epigenetic drift is not caused by underlying changes in blood cell subtype proportions . Indeed , focusing on the 5% most significant age-associated DNA methylation changes , we observed that over 80% of these changes retained statistical significance in an analysis adjusted for shifts in blood cell subtype proportions . Intriguingly , we observed that age-DMRs were enriched most strongly for CpG island hypermethylation and that this enrichment increased upon adjustment for cellular heterogeneity ( Fig . 1 ) . We note that although opensea and shore/shelf probes were generally much more numerous among age-DMRs ( Figs . 1B & 1E ) , that this seems to only reflect the bias of the 450k array towards this class of probes . Normalizing for this bias , as we did here , shows that CGIs have the highest probability of being an age-DMR ( Figs . 1A & 1D ) . The stronger enrichment of CGIs in the adjusted analysis is also consistent with two previous observations: First , that the main age-associated shift in blood cell subtype composition , ie . an increase in the granulocyte to lymphocyte ratio , is accompanied by a global loss of methylation [7] , and secondly , that a significant proportion of developmental DMRs are distal to promoter CGIs and TSSs [19] . Thus , if a substantial proportion of the myeloid/lymphoid cell subtype specificity is conveyed by DNA methylation patterns in shore/shelf and open-sea regions , then this would explain why these regions are less enriched after adjustment for the age-associated myeloid/lymphoid skewing . All these are important observations , because it suggests that the mechanisms leading to the age-associated modulation of the DNA methylome are very distinct to those implicated in development and differentiation . This is perhaps not surprising , since otherwise drift , which is already prominent in early life [42 , 43] , would have a dramatic effect on normal tissue homeostasis well before the normal aging effects become visible [10] . Our integrative analysis with a large blood gene expression data set further supports this view , since we found that drift does not significantly alter in-cis gene expression levels , a result which is also consistent with a recent age-related gene expression study conducted in lymphoblastoid cell-lines [44] . In fact , quite remarkably , we observed that drift may act to stabilize pre-existing baseline levels of gene expression . Indeed , we found that age-associated hypermethylation at CGIs preferentially targets genes that are not expressed in blood tissue , whereas age-associated hypomethylation correspondingly targets more highly expressed genes ( Figs . 4A & B ) . This confirms an earlier observation made by Day et al with the older Illumina 27k beadarrays [41] , and was further validated in another large gene expression data set ( Fig . 4B ) . It is important to point out that our observation that a significant proportion of epigenetic drift is not caused by changes in blood-cell type composition does not contradict the observations and recommendations of Jaffe and Irizarry [14] . As shown here , adjustment for blood-cell type composition does remove some of the epigenetic drift attributable to the increase in the granulocyte to lymphocyte ratio ( Fig . 1F & S2_Fig in S1 Text ) . Most importantly , however , it should be clear that the effect of cell type compositional changes on inferred DNA methylation patterns will depend on two main factors: the tissue type and the phenotype being considered . Thus , when comparing patterns of DNA methylation in blood tissue between cancer patients and healthy controls , there is a strong need for adjustment because the presence of the tumour induces dramatic changes in blood-cell type composition [12 , 14 , 15] . Likewise , when comparing DNAm patterns in blood of Rheumatoid Arthritis ( RA ) patients to that of controls , adjustment is critical as demonstrated by Liu et al [11] . Our data strongly supports the view that age has a much less dramatic effect on changes in blood cell type composition , compared to cancer-presence or RA . Another important contribution is the demonstration of large megabase-scale blocks of age-associated hypomethylation , covering 14% of the genome , and , importantly , that only a subset of these blocks are enriched for age-associated CGI hypermethylation ( Fig . 2 , Table S1 in S1 Text ) . Interestingly , we also found that age-associated hypomethylated blocks preferentially undergo hypomethylation in cancer compared to other open sea regions which do not change with age ( Fig . 3 ) . However , we also observed some differences in relation to what has been observed in cancer . Notably , in cancer most hypomethylated blocks are enriched for CGI hypermethylation [18] . In contrast , CGI hypermethylation within age-hypomethylated blocks was only seen for a relatively small fraction of blocks . Given that most of the age-associated CGI hypermethylation is independent of blood cell subtype , it is highly plausible that these specific changes are also present in other tissue types . Indeed , that the age-hypermethylated component of drift appears to be specially independent of tissue type was an observation made by us previously with Illumina 27k beadarrays [2] and further confirmed by others [28] . Thus , it is of interest to consider the specific genes targeted for aberrant DNA methylation within these blocks . For instance , we identified a hypomethylated block on chromosome-4 , containing age-hypermethylated CGIs targeting the promoter of the HAND2 gene . This transcription factor has been shown to be causally implicated , through DNAm induced silencing , in the development of endometrial cancer [29] . Specifically HAND2 mediates the tumour suppressive effects of progesterone [29] . Interestingly , age is also one of the main risk factors for endometrial cancer , hence it is plausible that age-associated HAND2 promoter methylation , if present in endometrial tissue , could be a contributing factor to endometrial cancer risk . Indeed , there is already prior evidence that HAND2’s promoter undergoes age-associated DNA hypermethylation in epithelial tissues [1] . Interestingly , the effect of HAND2 methylation on endometrial cancer risk is mediated by methylation and silencing of HAND2 in endometrial stromal non-immune cells , increasing paracrine signaling through release of fibroblast growth factors thus sensitizing the endoemtrial epithelial cells to oncogenic estrogen [29] . That the DNA methylation of HAND2 in cancer tissue is not attributable to immune cell infiltration , is an important observation because it suggests that some of the common hypermethylation seen in aged blood tissue and in cancer tissue may not be due to immune-cell activation and tumour-infiltration of these activated immune cells . In this work we also performed an integrative analysis of epigenetic drift , as measured with Illumina 450k beadarrays , with ENCODE data , and specifically with ChIP-Seq transcription factor binding site data for a total of 58 TFs as determined in a human embryonic stem cell line . This analysis not only identified specific polycomb factors ( SUZ12 , EZH2 ) , previously already known to be targeted by age-associated DNA hypermethylation [2 , 3] , but also several interesting regulatory factors , including RBBP5 , NRSF , SIN3A , TAF1 , EGR1 and CTBP2 , some of which ( e . g . RBBP5 ) have not been previously implicated in aging . For instance , RBBP5 ( retinoblastoma binding protein-5 ) , a protein whose role in aging is only implied from homology ( JenAge AgeFactDB database ) [45] , is part of the MLL1/MLL complex , whose role is to methylate/di-methylate lysine-4 of histone H3 , which is a tag for epigenetic transcriptional activation . Thus , age-induced impairment of RBBP5 binding could lead to functional disruption of the MLL complex and hence to loss of H3K4 methylation , a well-known aging effect [46 , 47] . Interestingly , RBBP5 has also been shown to interact with TAF1 , TAF7 and TBP , all members of the transcription factor IID ( TFIID ) multiprotein complex , and all of which had binding sites enriched among age-hyperM DMRs , even in the multivariate analysis ( Fig . 6B & S8_Fig in S1 Text ) . Both TAF1 and TBP have entries in the GenAge HAGR database [48] , with TAF1 being essential for cell cycle progression . Another TF with an entry in the GenAge HAGR database is SIN3A: interestingly , SIN3A has been shown to function in histone deacetylase pathways [49] ( S9_Fig in S1 Text ) , as well as in the deacetylation of the c-MYC protein , thus contributing to its repression [50] . Of note , MXI1 , an interacting partner of SIN3A , was also enriched in our TFBS analysis ( Table S4 in S1 Text ) , although it lost the enrichment in the multivariate analysis ( Fig . 6B ) . Yet another TF with enriched binding sites among age-hyperM DMRs and with an entry in the GenAge database is EGR1 . This gene is an important regulator of the cell-cycle , with pro-apoptotic functions and acting upstream of TP53 ( S9_Fig in S1 Text ) . Finally , our list also included NRSF/REST , which has a role in suppressing genes which promote Alzheimer’s disease [40] . Thus , our data suggests age-associated functional disruption of REST , which may allow Alzheimer promoting genes to be expressed . In addition , we observed that CTCF was one of a few number of transcription factors whose binding sites were enriched among both age-hypomethylated DMRs , suggesting a global redistribution of chromatin patterns with age . Thus , overall , our integrative analysis points towards an age-associated disruption of DNA binding of transcription factors with important roles in histone deacetylation , histone methylation , chromatin architecture and tumour suppressor pathways . In this regard , it is worth emphasizing some of the key differences with the corresponding integrative ENCODE analysis performed in the context of cellular differentiation . For instance , pluripotency factors such as NANOG or OCT4 . and chromatin factors such as HDAC2 or CTCF2 , all of which play key roles in differentiation from hESCs , played a much less significant role in the context of aging , wheras the opposite is true for the polycomb factors EZH2 and SUZ12 .
In summary , this work has shown that age-associated DNA methylation changes seen in blood are largely independent of changes in blood cell type composition , and reflect patterns of change at different genomic length scales which are reminiscent of those seen in cancer . By integrating Illumina 450k with ENCODE data we have furthermore identified a number of candidate key transcription factors whose regulatory potential may be disrupted during aging .
For the analysis of aging we used one of the largest available data sets , which profiled over 656 whole blood samples using Illumina Infinium 450k beadarrays [5] . Data was downloaded from GEO and subjected to a stringent quality control analysis , including imputation of missing values , type-2 probe bias correction using BMIQ [51] , as well as assessment of the sources of inter-sample variation using a Singular Value Decomposition [12] . The top component of variation correlated with Source site of samples ( i . e . laboratory sample was processed ) , plate and ethnicity . Lower ranked components correlated with gender and age . Samples came from 4 different labs , were processed on 9 different plates ( indexed here as 1 , 2 , 3 , 5 , 6 , 8 , 9 , 10 , 11 ) and came from 2 different ethnic groups ( white caucasian and hispanic ) [5] . For the analysis relating to cellular development and differentiation , we downloaded the Illumina 450k data of GSE31848 from the GEO website . This data consisted of a total of 153 samples , encompassing human embryonic stem cell lines , induced pluripotent stem cells , primary cell lines and somatic differentiated samples from a range of different tissue types . Probes with more than 5% missing values across the samples were removed from analysis . Rest of missing data was imputed using the k-nearest neighbour procedure , as implemented in the impute R-package [52] . Because of the inherent bias of type-2 probes and our desire to analyse DNA methylation patterns spatially , we also adjusted this data using the BMIQ algorithm [51] . Cross-reactive probes and probes with SNPs in them were kept in the DNA methylation data analysis . We verified that the main results of this work are independent of whether these probes are included or removed . Transcription Factor ChIP-seq Uniform Peaks . narrowPeak files were downloaded from the UCSC ENCODE website ( http://hgdownload . cse . ucsc . edu/goldenPath/hg19/encodeDCC/wgEncodeAwgTfbsUniform/ ) . This track contains 690 ChIP-seq datasets representing 161 unique regulatory factors ( generic and sequence-specific factors ) . The dataset spans 91 human cell types and various treatment conditions . These datasets were generated by the five ENCODE TFBS ChIP-seq production groups: Broad , Stanford/Yale/UC-Davis/Harvard , HudsonAlpha Institute , University of Texas-Austin and University of Washington , and University of Chicago . Some TFs were generated by more than one group , or selected by different antibodies . In all cases , human Illumina 450K CpGs were mapped to TF binding sites . Only CpGs falling within the ChIP-Seq peak boundaries were considered as defining an overlap . Here we considered mainly two cell-lines: H1-hESC and HepG2 . The number of profiled TFs were 58 ( H1-hESC ) and 75 ( HepG2 ) . To integrate the blood DNA methylation data from Hannum et al , with gene expression , we first used a whole blood data set ( 233 samples ) generated using the Illumina HumanRef-12v3 array [33] . The normalized data was provided at the probe-level . Probes mapping to the same Entrez gene ID were averaged resulting in a 16345×233 gene expression data matrix . This data matrix underwent further quality control using a Singular Value Decomposition ( SVD ) method [12] to assess the sources of inter-sample variation . We found that the top PC correlated with beadchip and not age . In deriving differentially expressed genes ( DEGs ) associated with age using a linear regression model , we also observed statistical confounding , as deduced from the shape of the P-value histogram [53] . To address this problem , we identified and ranked age-DEGs using the ISVA-algorithm [53] . As a validation of the gene expression analysis , we also downloaded another relatively large gene expression data of 198 whole samples [36] . This data was generated with Agilent Whole Human Genome Microarrays [36] . The normalized data was provided at the probe-level . Probes mapping to the same Entrez gene ID were averaged resulting in a 19751×198 gene expression data matrix . This data matrix underwent further quality control using a Singular Value Decomposition ( SVD ) method [12] to assess the sources of inter-sample variation . In deriving differentially expressed genes ( DEGs ) associated with age using a linear regression model , we adjusted for sex , case/control status and smoking status . In order to obtain estimates of baseline levels of gene expression in blood we collected an expression data of cord blood and placenta samples [35] . This data was generated with Illumina Human Ref-8 beadarrays [35] . The normalized data was provided at the probe-level . Probes mapping to the same Entrez gene ID were averaged resulting in a 18342×183 gene expression data matrix . This data matrix underwent further quality control using a Singular Value Decomposition ( SVD ) method [12] to assess the sources of inter-sample variation . In this data set there were 64 cord blood samples , 54 placenta samples and 65 maternal blood samples . We restricted to analysis of cord blood and placenta samples from non-smokers ( n = 45 cord blood , n = 37 placenta ) . To identify age-DMRs in the Hannum et al data [5] , we used linear regressions with plate and gender as covariates . Note that we did not include source and ethnicity as covariates because these were fully correlated with plate . Indeed , plates-1 , 2 , 3 , 8 , 10 and 11 contained only caucasian samples , specifically their numbers were 90 , 84 , 69 , 81 , 17 and 85 . The other plates , i . e . plates-5 , 6 and 9 only contained hispanics ( 90 , 92 and 48 ) . It follows from this design that adjusting for plate automatically adjusts for ethnicity , but not vice-versa . Similarly , plate and lab were fully correlated , thus adjusting for plate also adjusts for source-site . Moreover , it is well known that adding highly correlated covariates in multivariate regressions may cause singular or near-singular predictor matrices and thus compromise statistical inference . For these reasons , and given that lower ranked SVD components also correlated with gender ( gender was equally distributed across plates , with each plate containing males and females ) , we used as covariates plate and gender . The linear regressions above were performed at the level of regional probe clusters , following the procedure of Aryee et al [26] . Specifically , 450k probes were first divided into 3 regional classes: open-sea , shore/shelves and CpG Islands ( CGIs ) . We then used the boundedClusterMaker function with maxGap = 500 and maxClusterWidth = 1500 on probes within each of these 3 classes , separately , resulting in probe clusters within each class . Beta-values for probes within a cluster were averaged , and linear regressions were performed for these averaged values , i . e one linear regression for each probe cluster . Finally , we also remark that we chose a linear model , mainly because previous works have shown how well linear models are able to predict age across a wide range of tissue types [2 , 6] . We followed the reference-based procedure of Houseman et al [15] , using the R-scripts and reference CpG list provided in that publication . To clarify , we ran the Reference-Based method using reference profiles for CD8+ T-cells , NK-cells , Monocytes , CD4+ T-cells , B-cells and Granulocytes , to infer cell-type proportions in the individual samples . These were subsequently used as covariates in the linear regression analysis . In order to see if age-associated DMRs occur on longer length ( i . e megabase ) scales , we followed the procedure of Aryee et al [26] , focusing on the previously constructed regional clusters mapping only to open-sea probes . To define blocks we used the clusterMaker function of the bumphunter package [26 , 27] with maxGap = 250000 and using the previously constructed regional open sea clusters as input . To find extended DMRs ( specifically hypomethylated block regions ) we used the bumphunter algorithm with bpSpan = 250000 and 500 bootstrap iterations . To test enrichment of transcription factor binding sites among age-hypermethylated and age-hypomethylated DMRs we used a one-tailed Fisher’s exact test . However , since two transcription factors can share the same binding sites , we devised a simple mulitvariate framework to assess if the association of a given TF binding site profile with age-DMRs is independent of the other TF binding site profiles . Specifically , denoting by tr the t-statistic derived from a linear regression of a regional cluster’s ( r ) average DNA methylation level against age , and denoting by bf the binding site profile of a given transcription factor f ( so that bfr = 1 if a binding site of the TF falls within region r , bfr = 0 otherwise ) , we performed the multivariate regression t = α 0 + ∑ f b f α f + ϵ ( 1 ) It follows from standard multivariate regression analysis , that the t-statistics associated with the estimated parameters α^f assess the association of any given TF’s binding profile with age-DMRs independently of other TF’s binding profiles .
|
Two well-known features of aging are the gradual decline of the body’s ability to regenerate tissues , as well as an increased incidence of diseases like cancer and Alzheimers . One of the most recent exciting findings which may underlie the aging process is a gradual modification of DNA , called epigenetic drift , which is effected by the covalent addition and removal of methyl groups , which in turn can deregulate the activity of nearby genes . However , this study presents the most convincing evidence to date that epigenetic drift acts to stabilize the activity levels of nearby genes . This study shows that instead , epigenetic drift may act primarly to disrupt DNA binding patterns of proteins which regulate the activity of many genes , and moreover identifies specific regulatory proteins with key roles in cancer and Alzheimers . The study also performs the most comprehensive analysis of epigenetic drift at different spatial scales , demonstrating that epigenetic drift on the largest length scales is highly reminiscent of those seen in cancer . In summary , this work substantially supports the view that epigenetic drift may contribute to the age-associated increased risk of diseases like cancer and Alzheimers , by disrupting master regulators of genomewide gene activity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Conclusions",
"Methods"
] |
[] |
2015
|
An Integrative Multi-scale Analysis of the Dynamic DNA Methylation Landscape in Aging
|
Naturally acquired clinical immunity to Plasmodium falciparum is partly mediated by antibodies directed at parasite-derived antigens expressed on the surface of red blood cells which mediate disease and are extremely diverse . Unlike children , adults recognize a broad range of variant surface antigens ( VSAs ) and are protected from severe disease . Though crucial to the design and feasibility of an effective malaria vaccine , it is not yet known whether immunity arises through cumulative exposure to each of many antigenic types , cross-reactivity between antigenic types , or some other mechanism . In this study , we measured plasma antibody responses of 36 children with symptomatic malaria to a diverse panel of 36 recombinant proteins comprising part of the DBLα domain ( the ‘DBLα-tag’ ) of PfEMP1 , a major class of VSAs . We found that although plasma antibody responses were highly specific to individual antigens , serological profiles of responses across antigens fell into one of just two distinct types . One type was found almost exclusively in children that succumbed to severe disease ( 19 out of 20 ) while the other occurred in all children with mild disease ( 16 out of 16 ) . Moreover , children with severe malaria had serological profiles that were narrower in antigen specificity and shorter-lived than those in children with mild malaria . Borrowing a novel technique used in influenza–antigenic cartography—we mapped these dichotomous serological profiles to amino acid sequence variation within a small sub-region of the PfEMP1 DBLα domain . By applying our methodology on a larger scale , it should be possible to identify epitopes responsible for eliciting the protective version of serological profiles to PfEMP1 thereby accelerating development of a broadly effective anti-disease malaria vaccine .
The surface of red blood cells ( RBCs ) infected with Plasmodium falciparum contains antigens of parasite origin that are highly immunogenic and genetically very diverse [1] . Diversity in variant surface antigens ( VSAs ) plays an important role in immune evasion and thus in prolonging infections: this affords parasites more opportunities to transmit to new hosts . Acquisition of antibodies to the most studied family of VSAs–P . falciparum erythrocyte membrane protein 1 ( PfEMP1 ) —is associated with protection against malarial disease [2 , 3] . Since PfEMP1 also plays a key role in pathology due to its property as an adhesion ligand to host cells [4] , this protein family makes an attractive target for vaccine development . PfEMP1 is encoded by approximately 60 var genes which are genetically diverse within and between parasite genomes , and which recombine , thus potentially presenting a challenge in finding a handful of antigens that could form the basis of a broadly effective vaccine . Despite their genetic diversity , however , var genes structure into distinct groups thus somewhat limiting this pool of variability . These groupings are based on chromosomal position and upstream sequence ( Ups ) [5 , 6]; combinations of domains and sub-domains ( domain cassettes , DC ) [5 , 7–9]; and homology in short sequence blocks found across the full gene ( ‘homology blocks’ , HB ) [7 , 9] or at positions of limited variability ( PoLV ) within the DBLα domain [10] . These different classification systems partially overlap [11 , 12] . Some var genetic groups have been consistently associated with severe disease , denoted here as ‘SM types’ , namely , those with Group A-type upstream promoters ( UpsA ) [13] , domain cassettes 8 and 13 [14] , the REY motif at the PoLV2 position of the DBLα domain [15] , and the presence of two cysteines ( cys2 ) between the PolV3 and PolV4 positions in DBLα [13 , 16] . PfEMP1 host cell adhesion phenotypes ICAM-1 , EPCR and rosetting have been mapped to expression of specific domains or domain cassettes thus providing a clear link between PfEMP1 diversity , adhesion phenotype and disease severity [4] . Antigenic properties of PfEMP1 are , by contrast , poorly understood . It is known that antibodies against var types associated with severe malaria , such as UpsA and DC8 , are readily detected in young children living in malaria-endemic regions and develop before antibodies to other types [13 , 15 , 17–20] . In addition , VSAs from parasites found in younger children and those with severe disease are more frequently recognized by sera than those from older or more immune children: that is , they are more ‘immunologically common’ [2 , 13 , 15 , 21 , 22] . However , the antigenic properties that differentiate these types are not known . It has been hypothesised that some var types–those found in younger and sicker children—have immunological properties that differentially affect the quality and efficacy of the PfEMP1 antibody response [13] . Possible reasons for stronger antibody recognition of these types are that they have higher immunogenicity , elevated transcriptional levels , greater protein abundance on the red cell surface , stronger antigenic conservation , wider cross-reactivity or better ability to grow inside the host and thus become immunodominant . Some of these hypotheses are supported by recent studies [13 , 15 , 20 , 23] . Here , we describe the antigenic diversity of PfEMP1 in relation to severe vs . mild malaria in order to understand this further . To do so , we borrow a technique used in the study of immune cross-reactivity of influenza A viruses called ‘antigenic cartography’ [24 , 25] that is used to annually update flu vaccines in order to cover the ‘antigenic space’ of the circulating pathogen population [26] . Our overall aim is to lay the foundation for inferring the subset of antigens that might form a broadly protective PfEMP1-based vaccine .
We measured sero-reactivity , both IgG and IgM , of anti-sera from 36 children with severe or mild malaria to a panel of recombinant proteins that represented part of the DBLα domain -the DBLα-tag—of dominantly expressed PfEMP1 types found in clinical isolates of 36 children ( Fig 1 , S1 Fig ) . 32 of the antigens derived from the 36 children that provided antisera . As expected given the endemicity of malaria in our study site in coastal Kenya , at the time of recruitment , children in our study had substantial prior history of malaria exposure , evident from their generally broad recognition and significantly higher sero-reactivity than unexposed control individuals to recombinant DBLα-tag proteins ( Fig 1 ) . Sero-reactivity of IgG with individual DBLα-tag proteins was generally higher than sero-reactivity of IgM at acute disease . This is most likely to be because IgG plasma antibodies reactive with individual DBLα-tags were generated during previous infections while IgM responses present induction of de novo responses from naïve B cells and , to a lesser extent , IgM memory B cells . We used analysis of variance to determine the principal factors affecting sero-reactivity: these revealed that both antisera and DBLα-tag antigens varied substantially in their average reactivity across antigens/sera , respectively explaining 39% and 12% of the total variation for IgG , and 13% and 7% for IgM under analysis Model 1 ( Fig 1 ) . These differences were maintained across timepoints ( Fig 1 , S1 , S2 and S3 Figs ) . In addition , there was strong specificity of reactivity of individual anti-sera to individual DBLα-tag antigens: this explained a further 24% and 15% of the total variation in IgG and IgM reactivity , respectively , after excluding data from homologous ( i . e . , derived from the same child ) DBLα-tag-antiserum pairs . These patterns of sero-reactivity were exploited to produce ‘serological maps’ that reflect shared patterns of reactivity of individual sera with all DBLα-tag antigens , and ‘antigen maps’ that reflect shared patterns of reactivity of individual DBLα-tags with all anti-sera . Distances between points in these maps correspond to differences in reactivity patterns between antigens or between sera , respectively . Serological maps revealed two distinct clusters which we denote serological Clusters I and II hereon . For IgG , these serological clusters were apparent at the acute stage ( Fig 2A ) and remained stable throughout the convalescent period ( Fig 2B , S1 & S4 Figs ) . By contrast , for IgM , serological clusters were not evident at the acute stage ( Fig 2C ) but , as convalescence progressed , despite the overall low sero-reactivity compared to IgG , two clusters emerged which were the same as those for IgG ( Fig 2D , S5 Fig ) . Serological clusters were still evident after pre-adjusting sero-reactivity data for mean antisera reactivity ( S4 and S5 Figs vs . Fig 2 ) and thus were not generated by systematic differences in mean reactivity of sera . Instead , they were driven by the highly specific nature of serum-DBLα-tag reactivities: this was most obvious for serological Cluster II antisera which reacted strongly with DBLα-tag antigens P6433 , P6408 and P6430 while serological Cluster I sera did not ( Fig 1 , S1 Fig ) . DBLα-tag antigen maps , by contrast , did not show strong clustering at any timepoint for either IgG or IgM ( S6 , S7 and S8 Figs ) . Instead , distances between DBLα-tag antigens were principally a function of differences in a given DBLα-tag antigen’s mean reactivity ( S6 Fig ) . Nonetheless , DBLα-tag antigens grouped into five groups ( denoted DBLα-tag antigen Clusters III to VII ) using a combination of hierarchical clustering ( S7 Fig ) and the proportion of antisera strongly recognized ( log10 OD > 2 ) as follows: Antigen Cluster III , ‘high reactivity’ with >75% anti-sera reactive with these DBLα-tags; Antigen Cluster IV , ‘high specificity’ with 40–60% of anti-sera strongly reactive and the remaining anti-sera with low reactivity; Antigen Cluster V , ‘medium specificity and reactivity’ with medium reactivity of anti-sera to a broad set of DBLα-tag antigens ( 19–64% ) that were largely non-overlapping with the highly reactive antigens that defined Cluster IV; Antigen Cluster VI with ‘low reactivity’ with <25% of anti-sera reactive; and Antigen Cluster VII with ‘zero reactivity’ with 0% of anti-sera reactive . Antigen clustering weakened upon pre-adjusting for antigen mean reactivity ( S7 and S8 Figs vs . S1 Fig ) . Antigenic clusters did not significantly associate with serological clusters ( S6 Fig , P = 0 . 20 by chi-squared test of association on 4 d . f . ) . Serological clusters almost perfectly aligned to clinical group and thus disease severity: whereas all serological Cluster I sera derived from patients with mild malaria attending the hospital outpatients department ( n = 16 ) , all but one ( n = 19 ) of the Cluster II sera derived from patients admitted to the hospital wards ( Fig 3A , P < 0 . 001 by chi-squared test on 1 d . f . ) . Since host age , parasite density at the time of sampling and average reactivity of the serum did not differ significantly between clinical groups ( Table 1 , P > 0 . 05 by t-tests ) , the association between serological clusters and disease is unlikely to have arisen from different levels of prior exposure in the two clinical groups . There was a significant effect of clinical group on date of sampling because all six plasma samples collected in 2008 were sourced from patients admitted to the hospital ( Table 1 ) . However , since the association between serological clusters and disease severity remained highly significant ( P < 0 . 001 ) after excluding data from the 2008 samples , this confounding was not responsible for the association , as might be the case , for example , if the frequency of severe malaria-causing genotypes in the population changed from year to year . DBLα-tag antigen clusters did not significantly associate with clinical group ( P = 0 . 36 by chi-squared test on 4 d . f . , Fig 3B ) . However , all three of the DBLα-tag antigens with very high specificity , reacting strongly to approximately half the plasma ( antigen Cluster IV ) , derived from patients with mild disease who had serological Cluster I profiles . DBLα-tag antigen cluster was not significantly associated with other host infection parameters ( Table 1 ) . At the acute stage of infection , IgG sero-reactivity of homologous antigen-antiserum pairs ( i . e . , both the plasma and parasite isolate from which the dominant DBLα-tag was cloned were sourced from the same child ) was significantly lower than for heterologous antigen-antiserum pairs ( P < 0 . 001 under Model 2 , Fig 4A vs . 4B and 4C ) . This effect disappeared in the convalescent phase during which antibodies to homologous DBLα-tags significantly increased ( P < 0 . 001 fitting timepoint as a linear covariate ) while those to heterologous DBLα-tags significantly declined ( see below ) . These findings are consistent with those from previous studies showing that children are more likely to become infected with PfEMP1 antigenic types not encountered previously , and that they subsequently mount antibody responses that are primarily specific to the infecting antigenic type [2 , 27–29] . There were no differences between serological Cluster I and II in the homologous IgG response ( P > 0 . 05 at all timepoints under Model 4 , Fig 4A ) . By contrast , Cluster I and II antisera responded differently to heterologous DBLα-tag antigens . Acute stage plasma belonging to serological Cluster I had lower IgG reactivity to heterologous DBLα-tag antigens from children with the same serological cluster than to antigens from the opposite cluster ( Fig 4B vs . 4C , P < 0 . 01 and P < 0 . 05 at the A and C2 timepoints under Model 3 ) , similar to the heterology effect for individual antigen-antisera pairs . By contrast , serological Cluster II had similar or higher IgG sero-reactivity to DBLα-tags from children with the same serological cluster than to antigens from the opposite cluster ( P = 0 . 04 for timepoint C3 and P > 0 . 05 at the A and C2 timepoints under Model 3 , Fig 4B vs . 4C ) . Serological Cluster II profiles further differed from serological Cluster I profiles in that–with the exception of responses to homologous antigens—antisera levels declined significantly during convalescence ( Fig 4B and 4C , P < 0 . 001 for Cluster II vs . P > 0 . 05 for Cluster I fitting timepoint as a linear covariate in Model 3 ) . Serological cluster and cluster heterology effects for IgM patterns were qualitatively similar , though weaker , to those for IgG ( Fig 4D , 4E and 4F ) . Combined , these results suggest that children develop one of just two qualitatively distinct serological profiles with respect to the DBLα-tag antigen which align with disease outcome of the infection . Serological Cluster I profiles appear to be broadly reactive , long-lasting and , while less effective in protecting against infection with non-SM than SM types , nonetheless are effective in preventing progression to severe disease . Serological Cluster II profiles , by contrast , while generally stronger at the time of acute infection , appear to be narrowly specific , transient , non-protective against infection with all types—both SM and non-SM—and ineffective in preventing progression to severe disease . Genetic maps based on the full amino acid sequence of the DBLα-tag region of PfEMP1 , like serological maps , revealed two genetic clusters ( denoted Genetic Clusters VIII and IX ) ( Fig 5 ) . These corresponded to the 2-cysteine vs . not-2-cysteine major genetic groupings described previously [10] ( Fig 5B , P < 0 . 001 by chi-squared test ) . Genetic cluster and PoLV group did not significantly associate with clinical group ( Fig 3C , P = 1 . 00 and P = 0 . 20 , respectively ) , serological cluster ( P = 1 . 00 and P = 0 . 13 , respectively ) , or DBLα-tag antigen cluster ( P = 0 . 91 and P = 0 . 35 , respectively ) or any of the host and infection variables described above ( Table 1 ) . Sliding window analysis of genetic maps based on 14 amino acid sub-regions in relation to serological , DBLα-tag antigenic and genetic clusters indicated that serological clustering was principally driven by genetic diversity between Segments 2 and 3 at the N-terminal portion of the DBLα-tag ending with PoLV2 ( Fig 6A and S9 Fig ) . This region comprises almost complete S2b and S2c sub-domains of DBLα domains as defined by Rask et al . ( 2010 ) [9] . High concordance between serological clustering and genetic diversity in this region was driven by the fact that all but one ( P6408 ) of the nine antigens that contained a REY motif at the PoLV2 position ( PoLV Groups 2 and 5 ) derived from patients with serological Cluster II profiles , while only one REY type was found in patients with serological Cluster I profiles ( Fig 7 , S9B and S9E Fig ) . Previous evidence suggests that parasites expressing PoLV/Cys Group 2 [10] and Group 5 [32] are likely to form rosettes , a cytoadhesion phenotype strongly implicated in severe disease [33–35] . This , together with our findings , suggests that epitopes in the S2b sub-domain of DBLα REY sequence type , or elsewhere in the PfEMP1 protein but in strong linkage disequilibrium with the latter , are the target of antibodies that are associated with a milder course of disease . Genetic clustering was strongly associated with sequence diversity in Segments 4 and 5 at the 3’ end of the S2c sub-domain of DBLα , consistent with location of the cysteines that distinguish two major genetic groups under the PoLV/Cys classification system [10] ( Fig 6B ) . We explored other genetic aspects of the DBLα domain for their relationship with disease and serological clusters . Recognition scores for homology blocks HB14 and HB64 were significantly elevated in DBLα-tags derived from parasites from patients with mild disease and serological Cluster I profiles , while those of HB54 were elevated in DBLα-tags derived from parasites from patients with severe disease and serological Cluster II profiles ( Table 1 ) . These three homology blocks span Segments 1 , 2 and 3 of the DBLα domain ( Fig 7 ) , thus corroborating the strong association between serological cluster and genetic diversity in the S2b sub-domain described above . Recognition scores of HB60 and HB36 were significantly positively associated with Genetic Clusters VIII and IX , respectively ( Table 1 ) , consistent with their location in Segments 4 and 5 in which variation in the number of cysteines that delineate these clusters occurs ( Fig 7 ) . The analysis above focused on sero-reactivity of plasma to the DBLα-tags of the dominant expressed PfEMP1 in clinical isolates . We next addressed whether underlying genetic characteristics of the full length PfEMP1 protein in the entire parasite population of a given child , such as expression of specific domain cassettes or upstream promoter sequences were associated with serological or antigenic clusters . We found that expression levels of domain cassettes DC8 and DC13 were significantly positively associated with severe disease and serological Cluster II or both ( Table 1 ) , consistent with previous studies [14 , 20] . Expression levels for three out of four other markers of var subtypes associated with severe disease in previous studies were significantly associated with disease in this study ( Table 1 ) : two of these markers define DBLα sub-types that belong to the upstream promoter Group A ( DBLa2/a1 . 1/2/4/7 and DBLa1 ) . Surprisingly , a generic group C marker that has been previously associated with mild disease was also elevated in children admitted to hospital: however , expression levels of Group C upstream promoter regions were low in all groups ( Table 1 ) . Global population frequency of the parasite’s var genes , as judged by number of highly similar sequences to the full DBLα–tag in the global var gene sequence database , did not associate with disease severity or clusters ( Table 1 ) . Thus , the association between disease severity and antigenic profile was not explained by the rarity of , and hence degree of prior exposure to the var gene products of the infecting isolate . Overall , our genetic analysis of the parasites in this study confirms the conclusion from previous studies that a subset of genetically defined PfEMP1 types–those with DC8 , two cysteines , and Group 2 and Group 5 REY types—are strongly predictive of severe disease . By introducing serology into analysis of this relationship , we have shown that protection against these SM types may depend on serological responses to epitopes in the S2b sub-domain of DBLα , or epitopes elsewhere in the PfEMP1 protein that are in strong linkage disequilibrium with these . By contrast , protection was not associated with putative epitopes in the S2c sub-domain which delineates the two major genetic groups by its number of cysteines , Cys2 and non-Cys2 .
This study has revealed that , despite different histories of malaria and diversity of antigenic types , serological responses to PfEMP1 in children fall into one of just two qualitatively different patterns which strongly associate with disease severity . In most children , serological responses to PfEMP1 ( here evaluated using recombinant DBLα-tags representing the dominantly expressed PfEMP1 ) are broadly reactive and long-lasting , but in some children–those that succumb to severe disease—responses are narrower and short-lived . This conclusion , based on the similarity of serological profiles between children to a set of related antigens rather than on the frequency of responses to individual antigens–as in most immuno-epidemiological studies—departs from the traditional view that protection against malarial disease is acquired in a piecemeal fashion through acquisition of specific antibodies to each of PfEMP1’s many antigenic types upon repeated exposure [36 , 37] . It shifts the focus in understanding severe malaria away from individual antigenic types towards the overall quality of the antibody response . As in many previous studies [13 , 15 , 17–20] , we observed that PfEMP1 ( here represented by DBLα-tags ) from parasites that caused infections which progressed to severe disease were generally better recognized by plasma than those from infections resulting in mild disease ( Figs 1 and 4 ) . How does this reconcile with our proposal that patients with severe disease mount serological responses that are less effective and less persistent ? Although var genes are very heterogenous in sequence within and between genomes , they share sequence similarities that define domains , homology blocks and binding-sites for endothelial receptors , as also reflected in short shared stretches of sequence similarities within the DBLα-tag region [38] . These sites could be targets of cross-reactive serological responses . Cross-reactive antibody responses can be short-lived when the corresponding antibodies originate from memory B cells that differentiate into short-lived plasma cells , or long-lived when antibodies originate from long-lived plasma cells . Our data suggest that children with mild disease maintain stable antibody responses to heterologous DBLα-tag antigens , with levels generally higher to DBLα-tag antigens derived from parasites of children with severe malaria , while children with severe disease have similar levels of serological responses to heterologous DBLα-tag antigens of all parasite types which drop significantly during convalescence . Given that memory B cells tend to leave germinal centres earlier than long-lived plasma cells during affinity maturation [39] , it seems likely that the short-lived serological Cluster II IgG responses originated from short-lived plasma cells that had lower affinity for heterologous DBLα-tags , and that these qualitative differences contributed to the association of serological profiles with disease severity that we observed here . Remarkably , serological profiles for IgM and IgG were similar at the convalescent stage . IgM memory B cells contribute—and sometimes are the earliest responders [40]–to the rapid boosting of antibody levels during re-infection . However , differentiation of naïve B cells in response to new antigenic variants increases levels of plasma IgM memory B cells too . Although the source of plasma antibodies to the different DBLα-tag antigens has to be further investigated , we thus hypothesise that the different sources and quality of memory B cells may underlie the reason for why children that have developed Cluster I serological profiles in response to previous infections are protected from severe disease , even when they become infected with SM types , while children with Cluster II profiles are not . We further note that the decoupling of the roles of antigen and serological responses in severe disease under this hypothesis contrasts with that for establishment of an infection which clearly depends on highly specific reactivity to individual antigenic types . Although severe disease is clearly linked with PfEMP1 type , such as expression of DC13 , DC8 or group A PfEMP1 associated with rosetting , expression of these SM-PfEMP1 types is not sufficient for the development of severe disease since they are also expressed during first infections , whether or not the child develops severe malaria , and because children readily induce antibodies against SM-PfEMP1 [14] . Our data suggests that this gap is filled by some qualitative feature of the host’s immune response that causes failure to develop an adequate anti-SM-PfEMP1 response . Possible reasons for this include host-derived factors that influence the general quality of the immune response such as haemoglobinopathies which alter display of PfEMP1 on the red cell surface [41]; parasite-induced epigenetic modification of immune cell function [42]; PfEMP1-mediated inflammation resulting in dysregulation of T- and B cell function [43] or antibody-dependent cellular cytotoxic function [44]; and RIFIN-mediated binding to LILRB1 expressed on B cells [45] . Clearly , underlying host genetic factors and cellular events that result in distinct serological profiles between children with mild vs . severe malaria disease require further investigation . Alternatively , the failure may be directly attributable to the parasite . Theoretical modelling of within-host infection dynamics has led to the hypothesis that in order to produce sequential expression of individual var genes during the course of an infection , the parasite must induce the immune system to produce long-lasting , cross-reactive antibodies as well as transient , highly specific antibodies to the current numerically dominant variant [46] . Empirical studies have shown that , in patients with severe disease , var gene expression is dysregulated , leading to the appearance of multiple var types simultaneously instead of sequentially: this has been linked to down-regulation of parasite histone deacetylase PfSIR2A , an epigenetic silencer of var gene expression [20 , 47] . The latter two observations , together with our results here , support a model in which hosts mount one of two types of immune response to PfEMP1 –one ( Cluster I-like ) in which var gene expression is well-regulated , this with assistance from long-lasting antibody seroreactivity , and which culminates in control of severe disease; and the other ( Cluster II-like ) in which many vars are expressed simultaneously , generating an antibody repertoire that is broader but less effective , presumably as a consequence of generating fewer antibodies to each type and which thus gives the appearance of transience . Such a model must include an antigen-antiserum specific component , however , since our conclusion that protected children mount broader and longer serological responses than unprotected children does not sufficiently explain the serological reactivity patterns in our data . We found that it was not the overall quantity of plasma antibody , but instead the highly specific nature of the immune profiles which delineated serological Cluster I from Cluster II and mild from severe disease patients . The serological specificity that drove these differences was particularly obvious among a small subset of DBLα-tag antigens ( P6433 , P6408 and P6430 , Fig 1 , S7 Fig ) . Perhaps surprisingly , these ‘indicator’ antigens were only recognized by children with severe disease and all of them derived from children with mild disease ( S1 Fig ) . We therefore interpret strong responses to this subset of antigens in children with severe malaria as an indication of , rather than the cause of , a defective immune response . A variant-specific basis for the relationship between serological antibody profile and disease outcome is further supported by the results of our analysis relating distinct serological profiles to genetic variation within PfEMP1 . We mapped diversity in serological responses to genetic diversity in HB14 , a short region of semi-conserved sequence in the S2b sub-domain of PfEMP1’s DBLα domain . HB14 spans a hypervariable loop ( Fig 6 & S9 Fig ) between two alpha helices that form the core structure of this sub-domain: these are marked by conserved sequences HB3 at the 5’ end and HB5 at the 3’ end . HB5 is believed to be frequently exposed on the surface of PfEMP1 [9]: the adjacent hypervariable region marked by HB14 may also be frequently exposed to the host’s immune system . Just downstream from HB14 and marking the beginning of HB5 is a short motif–REY or RED . REY defines genetic sub-groups ( PoLV Groups 2 and 5 ) which have previously been reported to be associated with severe malaria , particularly rosetting [32] . REY-type HB5 tend to have shorter segments in the HB14 region ( Fig 7 ) . Sero-reactivities against REY-containing DBLα-tag sequences are acquired early in life in children living in Papua New Guinea [48] . Taken together , these findings suggest that the length and/or diversity of the hypervariable loop exposed on the surface of PfEMP1 , as defined by HB14 , might provoke antibody responses that qualitatively differ in their efficacy in controlling progression to severe disease . Since we did not sequence the full length of the var genes , we could not determine the complete set of genetic variants that best predict serological profile and how these relate to Group A , DC8 and DC13 var types that generate severe malaria . Given the genetic structure of var genes , it is likely that HB14 and HB5 in the DBLα domain are in linkage disequilibrium with cassette types in other domains . A much larger study is required to identify all the potential epitope sites in PfEMP1 that could explain the dichotomous serological profiles that we have revealed here . Such a study would constitute a significant step towards the development of a broadly effective PfEMP1-based malaria vaccine . Unlike vaccines for other highly genetically diverse pathogens such as influenza A , in which antigen components are selected to cover the full antigenic diversity of the currently circulating population , and which require updating for each global hemisphere every year [25 , 26] , the challenge will be to define PfEMP1 types that elicit the long-lasting , protective version of the antibody repertoire , as found in most children in malaria-endemic areas , and which we have begun to define here .
Children were recruited into this study after written informed consent from their parents or guardians . Ethical permission for the study was granted by the National Ethical Review Committees of Kenya ( Kenya Medical Research Institute protocol SSC1131 ) and the Oxford Tropical Research Ethics Committee ( OXTREC protocol no . 30–06 ) . Thirty-six children under the age of 5 years presenting with acute malaria at the Kilifi County Hospital , Kilifi , Kenya each donated a venous blood sample at recruitment ( Acute , A ) , at 4 weeks ( Convalescence 1 , C1 ) and at 16 weeks ( C2 ) after the acute malaria episode . Children were either admitted to the hospital ward according to pre-established criteria for severe malaria ( Blantyre coma score <5 , Hb<5g/dl or respiratory distress [49] ) or moderate malaria ( admitted to the ward but without any of these severe malaria syndromes ) : otherwise they were diagnosed as ‘mild malaria’ , treated with anti-malarial drugs at the Outpatient Department ( OPD ) and sent home . Plasma and parasite material were collected upon recruitment and separated and stored using standard procedures [50] . The dominant expressed var gene in each isolate at the time of recruitment was identified based on the DBLα-tag as described previously [50 , 51] . For 36 of the isolates for which a dominant var transcript was able to be identified , DBLα tags were cloned and expressed in BL21 ( DE3 ) pLysS E . coli cells to give recombinant proteins ( ‘antigens’ ) as previously described [50] ( S1 Table ) . Each antigen was screened for sero-reactivity to each of 36 plasma ( ‘antisera’ ) collected at each of the three timepoints ( A , C1 and C2 ) . Thirty-two of these plasma samples came from the same children from which the 36 recombinant proteins derived . Sero-reactivity was measured by coating 96-well NuncTM MaxisorbTM immunoplates plates overnight at 4°C with recombinant protein diluted in Tris-buffered saline ( TBS ) at a concentration of 1μg/ml . Plates were blocked with 3% bovine serum albumin ( BSA ) in TBS , washed and then incubated in duplicate at a 1:200 dilution of individual sera diluted in 1% non-fat milk powder in TBS . After incubation for 2h at room temperature , plates were washed , then bound antibody was detected with mouse anti-human IgG alkaline phosphatase antibody or AffiniPure donkey anti-human IgM alkaline phosphatase antibody ( Jackson ImmunoResearch Laboratories , Inc . ) diluted 1:5000 in 1% non-fat milk powder in TBS . The reaction was developed using o-Phenylenediamine dihydrochloride ( Sigma-Aldrich ) and then stopped with 2M sulphuric acid . Optical density ( OD ) was read at 450nm in an ELISA microplate reader ( BioTek Synergy 4 , BioTek Instruments ) . Prior to building antigen and serological maps , sero-reactivity levels ( log10 OD values ) were explored for systematic effects of timepoint , DBLα-tag antigen and antisera by conducting analysis of variance fitting a series of mixed linear models in the lme4 package in R [52] . Model 1 included a fixed effect for timepoint ( sometimes fitted as a fixed-level factor and sometimes as a continuous variable ) and random effects for antigen and antisera within antigen: the latter term thus tested for interactions between antigens and antisera . Model 2 was as for Model 1 but substituting antisera within antigen with a fixed effect for ‘heterology’ , i . e . , whether the antigen-antisera pair was from the same patient ( homologous ) or not ( heterologous ) . Model 3 , which was only applied to data from heterologous antigen-antisera pairs , fitted fixed effects for timepoint , serological cluster of the antisera ( see below ) , and serological cluster of the antigen , with all possible interactions between these , and random effects for antigen and antisera . Model 4 , which was only applied to data from homologous antigen-antisera pairs , fitted fixed effects for timepoint , serological cluster of the antisera , the interaction between these and a random effect for antisera . Model 5 fitted fixed effects for timepoint and antigen cluster ( see below ) and random effects for antigen and antisera . Models 6 and 7 were as for Model 5 but fitting genetic cluster ( see below ) or clinical group instead of antigen cluster . Means for fixed effects after adjusting for other fixed and random effects ( least-squares means ) were computed using the lsmeans package in R [53] . Pairwise contrasts between fixed effects of interest were formed using the same package with no adjustment for multiple testing . The variation due to antigen , antisera and their interaction was described by the ratio of the corresponding variance estimate ( from the random effect ) to the total variance remaining after accounting for fixed effects . Significance levels of all fixed and random effects were determined by likelihood ratios from models with and without the term of interest included . Estimates presented are from models in which interaction terms with P > 0 . 05 were removed . Associations between categorical variables ( disease severity class ( ‘clinical group’ ) , antigen cluster , serological cluster and genetic cluster , as defined by maps ( see below ) ) and normally distributed host and infection-related variables ( parasite density , age , date of collection at the time of recruitment ) were tested for significance using F-tests in a fixed effect analysis of variance with the latter as the dependent variables . Associations between categorical variables and non-normally distributed traits ( DNA sequence global population frequency , homology block scores , relative gene expression levels for domain cassettes , see section below on genetic characteristics ) were tested for significance using a two-sided Wilcoxon rank sum non-parametric test implemented by the ‘wilcox . test’ function in the R stats package [54] . Tests for relationships between clusters and categorical variables were performed using chi-squared tests of association using the ‘chisq . test’ function in the R stats package with Monte Carlo simulation for computation of P-values [54] . Antisera ‘maps’ ( a representation of pairwise antisera ‘distances’ , with respect to the antigens they ‘read’ antigens ) were generated from the 36 x 36 matrix of sero-reactivity data using multidimensional scaling ( MDS ) implemented by the ‘cmdscale’ function in the stats package of R [54] . The elements of the distance matrix used as input for MDS were computed as ( dij/dmax ) 2 where dij is the Euclidean distance between the vector of sero-reactivities of antisera pairs i and j and dmax is the maximum Euclidean distance for all pairs , calculated using the ‘adjacency’ function in the WGCNA package in R [55] . Maps were fitted in two dimensions and constructed from data with and without adjusting for mean differences in sero-reactivity across antigens of each antisera . Two further aids to visualization of antisera diversity were used . First , heatmaps of the distance matrix were created using the pheatmap package in R [56] and plotted with dendrograms based on hierarchical clustering ( ‘hclust’ function using the ‘average’ method in the R stats package [54] . Second , networks were constructed from the lowest 20% of distances in the distance matrix using the Davidson-Harel algorithm in the igraph package in R [57] . Antigen maps were similarly constructed but based on distances between pairs of antigens instead of antisera . Note that the application of ‘antigenic cartography’ here differs from its original use [24 , 25] in that the plasma were polyclonal rather than monoclonal due to prior history of malaria in study children and the fact that multiple PfEMP1 types are expressed within the lifetime of a single infection . We did not attempt to adjust for previously existing levels of antibodies since our main interest was to relate existing antisera profiles at the time of infection , rather than responses to the current infection , to disease severity . Genetic maps were constructed by MDS of the distance matrix based on amino acid sequence similarities after alignment using the clustalw algorithm implemented in the Geneious software [58] with gap open and extension penalties of 12 and 3 , respectively . The genetic distance matrix was calculated using the ‘seqidentity’ function in the bio3d package in R [59] . Antigenic , serological and genetic clusters were defined by a combination of visual inspection of MDS maps and hierarchical clustering of the distance matrices . Antigens were categorized based on their amino acid sequence into six ‘PoLV/Cys2’ groups ( Groups 1 to 6 ) based on motifs at four ‘positions of limited variability’ ( PoLV ) within the DBLα domain and by number of cysteine residues ( Cys ) - 2 vs . not-2—in the region between the third and fourth PoLV ( PolV3 and PolV4 ) as described previously [10 , 51] . The two major groups defined by number of cysteines ( Groups 1–3 vs . Groups 4–6 ) further divide into those containing a REY vs . non-REY motif at the second PoLV ( PoLV2 ) ( Groups 2 and 5 vs . Groups 1 , 3 , 4 and 6 ) . REY types are shorter than non-REY types and are further distinct in their sequence between PoLV1 and PoLV2 [10] . Antigens were also classified for the presence of predicted ‘homology blocks’ by analyzing the DBLα-tag sequences on the “varDom” server [9 , 31] . The presence of specific domain cassettes , particularly ( DC ) 8 and 13—combinations of protein sequences from the DBLα , DBLα , DBLα and CIDR domains of PfEMP1 that have been previously strongly associated with severe malaria [14 , 20]–were determined using real-time quantitative PCR for each parasite isolate as described previously [14] . Upstream promoter type ( A , B or C ) of the antigens was likewise determined by real-time PCR [9 , 14] . Nucleotide sequences of each DBLα-tag was blasted against the DNA sequence in the global var gene database [30] . The number of var genes in the database which contained DBLα-tag nucleotide stretches with 95% or more identity to each DBLα-tag sequence was counted to estimate the ‘global population frequency’ of the sequence . To determine whether some sub-regions of the DBLα-tag sequence might better explain antigenic or antisera diversity than others , a sliding window analysis was performed in which a genetic sub-map built from sequence similarity based on windows of 14 amino acids , each offset by one , was assessed for concordance with the consensus antigenic or antisera map . Concordance was statistically evaluated by fitting a linear model to the distance matrix of the genetic sub-map with antigen or serological cluster as a fixed effect using the ‘adonis’ multivariate analysis of variance function in the vegan package in R [60] . 1200 permutations of the data were performed to determine significance levels . Concordance between genetic and antigenic or antisera maps was also assessed by computing the sum of squared differences ( SS ) between point locations in the two maps after rotating and scale transforming the antigenic or antisera map such that it minimized the SS , as implemented by the ‘procrustes’ function in the vegan package in R [60] . The P-value for observed goodness-of-fit statistic , R2 ( = 1-SS ) , was compared to that of its empirical distribution generated from 1200 random permutations of the antigenic or antisera map using the ‘protest’ function in the vegan package This second statistical test was applied to maps built both across and within antigenic or antisera clusters .
|
Immunity to human malaria parasites never fully protects against infection , even after a lifetime of exposure . By contrast , protection against severe disease occurs early in life in malaria-endemic areas . Both anti-infection and anti-disease immunity depend on antibody responses to proteins expressed by the parasite on the red blood cell surface which cause pathology . These proteins are extremely diverse thus creating a problem for designing a widely effective vaccine . Despite its importance to vaccine design , however , it is not yet known whether protection against malaria depends on accumulation of exposure to each of the many antigenic types , or whether this is accelerated due to cross-reactivity between antigenic types or some other mechanism . In this study , by applying a novel technique used for describing antigenic diversity in influenza viruses–antigenic cartography—we make the surprising discoveries that children’s serological responses to a panel of diverse surface antigens fall into one of just two qualitatively distinct patterns , and that these almost perfectly predict severity of disease . These alternative serological profiles were found to associate with genetic variants within a small region of the protein . By using the methodology presented here to define the antigenic types and their underlying genetic variants that give rise to the protective version of the serological response , progress towards development of a broadly protective vaccine against severe malaria might be accelerated .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"serum",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"immunology",
"tropical",
"diseases",
"parasitic",
"diseases",
"sequence",
"motif",
"analysis",
"molecular",
"biology",
"techniques",
"antibodies",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"immune",
"system",
"proteins",
"serology",
"gene",
"mapping",
"bioinformatics",
"proteins",
"molecular",
"biology",
"biochemistry",
"blood",
"anatomy",
"physiology",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"malaria",
"immune",
"serum",
"genetics",
"of",
"disease"
] |
2019
|
Antigenic cartography of immune responses to Plasmodium falciparum erythrocyte membrane protein 1 (PfEMP1)
|
The evolution of heteromorphic sex chromosomes creates a genetic condition favoring the invasion of sex-ratio meiotic drive elements , resulting in the biased transmission of one sex chromosome over the other , in violation of Mendel's first law . The molecular mechanisms of sex-ratio meiotic drive may therefore help us to understand the evolutionary forces shaping the meiotic behavior of the sex chromosomes . Here we characterize a sex-ratio distorter on the X chromosome ( Dox ) in Drosophila simulans by genetic and molecular means . Intriguingly , Dox has very limited coding capacity . It evolved from another X-linked gene , which also evolved de nova . Through retrotransposition , Dox also gave rise to an autosomal suppressor , not much yang ( Nmy ) . An RNA interference mechanism seems to be involved in the suppression of the Dox distorter by the Nmy suppressor . Double mutant males of the genotype dox; nmy are normal for both sex-ratio and spermatogenesis . We postulate that recurrent bouts of sex-ratio meiotic drive and its subsequent suppression might underlie several common features observed in the heterogametic sex , including meiotic sex chromosome inactivation and achiasmy .
Sex chromosomes are believed to evolve from a pair of autosomes [1–3] . An incipient Y chromosome , like an autosome , is largely euchromatic and free to recombine , except for a small region determining sex , as exemplified by species such as the papaya plant [4] and the medaka and stickleback fish [5 , 6] . On an evolutionary time scale , the nonrecombining region of the Y will generally expand to include most or all of the chromosome , accompanied by an accumulation of transposable elements and other repetitive sequence , as well as mutational inactivation of most of the protein-coding genes . Only a small number of genes remain active in a mature Y chromosome , such as that in humans or Drosophila . Some Y-linked genes are vestiges of the degeneration process , while others have originated from autosomes as a result of recruiting male-specific genes such as those that function in spermatogenesis [7–10] . Accompanying the evolution of sex chromosomes , at least two problems of biological significance arise . One problem is the unequal gene dosage of sex-linked genes between the XY sex and the XX sex . Because of Y degeneration , most genes on the X have only one active copy in the XY sex but two in the XX sex . Myriad strategies to compensate the dosage inequality have been exploited by various species , and some of these mechanisms are now understood in molecular detail in model organisms of fly , worm , and mouse [11] . Another substantial but less obvious problem consists of genetic conflicts over the sex ratio among various parts of a genome , which would allow optimal transmission of their own genes . A corollary to sexual reproduction is Fisher's well-known principle that the sex ratio must be equal for a panmictic population of dioecious species [12] . However , as noted long ago , Fisher's principle applies only to autosomal genes but not to sex-linked genes . Genes linked to one or the other sex chromosome would have a selective advantage were the sex ratio in the population skewed [13] . Because of the genetic isolation between the sex chromosomes , mutations biasing the sex ratio can easily accumulate and enhance each other as long as their deleterious effects are offset by their biased transmission . Thus , the evolution of sex chromosomes leads to an intrinsic conflict among the X , the Y , and the autosomes with regard to sex ratio . Many cases of sex ratio distortion ( sex-ratio hereafter ) have been documented , particularly in taxa where intensive laboratory investigation is possible [14] . Because of the biased sex ratio , suppressors unlinked to a distorter are strongly selected to restore the Fisherian sex ratio [15] . The occurrence of sex-ratio in a population can often be transient and easily escape notice . However , recurrent bouts of sex-ratio invasion and suppression can modify the genetic architecture of gametogenesis to such an extent that hybrid incompatibility can be driven to evolve among isolated populations . In other words , genetic conflicts can be a key mechanism for speciation [16–18] . Several cases of sex-ratio have been reported in D . simulans [19–23] . In a companion paper , we reported the cloning of an autosomal sex-ratio suppressor [24] . As Fisher's principle predicts , there must exist an X-linked sex-ratio distorter to which this suppressor corresponds . Here we report the characterization of such a distorter . More generally , we speculate that sex-ratio distortion might underlie the evolution of meiotic sex chromosome inactivation and achiasmatic meiosis , two biological phenomena whose evolutionary origins still remain mysterious .
We previously cloned a D . simulans gene , not much yang ( nmy , polytene chromosome position 87F3 ) , in which the homozygous male mutant displays a female-biased sex ratio . This gene belongs to the Winters sex-ratio system , one of three independent sex-ratio systems found in this species [24] . We inferred that the wild-type ( Nmy ) function is a suppressor of sex-ratio distortion , and that there must be a corresponding X-linked sex-ratio distorter according to Fisher's principle of sex ratio evolution . By happenstance , we found an X chromosome that did not express the sex-ratio phenotype in homozygous nmy males ( Figure S1 ) . This X chromosome was thought to have a loss-of-function mutation in the gene ( s ) causing sex-ratio . We designate the mutant gene as distorter on the X ( dox ) . Other X chromosomes , including one from the stock y wam v2 f66 , did express sex-ratio when tested in the nmy background and were postulated to carry the distorting allele Dox ( Figure S2 ) . A preliminary mapping of dox was carried out through the scheme described in Figure 1 . Recombinant X chromosomes were tested for sex ratio ( proportion of females or k ) in the nmy background ( I in G5 of Figure 1 ) as well as in the nmy/+ background as control ( II in G5 ) . A total of 148 X chromosomes tested could be grouped into eight genotypic classes ( Figure 2A–2H ) . Several inferences can be drawn from the results . First , there are two and one recombinants in classes E and F , respectively , which are exceptional and thus informative in the mapping of dox , allowing it to be placed closely proximal to v at a distance of about 6% ( 3/51 ) of the w–v interval . Second , the major sex-ratio distorter shows less strength of distortion when a gene in the vicinity of f is absent . The reduced distortion can be inferred by comparing classes A ( k ± standard error of the mean [SEM] = 0 . 803 ± 0 . 014 , n = 22 ) , D ( 0 . 797 ± 0 . 019 , n = 7 ) , and F ( 0 . 786 ± 0 . 020 , n = 22 , excluding one losing the major distorter ) with class G that has a significantly lower sex ratio of 0 . 655 ± 0 . 013 ( n = 24 ) ( t-test , p << 0 . 001 ) . The first three classes have similar sex ratios ( analysis of variance [ANOVA] , p = 0 . 748 ) . We call the gene near f an enhancer of Dox ( E ( Dox ) ) because it alone does not cause sex ratio distortion ( class H , 0 . 509 ± 0 . 004 ) . In light of the above reasoning , the five class B recombinants showing sex-ratio probably have an inferred genotype of w Dox v2 e ( Dox ) ( 0 . 686 ± 0 . 018 ) , the same as class G with respect to sex-ratio distorter and enhancer ( t-test , p = 0 . 157 ) . Finally , based on numerous SNP sites found between the two parental X chromosomes , we genotyped a selected subset of the 148 chromosomes and narrowed the location of dox to a region of 215 kb between CG15316 ( 8E1–4 ) and nej ( 8F7–9 ) , which falls within an interval defined by two visible markers lz and v . The fine mapping of dox began with the construction of two X chromosomes of lzs Dox v2 f66 and y wam dox , whose phenotype with regard to Dox was confirmed by testing in a homozygous nmy background through the scheme described in Figure 1 ( G3 through G5 ) . A cross of + + lzs Dox v2 f66/y wam + dox + + females to Ubx/D males was set up , and 324 recombinants with crossovers between lz and v were obtained . We picked 22 lzs and 21 y wam v2 f66 X chromosomes with crossovers falling between CG15316 and nej to further test their sex-ratio phenotype in a homozygous nmy background , again using the scheme in Figure 1 ( G3 – G5 ) ( Figure 3A ) . The cross in G4 was carried out at 18 °C so that the sex-ratio phenotype of Dox can be fully expressed in G5 [24] . Each of the 43 recombinants was unambiguously classified as either Dox or dox . Four SNP markers were found in the CG15316–nej region ( Table S1 ) , and these were used to demarcate the crossover points for the 43 recombinants . There are two Dox and five dox lines , with their crossovers falling between the markers 5dox_III and C14/C17 . We sequenced ∼31 kb embracing this region ( Figure 3A ) . The two parental alleles are identical for the 20 , 791 bp within the 5dox_III–C14/C17 interval , except for a deletion of 105 bp ( Δ105 ) in dox ( Figure 3B ) . We confirmed the predicted presence or absence of the Δ105 element in the final seven informative recombinants . We sequenced six other D . simulans strains in the region between the primer pair DoxF4-DoxR4 that spans the Δ105 sequence ( Figure 3B; Text S3 ) . Two types of haplotypes were recognized . One is from the SR6 X chromosome that carries the Paris sex-ratio distorters [24 , 25] . Three copies of a 360-bp repeat were found within this haplotype . The other type is shared by all the other strains , with an insert of 3 , 833 bp found within the last 360-bp repeat . This 3 , 833-bp fragment has sequences homologous to the last three exons from the gene CG32702 of D . melanogaster . The CG32702 ortholog is missing in the current annotation of D . simulans genome ( Release 1 . 0 , http://genome . ucsc . edu/ ) . However , we did obtain a sequence of 18 . 7 kb covering the orthologous CG32702 region in D . simulans as well as its full length cDNA of 11 , 550 bp ( Figure 3B and Text S1 ) . The transcript consists of 15 exons , largely agreeing with the computational annotation of this gene in D . melanogaster , except for differences in two splice sites and one extra exon at the 5′ end . Apparently , this 3 , 833-bp insert ( designated Tp3833 ) was duplicated and transposed from a sequence of 3 , 549 bp ( designated Tp3549 ) in the 3′ region of CG32702 . Note that one copy of the 360-bp repeat is also present next to Tp3549 , suggesting that this repeat may have facilitated the transposition . The last two exons and part of exon 13 ( Ex13 ) of CG32702 are still intact in the Tp3833 region ( CG32702d , Figure 3B ) . Sequences from the homologous region between DoxF4-DoxR4 were obtained from one strain of each of the sibling species D . sechellia , D . mauritiana , as well as D . melanogaster . All species resemble SR6 in having various copy numbers of the 360-bp repeat ( Figure 3C ) . A phylogenetic analysis of these 360-bp repeats shows a monophyly of the eight copies from D . melanogaster , but a reticulate relationship among the rest , suggesting shared evolutionary history of this intergenic region in the D . simulans clade ( Figure S3 ) . However , it remains to be determined whether a Tp3833-like sequence can be found in D . mauritiana . The existence of a functional Nmy strongly suggests that a corresponding sex-ratio distorter like Dox may still segregating in this species [24] . Within Tp3549 , a fragment of 1 , 458 bp replaces a fragment of 1 , 408 bp downstream of the 3′ end of CG32702 in D . melanogaster , and these two sequences have no homology ( Figure 3B ) . Database searches suggest that the 1 , 458-bp sequence is absent from D . sechellia , D . yakuba , and D . erecta , but some similar fragments of 300–600 bp can be found dispersed in these genomes , often in multiple copies . Transcripts within Tp3549 were detected , and a new gene , which we designate as Mother of Dox ( MDox ) , is defined ( see below ) . Tp3549 and Tp3833 thus represent a fluid portion of the Drosophila genome that occasionally gains new functions . Initially the truncated version of CG32702 ( CG32702d ) appeared to be the best candidate for Dox because of its perfectly conserved open reading frames ( ORFs ) and intron–exon boundaries ( Figure 3B ) . However , we have not detected transcripts from CG32702d . Extensive 5′- rapid amplification of cDNA ends ( RACE ) experiments using gene-specific primers targeting CG32702d all failed . A 3′-RACE experiment did recover cDNAs , but they could be transcribed from the 3′-end of CG32702 , not of CG32702d . There is a divergent site ( C/A ) between Tp3549 and Tp3833 in the 1 , 919-bp region corresponding to the last three exons of CG32702 . Using the primers CG32702seqF26 and CG32702seqR26 ( F26 and R26 in Figure 4; Text S3 ) , only the CG32702 sequence can be amplified from cDNA ( Figure S4 ) . CG32702d is therefore unlikely to be transcribed or its expression is too low to be detected by reverse-transcription PCR ( RT-PCR ) . On the other hand , we have detected transcripts that cover the region of Δ105 in the opposite direction of CG32702d ( Figure 4 ) . Two transcripts from the allele Dox were recovered with either four or three introns . Their full lengths are 2 , 781 bp and 2 , 690 bp , respectively . From the allele dox , we have also recovered two full-length cDNAs identical to those of Dox , except that the exon III , 42 bp in length , is missing because of the deletion Δ105 ( Figure 4 ) . This 42-bp element is tandemly repeated in the cDNA of Dox but has only one copy in that of dox . Within Tp3549 , we have also recovered a full length cDNA antisense ( 2 , 564 bp ) to the 3′ end of CG32702 ( Figure 4 ) . MDox like Dox also has three introns in exactly the same sites , as well as the tandem repeats of 42 bp present in its cDNA ( Figure 4 ) . Surprisingly , all transcripts from the Dox and MDox loci have very limited coding potential . The largest ORFs of MDox in all three frames are shown in Figure 4 , and all but one fail to match any known sequences by BLASTX searches through the nr database . The one ORF that was predicted by Genscan encodes 62 amino acids ( aa ) , and this ORF highly matches ( 57/62 or 91% identity ) the C terminus of a D . melanogaster gene named CG8664 ( located in region 15F7 , proximal to the gene f ) . Similarly , only one of the largest ORFs of Dox has BLASTX hits . This is again the ORF of 69 aa predicted by Genscan and is homologous to the 62 aa ORF of MDox , although only part of it matches to CG8664 ( 38/44 or 86% identity ) due to a frameshift mutation . CG8664 has no known biological functions or phenotypes . In the orthologous position of CG8664 in the current D . simulans genome annotation , a fragment of 2 , 084 bp , instead of a CG8664 homolog , has been found . Part of this 2 , 084-bp fragment , approximately equivalent to the 1 , 458-bp element mentioned above ( Figure 3B ) , is recognized and has a high similarity ( 99 . 3% identity ) to a region within Dox . If the existence of this partial paralog of Dox in the f region is confirmed by experiment , it would be interesting to test it as the candidate gene for E ( Dox ) . The pair of tandem 42-bp elements essential for a functional Dox are located within an ORF of 157 aa ( Figure 4 ) . A 14-aa domain encoded by this 42-bp element has no known functions . If this ORF is ever translated , the tandem 14-aa domains appear to be required for the wild-type function of Dox as a sex-ratio distorter . Coding or noncoding , the molecular mechanism underlying the effect of Dox in rendering Y-bearing sperm dysfunctional awaits further experimental investigation . The wild-type function of MDox is not known , although the presence of the critical tandem repeats of the 42-bp element suggests its biochemical similarity to Dox . The phenotype of Dox as a sex-ratio distorter is uncovered if its suppressor , Nmy , is nonfunctional ( Figure 5A ) [24] . The Nmy transcript appears to form a stem-loop structure with a double-stranded RNA ( dsRNA ) stem of 345 bp , and small interfering RNAs ( siRNAs ) produced from the dsRNA stem could target and suppress Dox [24] . Hence , homology between Dox and Nmy is anticipated . Indeed , sequence comparisons suggest that Nmy originated from Dox through a retrotransposition event [24] . Specifically , the 345-bp dsRNA sequence from Nmy has extensive homology to the potential ORF of both Dox and MDox that contains exon III ( Figures 5B ) . The critical 42-bp element falls within an 85-bp region that has a perfect match with the stem region ( positions 264–390 in the alignment of Figure 5C ) . Whether or how either Dox or MDox is regulated by these hypothetical siRNAs is currently under investigation . The possibility that Dox evolved solely as a sex-ratio distorter and for no other reasons is supported by the normal phenotype of the double mutant dox; nmy . We have shown previously that the etiology of the Winters sex-ratio is the degeneration of the Y-bearing spermatids during their maturation , as observed both through transmission electron microscopy ( TEM ) and through light microscopy [24] . We carried out similar observations of the spermatogenesis of dox; nmy males at 16 °C . All stages of spermatid maturation appear to be normal as also found in Dox; Nmy wild type ( Figure 6A–6D , in comparison to Figures 4 and 5 in [24] ) . Quantitatively , 5 . 7% ( n = 1108 ) of spermatid heads appear to be abnormal under TEM , in a proportion similar to wild-type Dox; Nmy ( 5 . 8% , n = 1903 ) . With 4' , 6-diamidino-2-phenylindole ( DAPI ) staining , no abnormal spermatid head was observed among the 1 , 416 heads examined . Consistent with these cytological observations , the sex ratio of progeny from the dox; nmy males at 16 °C was 54% . As a comparison , a dox; Nmy male was similarly examined . No abnormal heads were observed among 1 , 058 spermatids , and the sex ratio when tested was also 54% . All the evidence together suggests that Dox is not an essential gene and is fully dispensable . Nmy is also dispensable if Dox is absent . The fate of a sex-ratio system can be loss , fixation , or stable polymorphism . Apparently , the Winters sex-ratio is still segregating in D . simulans [24] . The same is true for the Paris sex-ratio system that has been found in the same species [21] . Evidence from molecular population genetics shows that the Paris SR6 X chromosome has swept through African and Indian Ocean islands only recently ( less than 20 thousand years ago [ka] ) [25] . The presence of a functional Nmy suppressor in D . mauritiana suggests that the Winters sex-ratio evolved in the ancestor of the D . simulans clade [24] . The following genetic evidence will enforce the above conclusions and help to compare the evolutionary history of these two sex-ratio systems . We have introgressed the Y chromosome of D . sechellia into D . simulans ( D . sim Y[sech] ) in a background isogenic to simB ( Figure S6 ) . The success of this introgression was confirmed by fingerprinting with a Y-specific probe Y5g ( Figure 7A and Figure S2 ) . The D . sechellia Y chromosome was thus tested against the driving effect of either Dox ( Winters ) or SR6 ( Paris ) . The Dox/Y[sech] male expresses sex-ratio if nmy is homozygous , but does not if one copy of the functional Nmy gene is present ( Figure 7B ) . Hence the D . sechellia Y chromosome is equally sensitive to Dox as is the D . simulans Y chromosome . Intriguingly , SR6/Y[sech] males exhibit male-biased sex ratio distortion ( k = 0 . 33 ) . Unfortunately , similar introgression of the Y chromosome from D . mauritiana cannot be made because D . sim Y[mau] is sterile [26] . The above observations are consistent with the earlier estimate that the origin of Dox predates speciation among D . simulans , D . mauritiana , and D . sechellia about 200–400 ka [27] , whereas SR6 arose in D . simulans after the speciation [24 , 25] . Assuming that an Y-linked distorter causing male-biased sex ratio distortion has little chance of persistence as compared to an X-linked one causing female-biased sex ratio distortion [13 , 28] , we suppose that the Y[sech] still bears sensitive sequence to Dox as in the ancestral Y of the three species . We suggest that the male-biased sex ratio expressed by SR6/Y[sech] is a sign of evolutionary independence between the Y[sech] and the SR6 distorters . The etiology of SR6 has been attributed to loss or breakage of the D . simulans Y chromosome during meiosis II in SR6/Y[sim] males [29 , 30] . Our results support earlier findings by showing that male progeny from the SR6/Y[sim] father are sterile at a frequency of 19% . However , the frequency of sterile male progeny from an SR6/Y[sech] father is only 3% ( Fisher's exact test , p < 0 . 0001 ) , a number that is not different from the control ( 2% , Fisher's exact test , p = 0 . 284 ) ( Figure 7C ) . It is possible that SR6 does not cause loss or breakage of the Y[sech] , hence the etiology of the male-biased sex ratio may be different from similar male sex-ratio ( msr ) cases reported in D . pseudoobscura [31] and in D . affinis [32] , where a large number of nullo XY sperm are produced . The unique cytological mechanism underlying the male-biased sex ratio in SR6/Y[sech] males again suggests that the unequal sex ratio is a neomorph created by a genetic incompatibility between the two chromosomes , rather than a shared evolutionary history of sex-ratio .
The gene structures of Dox and Nmy strongly suggest that an RNAi mechanism is involved , just as in numerous transgenic studies where inverted repeats ( IR ) were used to silence target genes in eukaryotes ( e . g . , [33] ) . Most likely , the suppression of Dox by Nmy is through a classic RNAi pathway , also known as post-transcriptional gene silencing ( PTGS ) , which has been under intensive genetic and biochemical studies ( reviewed in [34] ) . In essence , 21–23 nucleotide ( nt ) siRNAs processed from dsRNA are responsible for guiding the active RNA-induced silencing complex ( RISC ) to homologous mRNA , resulting in the latter's subsequent cleavage [35] . The PTGS model for Dox/Nmy interactions can be readily tested by comparing the steady-state mRNA levels of Dox between Dox; Nmy and Dox; nmy males , and by detecting the binding of specific siRNAs with the RISC components . Because PTGS happens in the cytoplasm , and spermatid nuclei within a cyst share the same cytoplamic syncytium , the final gene product of Dox likely has a localized deleterious effect in the Y-bearing spermatid nuclei , whereas the presence of a Y or absence of an X must provide the primary cue that eventually leads to abnormal maturation of the Y-bearing sperm heads . Alternatively , a different type of RNAi mechanism could be involved in the Dox/Nmy interaction . A class of small RNA in the size range 24–29 nt has been identified as silencing intermediates in the control of repetitive sequences such as retrotransposons [36] . Unlike the classic RNAi machinery , the core proteins do not require DCL-1 , DCL-2 and AGO2 , and a different type of RNAi pathway ( repeat-associated small interfering RNAs or rasiRNAs ) has been proposed [36] . The rasiRNA pathway has also been shown to be responsible for silencing a possible cryptic sex-ratio meiotic drive distorter , Ste , in D . melanogaster [37] . The Y-linked Su ( Ste ) suppresses the deleterious effects of the X-linked Ste , including male sterility and meiotic drive [38] . Both Ste and Su ( Ste ) consist of repeats that share extensive homology , and rasiRNAs were shown to be the information carrier for the target specificity [36 , 39 , 40] . A third possible mechanism for silencing Dox might be at transcriptional level in a manner of co-suppression as first observed in plant transgenics , where the expressions of both an endogenous gene and the homologous transgene were down-regulated [41] . This type of transcriptional gene silencing ( TGS ) has been demonstrated in Drosophila [42 , 43] , and it requires physical contacts between homologous sequences and Polycomb group ( PcG ) proteins [44] . Note that an intact pair of inverted repeats is not required for an efficient TGS ( e . g . , [44] ) . In our case , nmy[1427] is a loss-of-function mutation that does not have an intact pair of inverted repeats but does have a 1 . 2-kb sequence paralogous to Dox , arguing against this type of TGS as a strong candidate mechanism for silencing Dox [24] . The X chromosomes of many species are condensed precociously in prophase of meiosis I when active transcription peaks in the autosomes [45] . The existence of meiotic sex chromosome inactivation ( MSCI , also known as X chromosome allocycly ) has been well established in several model organisms , either directly through the observation of precocious heterochromatin sex bodies [46–48] , or indirectly from genetic analysis of X-autosome translocations [49] as well as with genome-wide gene expression studies [50–55] . Recently , MSCI has been demonstrated in D . melanogaster by assaying transgenic expressions in the X chromosome [56] . Though sex bodies are the direct evidence for MSCI , they have not been observed in most species examined so far , including Drosophila [57] . The status of MSCI may be assayed with more sensitive methods such as the detection of histone modifications that relate to transcriptional activity ( e . g . , [48] ) . There are several hypotheses for the evolution of MSCI . One hypothesis is that MSCI evolves because of a need to suppress recombination between the two sex chromosomes [58] . Another hypothesis was coined as the SAXI hypothesis ( sexual antagonism and X inactivation ) . Because the X spends 2/3 of its evolutionary history in females , the X will be depleted of male-specific genes , and a feminized X would be under selection to be silenced during male meiosis [59] . A third hypothesis has been suggested in the light of the discovery of the meiotic silencing of unpaired DNA ( MSUD ) in Neurospora crassa [60 , 61] . MSUD is reasoned to have evolved for defending against invasion of transposons [62–64] . The connection between MSCI and MSUD is supported by the silencing of unpaired chromosomal fragments in the mouse and worm [65 , 66] . Each of the three hypotheses for MSCI captures only some specific features consequent to the evolution of sex chromosomes , and hence provides only a partial and proximate explanation for the evolution of MSCI . The ultimate cause of MSCI , of course , must be the degeneration of the Y or W chromosome . Following this line of reasoning , we propose yet another hypothesis that we call “the drive hypothesis” for the evolution of MSCI: there is a constant requirement for silencing sex-linked genes including potential sex-ratio distorters during meiosis because of an intrinsic conflict over the sex ratio within a genome accompanying sex chromosome evolution . We believe that the drive hypothesis provides mechanistically superior explanations to the other hypotheses for the evolution of MSCI for the following specific observations: ( 1 ) There is a predominant pattern of generating testis-specific genes through retrotransposition between the X chromosomes and autosomes [67–70] . Some of the retrotransposed sequences might be involved in creating new distorters and suppressors . ( 2 ) MSCI in the worm and mouse is dependent on a putative RNA-dependent RNA polymerase , suggesting the involvement of an RNAi-like mechanism [71] . The Dox/Nmy case provides a strong mechanistic connection between meiotic drive and the evolution of MSCI . Admittedly , it is not an easy task to test empirically and discriminate among the hypotheses described above . Although the drive hypothesis emphasizes the importance of meiotic drive in the evolution of MSCI , other sex-chromosome–specific features such as suppressed recombination , degeneration , and depletion of sex-specific genes could well be different facets of the same evolutionary process . These features may share biochemical components and have reinforced each other over evolutionary time . In addition to MSCI , achiasmatic meiosis is another evolutionary oddity that might also be rooted in sex chromosome evolution . An achiasmatic meiosis has no crossovers between homologs , even of autosomes . Between 20 and 30 independent evolutionary occurrences of achiasmatic meiosis have been recorded , and all of them are observed in the heterogametic sex [72] , with one possible exception in Tigriopus californicus [73 , 74] . Nevertheless , it is reasonable to generalize that achiasmatic meiosis evolves only in the heterogametic sex , as a result of occasional spillover of the molecular machinery that is simultaneously responsible for both suppressing meiotic drive and recombination between the sex chromosomes . This connection between achiasmy and heterogamety was actually presaged by J . B . S . Haldane and J . S . Huxley [75 , 76] , but here we specify meiotic drive as the major evolutionary cause . One would wonder why there are so many sex-ratio meiotic drive cases reported in Drosophila . Other than the geneticists' proclivity for counting flies in lab , does achiasmatic male meiosis in this genus also contribute to the abundance of meiotic drive systems ? Meiotic drive in ZW females is mechanistically different from that in XY males . For XY males , sex-ratio meiotic drive can be achieved either by loss of a sex chromosome during meiosis ( meiotic drive sensu stricto ) or by abnormal postmeiotic development ( gametic drive or meiotic drive sensu lato , as used throughout this article ) . To prevent gametic drive , stringent control of X-linked genes must be achieved through means such as MSCI . However , in ZW females , only one of the four haploid products will eventually end up in the functional egg during oogenesis and meiotic drive sensu stricto would be an easier means to achieve biased transmission than gametic drive . For the ZW species , centromere structure and centromere-binding proteins may play an important role in meiotic drive ( “centromere drive” ) [77] , and there may be no particular need for evolving MSCI as a defense against gametic drive , although in principle , occasional female gametic drive may still evolve such as a polar body distorter that can kill an egg . Indeed , there is a general lack of cytological observation of MSCI in the ZW females [57 , 78] , with seemingly one proved exception in the caddis fly Glyphotaelius pelludidus [79] . Consistent with this observation , there are only two independent origins of ZW achiasmy [72] . Another explanation for the absence of Z allocycly might be due to a lack of selection pressure to evolve sex-ratio distorters on the Z chromosome , because male-biased sex-ratio distorter cannot persist long in a population [13] , therefore it might have left little trace of impact on genomic evolution prior to its disappearance . The question of whether the Z chromosome lacks MSCI deserves special attention because its general absence would be inconsistent with the SAXI or MSUD hypotheses .
The D . simulans stocks are: ( 1 ) y wam v2 f66 from the Tucson Drosophila Stock Center; ( 2 ) w; e and simB ( w; nt; III ) [80]; ( 3 ) C ( 1 ) RM y w/lzs and Ubx/D from J . Coyne; ( 4 ) SSR12-2-7 ( w; nt; nmy ) [24]; and ( 5 ) Paris sex-ratio X chromosome SR6 and its standard ST8 as described previously [81] . The SR6 and ST8 X chromosomes are maintained by backcrossing males to females of the stock C ( 1 ) RM y w/lzs every generation . The D . sechellia stock 3588 is from A . Clark [23] . The D . mauriatiana × D . simulans introgression lines have been described before [18 , 80] . The following lines were used in this study: heterozygous introgression lines P40–46 nmy , P38–11 nmy , P38H77 Nmy , P40L12 nmy , and P40B13 Nmy , all having the genotype w; nt; P/ III ( Nmy ) where P represents the various semi-dominant P-element transgenes P[w+] marking the introgressed D . mauritiana material nearby , and III ( Nmy ) represents the third chromosome III with Nmy . These chromosomes are maintained by backcrossing P/III males to simB females every generation . The SSR ( skewed sex ratio ) line Q15 . 3 is from a previous D . simulans × D . sechellia hybridization experiment [23] . Several stocks were constructed for this study: C ( 1 ) RM y w/w; nt; III by backcrossing females of C ( 1 ) RM y w/lzs to simB males for >19 generations . C ( 1 ) RM y w/w dox; nt; nmy was constructed through a scheme described in Figure S5 . Another stock , w/Y[sech]; nt; III , which is isogenic to simB except that the Y chromosome is from D . sechellia 3588 , was constructed through a scheme described in Figure S6 . The stock y wam was constructed from the stocks w; e and y wam v2 f66 . Another stock , C ( 1 ) RM y w /lzs v2 f66 , was constructed from C ( 1 ) RM y w/lzs and y wam v2 f66 . All flies were reared on cornmeal-molasses-agar medium sprinkled with yeast grains at room temperature ( 22 ± 1 °C ) unless otherwise indicated . The sex ratio of a male was scored by mating this male with three tester virgin females , usually of the stock w; e , for 7 d before clearing all adults . The progeny were sexed and counted three times until the 19the day . Sex ratio ( k ) was calculated as percentage of females . SNPs were discovered by sequencing 500–1 , 000–bp PCR products from relevant X chromosomes . The primers were designed by targeting the D . melanogaster genome ( http://www . flybase . org/ ) , and the virtual PCR products were compared to the D . simulans sequences ( http://genome . wustl . edu/tools/blast ) for correcting any mismatches within the primers ( Table S1 ) . Genotyping was done directly by sequencing . Some other key reagents/kits are: LA Taq long PCR kits ( Takara ) ; EZ-Tn5 Insertion Kit for sequencing large DNA fragments ( Epicentre ) ; Lambda ZAP II vector for genomic library ( Stratagene ) ; TRIZOL Reagent for RNA isolation , SuperScript II Reverse Transcriptase and 3′ or 5′–RACE kits ( Invitrogen ) . Light microscopy and TEM procedures have been described previously [24] .
All sequences have been deposited in the GenBank database ( http://www . ncbi . nlm . nih . gov/Genbank/index . html ) and have been assigned the accession numbers EF596886-EF596899 .
|
Mendel's first law of genetics states that two alleles of a heterozygote are transmitted to the next generation at an equal ratio . The cornerstone of population genetics , this law states that the evolutionary fate of genetic variants is solely governed by their contribution to the good of their carriers . However , meiotic drive genes—which skew transmission in their own favor—can evolve under certain circumstances , even though they cause harm to the genome as a whole . Meiotic drive elements are often enriched on the two sex chromosomes ( i . e . , the X and the Y ) because of a lack of recombination between them . Here we describe the genetic and molecular characterization of a meiotic drive distorter on the X chromosome in Drosophila simulans . This distorter apparently formed de nova from yet another new gene . To fight back against this harmful distorter , the D . simulans genome has evolved an ingenious mechanism based on DNA sequence homology . We postulate that repeated meiotic drive invasion and its suppression could be a major mechanism for genome evolution , underlying the ultimate cause for the inactivation of sex chromosome during meiosis and the occasional loss of recombination ( achiasmy ) , which is observed only in the heterogametic ( XY ) sex .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"cell",
"biology",
"evolutionary",
"biology",
"molecular",
"biology",
"genetics",
"and",
"genomics"
] |
2007
|
A sex-ratio Meiotic Drive System in Drosophila simulans. II: An X-linked Distorter
|
Genome-wide association studies using commercially available outbred mice can detect genes involved in phenotypes of biomedical interest . Useful populations need high-frequency alleles to ensure high power to detect quantitative trait loci ( QTLs ) , low linkage disequilibrium between markers to obtain accurate mapping resolution , and an absence of population structure to prevent false positive associations . We surveyed 66 colonies for inbreeding , genetic diversity , and linkage disequilibrium , and we demonstrate that some have haplotype blocks of less than 100 Kb , enabling gene-level mapping resolution . The same alleles contribute to variation in different colonies , so that when mapping progress stalls in one , another can be used in its stead . Colonies are genetically diverse: 45% of the total genetic variation is attributable to differences between colonies . However , quantitative differences in allele frequencies , rather than the existence of private alleles , are responsible for these population differences . The colonies derive from a limited pool of ancestral haplotypes resembling those found in inbred strains: over 95% of sequence variants segregating in outbred populations are found in inbred strains . Consequently it is possible to impute the sequence of any mouse from a dense SNP map combined with inbred strain sequence data , which opens up the possibility of cataloguing and testing all variants for association , a situation that has so far eluded studies in completely outbred populations . We demonstrate the colonies' potential by identifying a deletion in the promoter of H2-Ea as the molecular change that strongly contributes to setting the ratio of CD4+ and CD8+ lymphocytes .
The design of an ideal population for gene mapping involves balancing the avoidance of rare alleles with the requirement for rapid linkage disequilibrium ( LD ) decay . High rates of LD decay are found in populations with large effective population sizes and many generations of random mating that accumulate recombinants to break up correlations between genotypes . Unfortunately , a necessary corollary is the presence of rare alleles as allele frequencies drift to extremes and new , rare , alleles arise as a consequence of mutations . The more rare alleles in a population , and the more they contribute to phenotypic variation , the more difficult it will be to detect quantitative trait loci ( QTLs ) using genome-wide association strategies that genotype only common alleles [1] . The best strategy might seem to be to choose animals from highly divergent populations , such as wild mice caught in many locations [2] , or from inbred lines derived from highly genetically divergent progenitor strains . This maximizes genetic diversity and seeks to overcome the limitations of using only a subset of the variation present in wild populations . However , mice from different populations will have a high proportion of private variants present in one population only . LD decay for the latter private variants will depend solely on recombinants accumulated during the creation of the colony , while LD decay for the former , common , variants is boosted by the ancestry of the founding populations . Furthermore , the power to detect a genetic effect increases with the minor allele frequency ( MAF ) of the causal variant . It follows that high power and mapping resolution is best obtained by using animals from the same mating population to reduce the number of private alleles . A related phenomenon is population structure , caused either by recent admixture or uneven degrees of relatedness , both of which should be avoided . Commercial mouse breeders maintain large colonies of outbred mice that may have the necessary genetic structure . LD in some outbred stocks has been shown to allow high-resolution mapping [3] , sufficient to identify genes [4] . Importantly , most outbred stocks are known to derive from animals from a single population , such as the ‘Swiss’ stocks which descend from two males and seven females imported from Lausanne , Switzerland [5] , indicating that the proportion of private alleles may be low . Figure 1 summarizes the known relationship between commercially available outbred stocks as of 2007 ( the time of this study ) and additional information is given in Text S1 . However , without systematic characterization of the genetic architecture of commercially available outbred mice it is not possible to evaluate the suitability of any particular colony for genome-wide association . In this paper we evaluate 66 populations to determine ( i ) whether inbreeding and population structure preclude the use of the population; ( ii ) whether linkage disequilibrium ( LD ) enables high-resolution mapping; ( iii ) whether the proportions of common and rare variants are favorable for genome-wide association mapping . In order to assess the latter , we tested the hypothesis that the outbred colonies are descended from a common source: the laboratory inbred strains . Populations in which this assumption holds true , and which have low levels of LD , are the most suitable for high-resolution mapping . Finally , we show how commercially available outbred mice can be used to go from genetic association to molecular lesion by identifying a deletion in the promoter of H2-Ea as the molecular change that strongly contributes to setting the ratio of CD4+ and CD8+ lymphocytes .
Table 1 lists the populations that we obtained for this study and the numbers of animals we used . We included three control populations , with known genetic characteristics: 12 Heterogeneous Stock ( HS ) mice [6] , 109 Collaborative Cross ( CC ) mice [7] , 94 inbred strains [8] and a population of wild mice caught from multiple sites in Arizona that is likely to represent a fully outbred population , similar to that used in a human genome-wide association study ( GWAS ) [9] . We use the term “colony” to mean a population of mice maintained as a mating population at a single location , and “stock” to mean a collection of colonies that are given the same stock designation by the breeders . For example , HsdWin:CFW-1 and Crl:CFW ( SW ) are two colonies from the same stock ( CFW ) . We follow the nomenclature for outbred stocks [10] , but add a two letter code for the country of origin and , when there are several cohorts from the same site , a code for the production room: e . g . Crl:CFW ( SW ) -US_P08 . We analyzed all colonies with 351 markers at two loci on chromosome 1 ( 131 . 6–134 . 5 Mb and 172 . 6–177 . 2 Mb ) , one locus on chromosome 4 ( 136 . 2–139 Mb ) , and one locus on chromosome 17 ( 32 . 6–38 . 9 Mb ) ( marker details are given in Table S1 ) . The loci were chosen because they include large effect QTLs detected in a mapping study in Heterogeneous Stock ( HS ) mice [6] that are easy and inexpensive to phenotype ( large effect QTLs , explaining more than 10% of the phenotypic variance , can be mapped with about 200 outbred animals ) . The QTLs were for serum alkaline phosphatase ( ALP ) on chromosome 4 , the ratio of CD4+ to CD8+ T-lymphocytes on chromosome 17 , concentration of high-density lipoproteins ( HDL ) in serum on chromosome 1 , and mean red cell volume also on chromosome 1 . The region on chromosome 17 includes the MHC , highly polymorphic in wild populations and a sensitive indicator therefore of any loss of heterozygosity . While these four loci constitute less than 1% of the genome , if QTLs cannot be mapped at high-resolution here , it is unlikely that colonies will be suitable for genome-wide mapping ( we also carried out genome-wide analyses in a subset of animals to test this assumption ) . SNPs at the four loci were spaced so as to allow us to make inferences about both long and short range LD . We assessed the extent of inbreeding and population structure , genetic drift over time , linkage disequilibrium , the proportions of common and rare variants and the extent of genetic differentiation between colonies . Each factor influences the value of a colony for genetic mapping . High rates of inbreeding make colonies less suitable for mapping because they contain fewer segregating QTLs . Table 1 gives four measures of inbreeding: mean minor allele frequency ( MAF ) , heterozygosity ( inbred colonies will score low on this measure ) ; the percentage of markers that failed a test of Hardy Weinberg equilibrium ( HWE ) [11] ( colonies that consist of inbred but unrelated individuals , will have high scores ) and a coefficient of inbreeding that compares the observed versus expected number of homozygous genotypes [12] . Four colonies are almost inbred ( with heterozygosities less than 5% ) : NTac:NIHBS-US , ClrHli:CD1-IL , Hsd:NIHSBC-IL , BK:W-UK . A further five colonies have heterozygosities between 5% and 10% and so are unlikely to be useful for mapping . Three colonies have inbreeding coefficients greater than 20% ( HsdHu:SABRA-IL , Sca:NMRI-SE_10an , HsdOla:MF1-IL ) and a further seven have values greater than 10% ( Table 1 ) . None of these colonies are suitable for genetic association studies . Colonies that consist of a mixture of relatives ( such as siblings , half siblings , cousins , second degree and third degree relatives ) will be difficult to use for mapping because the differing degrees of genetic relatedness introduce population structure . We looked for evidence of this using multi-dimensional scaling of identity by state ( IBS ) pairwise distance matrices [13] . Overall , we found two or more clusters in nineteen populations ( marked as such in Table 1 ) ( results for all populations are shown in Figure S1 ) . However , while we can observe gross population structure at the markers tested , our power to detect more subtle effects is limited as accurate determination with Fst less than 0 . 01 requires more than 20 , 000 markers [14] ( Fst is a measure of genetic diversity within and among populations [15] ) . We carried out genome-wide analyses in six colonies judged to be most suitable based on the 351 SNP analysis ( sparse set of SNPs ) . Three populations were genotyped using the 600K Affymetrix Mouse Diversity Array [16] . Three more populations were analyzed using a precursor to this array that , after removing poorly performing markers , gave approximately 170 , 000 genotypes . Results are given in Table 1 . To compare results from the high-density arrays to those obtained from using 351 markers , we made 1 , 000 random samples of 351 markers from the dense marker sets ( from four regions matched in size ) , measuring heterozygosity and inbreeding in each sample . Using the samples we calculated the distribution of each statistic . The mean of the distribution coincided with the value obtained from the whole genome analysis . We then found the percentile position on this distribution of the results we obtained from the 351 markers . In all cases , the results lay within 10% of the distribution mean , indicating that results from our sparse marker set are representative . One potential concern surrounding the use of outbred colonies is that their genetic constitution is not stable and will fluctuate over time , due to unintended directional selection and random genetic drift . Table 1 demonstrates that most colonies are maintained with population sizes of many thousands , which should reduce the effects of shifting allele frequencies . We tested whether this was so by re-sampling six colonies at least one year after our initial analysis and in five cases found good agreement between heterozygosity , relatedness , and inbreeding measured on the two occasions ( Table 2 ) . In one case we noted a change in genetic structure . Results obtained from HsdOla:MF1-UK animals used in 2003 were different from those purchased in 2007: heterozygosity fell from 30% to 5% and the inbreeding coefficient rose from 3 to more than 30 . Due to infection , the colony had been reformed from a small number of re-derived founders , thereby introducing a severe population bottleneck and explaining the changes in genetic architecture . However , such drastic changes are unusual , are known to the breeders and can be ascertained in advance . Low LD is a requirement for high-resolution mapping . We assessed resolution at the four test loci by the LD decay radius , defined as the average physical separation in base pairs between SNPs beyond which the average squared correlation coefficient ( r2 ) drops below 0 . 5 ( Table 1 ) . Figure 2 shows results for all populations analysed ( there were insufficient polymorphic SNPs to calculate LD for NTac:NIHBS-US and ClrHli:CD1-IL ) . Average figures of LD decay mask variation between regions . For example HsdWin:NMRI-NL has a mean LD decay radius of just over 1 , but it will be of little use mapping the MHC region where LD is extensive . However , a region with high LD in one population may have low LD in another . This locus-to-locus variation means that no single population is ideal and that colony-specific genome-wide haplotype and recombination maps are needed . We explored genome-wide variation in LD in three colonies analysed with the 600K Mouse Diversity Array [16]: Crl:CFW ( SW ) -US_P08 , Crl:NMRI ( Han ) -FR and Hsd:ICR ( CD1 ) -FR . Mean block length varied between the three colonies: Crl:CFW ( SW ) -US_P08 79 . 2 Kb ( standard deviation ( sd ) 70 . 8 ) , Crl:NMRI ( Han ) -FR 39 . 53 Kb ( sd 58 . 7 ) , and Hsd:ICR ( CD1 ) -FR 51 . 1 Kb ( sd 79 . 5 ) . Block data for each chromosome is given at http://www . well . ox . ac . uk/flint-old/outbreds . shtml . Since there is on average about one gene per 100 Kb , gene-level resolution mapping is possible in these three colonies . Genome-wide association will be effective in colonies where all , or the majority of haplotypes are tagged by markers on a high-density array . The colonies' ancestry , as depicted in Figure 1 , suggests they contain a relatively limited set of haplotypes , present in inbred strains . We estimated the contribution of each inbred strain to each colony's genetic architecture by reconstructing the genome of each mouse as a probabilistic mosaic of the founders [17] . We used the Perlegen NIEHS genotypes [18] from 15 inbred strains and analysed all colonies at the four loci ( Figure 3 ) and performed genome-wide analyses in six colonies . While there is considerable variation between colonies , two general patterns are clear in both locus-specific and genome-wide analyses . First , in all colonies , the fraction of haplotypes accounted for by classical inbred strains ranges between 42% ( the NIHS colonies ) to 80% ( most ICR/CD1 ) . Second , the wild-derived strains ( WSB , CAST , MOLF ) contribute the least ( 3–5% ) . The NIHS stocks contain the highest contribution of the Swiss mouse FVB ( 25–35% ) . NMRI are 15–20% FVB and 15% 129 , CD1 about 15% FVB and MF1 only 5% . The CFW stocks all contain about 15% FVB . The genome-wide results are similar , except the overall contribution of 129 is closer to the other classical inbred strains . These results confirm that haplotypes in outbred colonies are predominantly the same as those found in classical laboratory inbred strains and suggest outbred stocks originated from mice genetically similar to inbred strains . The haplotype analysis might be subject to SNP ascertainment bias as only variants segregating among inbred strains were genotyped . Furthermore , ancestral haplotype reconstruction always finds representations of the outbreds' genomes as mosaics of a given set of inbreds; it does not test if the ancestral hypothesis is true in general , nor whether the set of founders is optimal in the sense of explaining the genome structure of outbred mice with the fewest recombinants and inbred strains . However , the ancestral hypothesis would be refuted if many SNPs segregated within the stocks that are not found in inbred strains . Colonies with high rates of these private alleles will be less suitable for genome-wide association studies . We assessed how many SNPs , missing in laboratory inbred strains , are present in the outbred colonies . We amplified and sequenced 22 fragments of about 1 . 2 Kb , from eight regions in a 5 Mb region previously sequenced on mouse chromosome 1 [19] and from a further 14 regions within the four QTLs described above . We sequenced 12 animals from three populations ( HsdWin:CFW-NL , Crl:CFW ( SW ) -US_K71 and HsdWin:NMRI-NL ) , 12 wild mice ( DNA provided to us by François Bonhomme , University of Montpellier ) and 10 classical inbred strains ( A/J , AKR/J , BALB/cJ , C3H/HeJ , C57BL/6J , CBA/J , DBA/2J , LP/J , I/LnJ and RIII/DmMobJ ) . We identified 120 SNPs ( Table S2 ) . Wild mice have an average of one SNP every 200 bp but this rate varies between colonies: HsdWin:CFW-NL and Crl:CFW ( SW ) -US_K71 have a frequency of one SNP every 350 bp , whereas HsdWin:NMRI-NL has one SNP every 520 bp . We compared this set with SNPs detected by whole genome re-sequencing of 13 inbred strains that are not wild-derived ( 129P2 , 129S1/SvImJ , 129S5 , A/J , AKR/J , BALBc/J , C3H/HeJ , C57BL/6N , CBA/J , DBA/2J , LP/J , NOD and NZO http://www . sanger . ac . uk/resources/mouse/genomes/ ) . We found three novel variants ( rate 2 . 5% ) in Crl:CFW ( SW ) -US_K71 and only one in HsdWin:CFW-NL and HsdWin:NMRI-NL ( rate 0 . 8% ) . The low fraction of novel SNPs suggests that known inbred strains can account for most of the genetic variation in the colonies tested . We took two approaches to determine whether these locus-specific results were representative of the rates of SNPs across the genome . First , we made a single library from four mice from the Crl:CFW ( SW ) -US_P08 colony , and sequenced sufficient short reads ( ∼100 bp ) to cover the complete genome at ten-fold coverage . We mapped all reads to the reference genome using MAQ , called SNPs using SAMtools[20] , [21] and identified a high confidence set of 2 , 554 , 879 SNPs . We again compared SNPs with the 13 inbred strains and found that 3 . 2% of the Crl:CFW ( SW ) -US_P08 SNPs were novel . In the second approach , we sequenced libraries of reduced complexity from pooled DNA samples , obtaining high coverage of a small fraction of the genome ( ∼2% ) . We validated the method by comparing the rate of novel variants found among 36 , 154 SNPs from a Crl:CFW ( SW ) -US_P08 reduced-representation library to the rate obtained from our whole-genome sequence described above . 11 . 7% of the SNPs in the Crl:CFW ( SW ) -US_P08 sequence were novel . Since the false discovery rate is estimated to be 8% [22] , this result implies a novel SNP rate of approximately 4% , consistent with the finding from the whole-genome sequence . We examined four animals from HsdWin:CFW-NL and HsdWin:NMRI-NL colonies and identified 4 , 885 and 16 , 724 SNPs respectively . 3 . 1% of SNPs from HsdWin:CFW-NL and 5 . 7% of SNPs from HsdWin:NMRI-NL were unique ( i . e . not found in the set of SNPs from the inbred strains ) . These percentages are consistent with there being few , or no novel SNPs in the HsdWin:CFW-NL and HsdWin:NMRI-NL colonies . Results from genome-wide sequence thus support the conclusions of the locus-specific results: about 95% of the polymorphisms in the colonies are derived from classical inbred strains ( those not derived from wild mice ) . This result is likely true for other colonies , meaning they contain a relatively limited set of haplotypes , consistent with the reconstructions of each mouse as a mosaic of inbred founders described above ( Figure 3 ) . Our genetic characterization of outbred colonies implies that while the same QTL alleles will segregate in different colonies , their frequencies may vary substantially , so that a QTL segregating in one colony may not be detectable in a second . We assessed the extent of genetic differentiation between colonies and stocks using principal components and Fst distances . We found extensive population differentiation: Fst between populations is 0 . 454 . No single feature , not stock , colony , producer or country of origin , satisfactorily accounted for genetic differentiation . The top panel of Figure 4 shows the relationships between colonies and the middle panel the relationship between stocks ( Figure S2 shows similar results obtained by principal components ) . We then characterized genetic relationships between colonies regardless of stock identity , using methodologies established in studies of human populations: we considered each colony as originating from K unknown ancestral populations and looked at values of K from 2 to 12 using a maximum likelihood method in the program FRAPPE [23] , [24] . Figure 4 ( bottom panel ) shows the results for K = 9 ( see http://www . well . ox . ac . uk/flint-old/outbreds . shtml for all values of K ) . At no value of K were we able to differentiate all stocks . Stocks differ primarily in the proportions of common ancestry , consistent with their descent from inbred strains . There is considerable variation within a stock , largely explained by variation between colonies , as shown for example by CD1 and NMRI stocks . Taken together the dearth of private alleles , we conclude that quantitative differences in allele frequencies are responsible for population differences . From the data in Table 1 we selected three colonies ( Crl:CFW ( SW ) -US_P08 , HsdWin:CFW-NL and HsdWin:NMRI-NL ) suitable for high-resolution mapping . We mapped four QTLs previously detected in HS mice [6]: serum alkaline phosphatase ( ALP ) on chromosome 4 , the ratio of blood CD4+ to CD8+ T-lymphocytes on chromosome 17 , the concentration of high-density lipoproteins ( HDL ) in serum on chromosome 1 , and mean red cell volume ( MCV ) on chromosome 1 . Since HS mice contain alleles derived from eight inbred strains we expect the QTL alleles also to be present in a proportion of the outbred colonies . We tested first whether QTLs could be detected under the assumption that the QTL alleles descend from inbred progenitors . To do so , we used the ancestral haplotype reconstruction described above and mapped QTLs with the HAPPY software package [17] . The detection of QTLs differed markedly between colonies . There was no evidence for association between any markers on chromosome 1 influencing MCV in any colony ( data not shown ) ; single marker association also failed to detect an effect for this phenotype . However , probabilistic ancestral haplotype reconstruction was successful in detecting QTLs for the other three phenotypes . By permutation , we obtained region-specific 5% significance thresholds for HsdWin:NMRI-NL , HsdWin:CFW-NL and Crl:CFW ( SW ) -US_P08 respectively for ALP of 2 . 8 , 2 . 4 and 2 . 6 , for HDL of 2 . 2 , 2 . 9 and 2 . 6 and for CD4+/CD8+ ratio of 3 . 3 , 2 . 9 and 2 . 1 , here expressed as a negative logarithm ( base 10 ) of the P-value ( logP ) . Results shown in Figure 5 exceed these thresholds for each phenotype , but not in every colony . The logP for association with ALP exceeded 2 . 5 for all colonies in a 400 Kb region between 136 . 9 Mb and 137 . 3 Mb on chromosome 4 with considerable variation in the strength of association ( logP of 11 . 5 in HsdWin:NMRI-NL and 2 . 7 in HsdWin:CFW-NL ) . One colony showed strong evidence for association with HDL ( HsdWin:CFW-NL ) with a logP>18; two colonies showed association at the chromosome 17 locus with CD4+/CD8+ T-cell ratio ( HsdWin:CFW-NL and Crl:CFW ( SW ) -US_P08 ) . The percentage of variance explained by each QTL is consistent with effect sizes for these phenotypes found in the HS [6]: 15% for CD4+/CD8+ T-cell ratio ( in Crl:CFW ( SW ) -US_P08 ) , 11% for HDL ( in HsdWin:CFW-NL ) and 18% for ALP ( in HsdWin:NMRI-NL ) . If the QTL alleles are identical in the three colonies , then a single trait effect for each founder strain , independent of colony , should fit the data as well as a model in which each colony had independent effects . This allows us to test , for example , whether the QTL alleles influencing ALP in all three colonies are the same . We found that a model for the single trait effect fitted the data as well as one allowing for independent effect . At the peak of association for ALP the P-value of the partial F test was 0 . 10; for HDL the P-value was 0 . 27 and for CD4+/CD8+ T-cell ratio , 0 . 92 . Our results indicate that the same QTL alleles are present in the different colonies and confirm that QTL mapping is possible on that assumption . We then assessed QTL mapping resolution . We wanted to know if our predictions of gene-level mapping ( based on estimates of haplotype block length ) were upheld in practice . HAPPY mapping results , in Figure 5 , indicate a region of over 1 Mb likely to contain each QTL . While this is a smaller region than observed in HS outbred mice ( where the mean size of QTL intervals is about 3 Mb ) it is larger than suggested by the mean LD decay radius ( of about 1 Mb ) . In fact , the size of the QTL interval is deceptive for two reasons: first , we may not have modeled the descent from the correct set of progenitors; second , in the absence of fine-scale recombination data , HAPPY mapping assumes a uniform genetic map , without hotspots , so that the localization is relatively imprecise . We resorted therefore to using single marker analysis and considering the LD structure of each region to determine the most likely position of the QTL ( Figure 5 ) . Analysis of the ALP QTL revealed in all colonies a large region of linkage disequilibrium extending from 136 . 7 to 137 . 3 Mb , consequently limiting mapping resolution . The region contains an alkaline phosphatase gene ( Akp2 ) at 137 . 3 Mb , but also an additional 9 genes . Mapping the QTL on chromosome 1 for HDL identified two peaks: rs13476237 at 173631526 and rs3709584 at 173177625 ( Figure 5 ) . In the colony showing association ( HsdWin:CFW-NL ) , r2 between these markers is low ( 0 . 21 ) and conditioning on the first marker failed to remove the effect attributable to the second ( F = 15 , df = 2 , 210 , logP = 6 . 1 ) . These results indicate that two separate effects contribute to the variation in HDL , one co-localizing with Apoa2 , already known to be involved in this phenotype , and the other over a region containing two genes , Cd48 and Slamf1 , neither previously implicated in the regulation of HDL levels . On chromosome 17 , we found a single peak of association for CD4+/CD8+ ratio at 34 . 49 Mb ( rs33573309 ) . Association with this marker is strongest in the Crl:CFW-US_P08 colony; r2 between rs33573309 and rs33699857 at 34550471 is 0 . 97 , but drops to less than 0 . 3 elsewhere , delimiting a region of 60 Kb containing four genes ( Figure 5 ) . Only two of these genes show a strong signal in the joint analysis . BC051142 ( a . k . a . Tesb ) is a testis-expressed EST of which little is known; on the other hand , H2-Ea encodes the alpha chain of the MHC class II Eαβ heterodimer , one of the two complexes which govern the selection and survival of CD4+ T cells , and is thus a highly plausible candidate . A number of mouse strains carry a null mutation of H2-Ea , most often through a 650 bp deletion in the promoter region [25] , [26] , and this deletion is tagged by rs33699857 . We confirmed that the promoter deletion is present in the Crl:CFW ( SW ) -US_P08 by examining reads from the whole-genome sequence lying between 34 , 485 , 333 and 34 , 483 , 847 bp ( http://www . well . ox . ac . uk/flint-old/outbreds . shtml ) . We designed primers to amplify across the deletion and tested for its presence in mice from HsdWin:NMRI-NL and the Crl:CFW ( SW ) -US_P08 colonies ( Figure 6 ) . We performed a complementation analysis to test H2-Ea , measuring CD4+/CD8+ ratios in mice in which the H2-Ea null mutation was complemented by introduction of the Eα16 transgene , which drives normal expression of Ea protein with the normal distribution [27] . As illustrated in Figure 6 , the presence of the transgene led to an increased representation of CD4+ cells relative to transgene-negative Eanull littermates , confirming the assignment . This increase was present in both the thymus ( single-positive mature thymocytes ) and spleen , indicating that the variation most likely affects positive selection and lineage commitment of CD4+ T cells .
Commercially available outbred mice are used primarily by the pharmaceutical industry for toxicology testing , on the assumption that they model outbred human populations , a view supported by limited genetic surveys [28] . In fact , very little is known about their genetic architecture and assumptions about the combined effects of fluctuating allele frequencies ( due to genetic drift ) and lack of genetic quality control have led some to argue against their use in genetic investigations [29] , [30] . Our catalogue of the genetic structure of commercially available populations makes a systematic evaluation possible for the first time . Our systematic evaluation of their genetic architecture reveals three important features . First , variation between colonies is large . Fst , a measure of variation within and between populations , is 0 . 454 ( in contrast , human populations values are typically less than 0 . 05 [31] ) . The source of this variation is not straightforward . Stock names ( such as NMRI or CD1 ) do not account for it , nor does the supplier , or the country of origin . While some stocks , such as TO and MF1 , do indeed have a unique genetic ancestry , many do not . Two likely causes are genetic bottlenecks during colony formation and genetic contamination . Thus , ICR colonies from Harlan and CD-1 colonies from Charles River Laboratories cluster together ( Figure 2 ) , having experienced a single bottleneck during their creation ( Figure 1 ) . Gene flow appears to have occurred between a number of stocks , as for example between CFW ( HsdWin:CFW ) and NMRI ( HsdWin:NMRI ) colonies of Harlan . Both were bred at the Winkelmann Versuchstierzucht GmbH & Co and could easily have been mixed . A similar story probably explains the close genetic relationship between RjHan:NMRI and RjOrl:Swiss . Apart from breeders' interventions , colony genetic architecture is stable over time . Mouse colonies are often believed to behave very much like finite island populations , so that , except for imposed bottlenecks ( as happened with the HsdOla:MF1-UK ) or the forcible introduction of new alleles , genetic variation will depend on the effective population size ( Ne ) . Assuming random mating , the time required for a neutral allele to go to fixation in a population , and hence to reduce heterozygosity , is approximately equal to four times Ne . Given that so many colonies are maintained with effective population sizes of many thousands , colony genetic architecture should be stable . Consistent with this view , our analyses of five colonies over two years found little evidence for changes in allele frequencies and LD values . One important caveat is the introduction by some breeders of systems to maintain heterozygosity by periodically crossing the colony to animals taken from a much smaller population , using a protocol called IGS ( International Genetic Standard [32] ) . In consequence , a small number of chromosomes are distributed widely throughout the population , introducing large regions of linkage disequilibrium that significantly reduce mapping resolution . Colonies subject to this protocol become useless for high-resolution genetic mapping ( documented in Table 1 ) . Second , the number of alleles segregating in colonies is relatively limited ( compared to a wild population ) . Almost all of the genetic variants can be found in classical laboratory strains . Both locus-specific and genome-wide sequencing support this conclusion and haplotype reconstruction demonstrates how variants in the outbreds can be modeled as descending from inbred progenitors . Third , in terms of mapping resolution , no mouse colony is comparable to a human population . Using an LD criterion , the best mapping resolution in any colony is at least twice that obtainable in human populations . Applying the same definition of a haplotype LD block as used in human LD studies , we found an average block size in three colonies of approximately 60 Kb . By contrast , in African populations average block length is 9 Kb , and 18 Kb in European populations [33] . These observations have important implications for the use of commercial outbreds for genetic mapping . First , the predominance of SNPs from classical inbred strains means that arrays designed using those SNPs , such as the Affymetrix 600K Mouse Diversity Array [16] , will capture the majority of genetic variation . Second , the extent of LD means that genome-wide coverage can be obtained with fewer SNP than in highly outbred and genetically heterogeneous populations: using 2 markers to tag each block and assuming an average block size of 50 Kb less than 200 , 000 markers will capture the majority of the variation in the genome , so the Affymetrix 600K Mouse Diversity Array [16] will be adequate . Third , resolution will fall short of gene level in some regions . But , since LD structure differs between colonies , high resolution mapping of a locus may be possible in one colony , but not in another – no single colony is ideal . However , mapping resolution is not the only useful measure of a colony's suitability for GWAS . Another critical measure is allele frequency . Large numbers of rare variants contributing to phenotypic variation in a population will make the trait difficult to map using standard GWAS designs . In this regard , our data reveal a favorable situation: QTL mapping , assuming a common set of founder strains , shows that the QTLs replicate between stocks in a consistent manner . These findings indicate that quantitative differences in allele frequencies , rather than the existence of private alleles , are responsible for the population differences . Furthermore , the limited sequence diversity means it is possible to impute the sequence of any commercially available mouse from the sequences of inbred strains . Thus , the full catalogue of sequence variation in a stock could be obtained by sequencing the inbred strains presumed to be founders for it , and genotyping the stock at a skeleton of SNPs . Therefore , we should be able to detect the effect of all variants , a situation that has so far eluded studies in completely outbred populations . Seen in this light , the relatively high degree of genetic differentiation between colonies becomes an advantage . The various genetic architectures available , with variation in QTL frequencies , LD extent and the position of LD blocks , mean that mapping in multiple populations will enable new strategies for gene identification in complex traits . Importantly , we have shown that , at least in the QTLs examined here , the same alleles contribute to variation in different colonies , so that when mapping progress stalls in one stock , another can be used in its stead . As a proof of principle , we have demonstrated the advantages of mapping in different colonies by detecting the same QTL influencing CD4+/CD8+ ratio and were able to refine this mapping to the gene level , transgene complementation helping to establish as the causal change the deletion in the H2-Ea promoter , a loss-of-function mutation that has long been fixed and segregates widely in the Mus species [26] . A strong genetic influence on the CD4+/CD8+ ratio in mice and human has long been known , predominantly reflecting the efficacy of positive selection [26] . Since MHC class-II molecules such as Ea condition the thymic selection of CD4+ T cells , they are thus highly plausible candidates . The homologous MHC class-II region ( HLA-DR ) has recently been shown to influence CD4+/CD8+ ratio in human blood [34] , providing cross-species validation of our result and an example of how results from mice can inform human genetic studies . A variety of resources are available for mapping complex traits in mice , each with its own advantages . The choice of which to use depends on the researcher's aims . We advocate commercial outbreds as a resource for finding genes . In some circumstances , as we have shown , it is possible to go from genetic association to a gene in a single step . The sequence variants in commercial outbred colonies are almost solely those present in classical laboratory strains , resulting in three advantages . First , it provides low LD: the colonies do not depend on recombinants accumulated since their foundation . Second , the relatively low genetic diversity increases power to detect a QTL , provided that it segregates , because there will be fewer QTLs overall . Simply put , in a population with ten variants the relative contribution of each is ten times the contribution of each locus in a population with 100 variants . Third , phenotypes known to show heritable variation among the classical laboratory inbred strains will show heritable variation within the outbred colonies . However , the relatively limited genetic diversity of the outbred colonies means that they do not model a fully outbred population; nor can they be used to assess the effect of all variants present in mouse populations . The colonies contain a relatively small subset of that variation . They are likely to have “blindspots” where little functional variation segregates . The creation of the Collaborative Cross ( CC ) , a large set of recombinant inbred lines derived from genetically diverse progenitors [7] , provides access to a more complete catalog of variation [35] , and also has the advantage of allowing researchers to interrogate the same genotype multiple times and hence accumulate an increasingly rich understanding of the relationship between genotype and phenotype . However , it is not yet clear to what extent CC animals will provide high-level mapping resolution , although simulations suggest it will be of the order of 1–2 Mb [36] . Assuming an investigator decides to use an outbred colony , which is the best to choose ? For single locus assays , for example attempting to refine a locus identified in a cross between two inbred strains , the choice will depend on whether the appropriate alleles are segregating at the locus , and this can be assessed by haplotype reconstruction from genotype data . The extent of genetic diversity between colonies ( ten times that between different human populations ) indicates that an appropriate colony will be found . However , genome-wide data will be needed from all colonies to enable a comprehensive assessment . For genome-wide association , which we think is the most likely use of the outbreds , choice will be guided by the genetic characterization provided here , most simply summarized by low LD , coupled with high mean minor allele frequency . Depending on the phenotype , an additional criterion may be the likelihood that heritable variation is present in a given colony; this could be determined either by family studies carried out with animals from the colony , or by determining whether strains contributing to the colony show phenotypic differences from published data , for example from the phenome project [37] . Our work , demonstrating the utility of the outbreds , is a starting place for ranking colonies according to their utility for genetic mapping . As costs fall , we anticipate that detailed characterization based on genomic sequence will become available and permit informed choices on the use of the colonies for genetic studies of complex traits in mice .
We contacted ten commercial providers of outbred strain of mice , including Harlan Sprague Dawley ( Hsd ) , Charles River Laboratories ( Crl ) , Taconic Farms ( Tac ) , Centre d'Elevage R . Janvier ( Rj ) , Ace Animal ( Aai ) , B&K Universal ( Bk ) , Hilltop Laboratory Animals ( Hla ) , Research and Consulting Company ( Rcc ) , Scanbur ( Sca ) , Simonsen Laboratories ( Sim ) , and we collected on average 48 tail samples from unrelated mice from each colony ( Table 1 ) , representing 90% of all commercially available colonies of outbred mice . 48 unrelated individuals from six colonies were resampled at least one year after the initial collection . We also collected samples from control populations: 109 Collaborative Cross ( CC ) mice provided by Fuad Iraqi ( Tel-Aviv University ) , 96 DNA samples of wild mice caught in the vicinity of Tucson ( Arizona ) provided by Michael Nachman , 12 unrelated Heterogeneous Stock ( HS ) DNA samples from our laboratory and 94 inbred strains purchased from the Jackson Laboratory . DNA was extracted from tail snips using a Nucleopure Kit ( Tepnel , UK ) . DNA quality and quantity was assessed using UV spectrophotometry ( Nanodrop ) and 0 . 8% agarose gel electrophoresis . We designed extension and amplification primers for 351 SNPs using SpectroDESIGNER . Oligonucleotides were synthesized at Metabion ( Germany ) ( Table S1 ) . We used the Sequenom MassARRAY platform for genotyping these 351 SNPs over 4 , 000 DNA samples and SpectroTYPER Version 4 . 1 for data analysis . The resulting genotypes were then uploaded into an Integrated Genotyping System ( IGS ) [38] . We also obtained genome-wide SNPs genotyping data for six colonies using Affymetrix arrays . Three populations were genotyped using the 600K Affymetrix Mouse Diversity Array [16] . Three more populations were analysed using a precursor to this array , a gift from Mark Daly . DNA was prepared , hybridized and genotypes obtained following the manufacturer's protocol . Data were stored in a relational database designed to manage genotypes and phenotypes [38] . Analyses were run either using software from the authors of each test or were implemented in R [39] . We tested Hardy-Weinberg Equilibrium ( HWE ) by the exact test [11] for all populations separately . Heterozygosity for each marker was calculated using PLINK [12] . We inferred individual ancestry proportions using a maximum likelihood method [24] in the program FRAPPE ( http://www . fhcrc . org/labs/tang/ ) . We used parameters described in [23] , running the program for 10 , 000 iterations , with pre-specified cluster numbers , from K = 2 to 12 . We found that independent runs yielded consistent results , with few additional clusters emerging after K = 9 . However , it should be noted that given the small set of markers and the inclusion of markers in LD , our estimates of ancestry are likely to be biased . Fst for all pairs of populations was calculated using the FDIST2 program [40] , [41] ( http://www . rubic . rdg . ac . uk/~mab/software . html ) . An identity-by-state ( IBS ) matrix for all individuals was calculated using PLINK [12] . Principal component analysis was carried out using this IBS matrix . Genetic relationships were represented as a tree using agglomerative clustering implemented in R [39] . Haplotype blocks were estimated using PLINK [12] which implements the block finding algorithm found in HAPLOVIEW [42] . We used Primer3 to design oligonucleotide primers and carried out PCR reactions with Hotstar Taq obtained from Qiagen . Each 50 µl PCR contained 50 ng of genomic DNA , 1 Unit of HotStar Taq , 5 pmol of forward and reverse primers ( synthesized at MWG Biotech , Ebersberg , Germany ) , 2 mM of each dNTP , 1× HotStar Taq PCR buffer as supplied by the enzyme manufacturer ( contains 1 . 5 mM MgCl2 , Tris-Cl , KCl and ( NH4 ) 2SO4 , pH 8 . 7 ) and 25 mM MgCl2 ( Qiagen ) . We ran the PCR reactions using a Touchdown ( TD ) approach . The temperature profile consisted of an initial enzyme activation at 95°C for 15 min , followed firstly by 13 cycles of 95°C for 30 sec , 64°C for 30 sec and 72°C for 60 sec , secondly by 29 cycles of 95°C for 30 sec , 57°C for 30 sec and 72°C for 60 sec , and finally by an incubation at 72°C for 7 min . PCR products were purified in a 96-well Millipore purification plate and resuspended in 30 µl of H2O . Two sequencing reactions were prepared for each DNA sample , one with the forward primer and one with the reverse primer using 50 ng DNA . The sequencing reaction consisted of an initial denaturation stage at 95°C for 1 min , followed by 29 cycles of 95°C for 10 sec , 50°C for 10 sec and 60°C for 4 mins . The PCR reagents were then removed from solution by an ethanol precipitation in the presence of sodium acetate . All sequencing reactions were run out on an ABI3700 sequencer and assembled by using phred/phrap [43] . Consed was then used for editing and visualisation of the assembly [44] . The libraries were prepared from 3–5 µg sample genomic DNA following the Illumina standard genomic library protocol up to the ligation step , where a modified adapter was used . The resulting constructs were digested overnight at 37°C with 20 units high-concentration HindIII restriction enzyme ( New England Biolabs ) in a volume of 50 µl . The digested libraries were purified on Qiagen MinElute columns . A complementary biotinylated adapter was ligated to the sticky ends before selecting the fragments of 200 to 500 bp on a 2% agarose gel . The constructs with a HindIII-specific adapter were purified using Streptavidin magnetic beads ( Invitrogen ) following the manufacturer's instructions . The beads were finally resuspended in 25µl 10mM Tris pH8 , of which 12 . 5 µl were used for the final PCR amplification ( 15 cycles ) using specific amplification primers and Phusion DNA polymerase ( Finnzymes ) . The resulting libraries were verified by TOPO cloning and sequencing before running them on an Illumina Genome Analyzer IIx for 38 cycles . Libraries of reduced complexity for SNP discovery were made from pooled DNA samples . Genomic DNA was subject to complete HindIII restriction enzyme digestion and ligation to linkers . Libraries were then sequenced on an Illumina Genome Analyzer IIx . Since it has been shown that this method has a false SNP discovery rate of about 8% [22] , we used three additional filtering criteria to increase confidence in SNP calls . First , following reports that SNPs falling at the ends of reads were unreliable , SNPs within three bases of the end of a read were discarded [22] . Second , SNPs that did not map to within 32 bases of a known HindIII restriction site were also discarded . Third , SNP detection is affected by read depth: where the coverage is less than fivefold , the proportion of novel SNPs rises to over 10% . We only report SNPs where the coverage is greater than tenfold . We analysed 200 animals from three colonies: Crl:CFW ( SW ) -US_P08 , HsdWin:CFW-NL and HsdWin:NMRI-NL . Blood samples were taken from a tail vein and we performed assays for serum alkaline phosphatase ( ALP ) , ratio of CD4+ to CD8+ T-cells , concentration of high-density lipoproteins ( HDL ) in serum and mean red cell volume using published protocols [45] . Where necessary , phenotypes were transformed into Gaussian deviates . Covariates ( such as gender , age , experimenter , time ) that explain a significant fraction of each phenotype's variance with ANOVA P-value<0 . 01 were included in subsequent statistical analyses . We use two mapping methods: a single point analysis of variance of each marker and a multi-point method . The single point method was implemented using linear modeling in R; the multipoint method is implemented in the R package HAPPY [17] . Region-wide significance levels are estimated by permuting the transformed phenotype values 1 , 000 times .
|
We show that commercially available mice are a resource for detecting single genes by genome-wide association . We surveyed 66 populations and identified those with properties conducive to high-resolution mapping . Importantly , we show that the same alleles contribute to variation in different colonies , so that when mapping progress stalls in one colony , another can be used in its stead . As a proof of principle , we detect the same QTL in different colonies influencing CD4+/CD8+ ratios and refine this mapping to the gene level . We show that a deletion in the promoter of H2-Ea is the molecular change that strongly contributes to setting the ratio of CD4+ and CD8+ lymphocytes . Our results make it possible for geneticists to make informed choices on the use of colonies for genome-wide association studies of complex traits in mice .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/animal",
"genetics",
"genetics",
"and",
"genomics/gene",
"discovery",
"genetics",
"and",
"genomics/genetics",
"of",
"the",
"immune",
"system",
"genetics",
"and",
"genomics/complex",
"traits",
"genetics",
"and",
"genomics/disease",
"models"
] |
2010
|
Commercially Available Outbred Mice for Genome-Wide Association Studies
|
Movement of the transcription machinery along a template alters DNA topology resulting in the accumulation of supercoils in DNA . The positive supercoils generated ahead of transcribing RNA polymerase ( RNAP ) and the negative supercoils accumulating behind impose severe topological constraints impeding transcription process . Previous studies have implied the role of topoisomerases in the removal of torsional stress and the maintenance of template topology but the in vivo interaction of functionally distinct topoisomerases with heterogeneous chromosomal territories is not deciphered . Moreover , how the transcription-induced supercoils influence the genome-wide recruitment of DNA topoisomerases remains to be explored in bacteria . Using ChIP-Seq , we show the genome-wide occupancy profile of both topoisomerase I and DNA gyrase in conjunction with RNAP in Mycobacterium tuberculosis taking advantage of minimal topoisomerase representation in the organism . The study unveils the first in vivo genome-wide interaction of both the topoisomerases with the genomic regions and establishes that transcription-induced supercoils govern their recruitment at genomic sites . Distribution profiles revealed co-localization of RNAP and the two topoisomerases on the active transcriptional units ( TUs ) . At a given locus , topoisomerase I and DNA gyrase were localized behind and ahead of RNAP , respectively , correlating with the twin-supercoiled domains generated . The recruitment of topoisomerases was higher at the genomic loci with higher transcriptional activity and/or at regions under high torsional stress compared to silent genomic loci . Importantly , the occupancy of DNA gyrase , sole type II topoisomerase in Mtb , near the Ter domain of the Mtb chromosome validates its function as a decatenase .
Translocation of the transcription machinery along the duplex DNA molecule causes axial rotation of the DNA duplex relative to the RNAP complex . Consequently , the structural complexity , topography and helical tension on the chromatin result in waves of positive supercoils in front of an advancing polymerase and accumulation of negative supercoils behind . The twin supercoiled domains [1] thus generated and the associated torsional strain are relieved by the action of different classes of topoisomerases [2] . Genetic and in vitro studies in Escherichia coli have implicated that negative supercoils are acted upon by relaxases i . e . Topo I ( Type I ) and Topo IV ( Type II ) while the positive supercoils are removed by the activities of DNA gyrase and Topo IV [2–4] . In parallel , the resolution of sister chromatids is carried out by stronger decatenation activity of Topo IV and assisted by a rather meager activity of DNA gyrase and Topo III [5–7] . Earlier studies aimed to unravel the influence of supercoiling and topoisomerases in the transcription process revealed their close connection [5 , 8–10] . A deficiency in the function of topoisomerase activities affects the supercoiling balance [11] resulting in inefficient transcription [12 , 13] . The genome-wide gene expression studies have linked the respective relaxation and supercoiling activities of Topo I and DNA gyrase in the modulation of gene expression profile , highlighting the global regulatory roles of topoisomerases [14–16] . However , the mechanism of their in vivo influence on gene expression is not understood due to inadequate information about their interaction with the genome . Direct demonstration of their occupancy on different genomic territories is not yet carried out which is a pre-requisite in determining their impact on supercoiling gradients generated in transcriptionally active regions spanning whole genome . Although low resolution genome-wide occupancy of DNA gyrase [17] and more recently that of Topo IV has been revealed in E . coli [18] , genomic landscape of both DNA gyrase and Topo I during transcription has not been elucidated on any bacterial genome so far . In order to achieve such comprehensive view of topoisomerase dynamics , function and factors affecting their distribution , we have deciphered in vivo genome-wide binding profile of both classes of topoisomerases together taking the advantage of non-redundancy of topoisomerases in Mycobacterium tuberculosis ( Mtb ) genome . Analysis of the Mtb genome [19] indicated the presence of a single Type I ( DNA topoisomerase I or Topo I ) and a single Type II enzyme ( DNA gyrase ) . Notably , the genome is devoid of any dedicated decatenase gene unlike E . coli and many other eubacteria . The presence of a single relaxase and a sole supercoiling enzyme in Mtb imposes on them the entire burden of managing the topological perturbations occurring during all DNA transaction processes . Accordingly , our earlier studies revealed that mycobacterial gyrase is a strong decatenase [20] and Topo I relaxation activity is stimulated by the nucleoid associated protein HU [21] . The absence of additional topoisomerases with overlapping or back-up function in Mtb ( unlike in E . coli ) was exploited in the present study to probe the in vivo topology-transcription dynamics . We have carried out the first genome-wide Chromatin Immunoprecipitation ( ChIP ) of both Topo I and DNA gyrase followed by sequencing ( Seq ) of the bound target DNA . ChIP-Seq analysis revealed the co-localization of Topo I , DNA gyrase and RNAP throughout the Mtb genome . We show that the distribution landscape of Topo I and DNA gyrase in the genome is driven by transcription induced supercoiling . We also demonstrate the recruitment of DNA gyrase at Ter region of the chromosome to carry out decatenation function .
Transcription generates waves of positive supercoils downstream and negative supercoils upstream of moving RNAP ( Fig 1A ) . The catalytic activities of Topo I and DNA gyrase ensure the removal of the resultant negative and positive supercoils , respectively . To investigate the interaction of topoisomerases ( both Topo I and gyrase ) with the Mtb genome , the global occupancy of Topo I and DNA gyrase together with RNAP was monitored by ChIP-Seq . Visualization of ChIP-Seq data of RNAP , Topo I and DNA gyrase on UCSC genome browser ( Fig 1B and S2 Fig ) revealed the distribution of Topo I , Gyrase and RNAP signals throughout the genome . As reported previously , the strongest RNAP signal was present at the putative promoter regions of the TUs [22] . Comparison of Topo I and DNA gyrase signals with the RNAP signals suggested their co-localization throughout the Mtb genome . Based on the peak detection criteria ( enrichment ratio; ER>2 . 0 ) , 190 peaks for Topo I and 73 peaks for DNA gyrase were identified , distributed throughout the genome . The classification of the loci enriched with Topo I and DNA gyrase suggested that the peaks were distributed in various functional groups ( S3 Fig ) . Importantly , the genes occupied by topoisomerases were transcriptionally active [22] . The MEME suite [23] was used to identify consensus sequences for Topo I and DNA gyrase from ChIP-Seq signals by scanning 300 bp encompassing the peaks ( S4A Fig ) . The gyrase binding motif was found to be GC-rich compared to the Topo I motif , indicating that the Topo I motif was relatively more prone to DNA melting . FIMO ( Find Individual Motif Occurrences ) detected 5299 and 9997 motifs ( p>0 . 0001 ) for Topo I and DNA gyrase respectively . The genome-wide distribution of Topo I and DNA gyrase binding motifs indicates that these enzymes can potentially bind throughout the genome ( S4B Fig ) . The UCSC genome browser view of the binding profile of Topo I , RNAP and DNA gyrase showed that at several places the peaks of these proteins were co-localized suggesting the association of topoisomerases with the RNAP enriched genes ( Fig 1B ) . To analyse further , the enrichment ratio ( ER ) across the protein coding genes were calculated by taking the ratio of mean read counts of ChIP-samples and Mock IP ( control ) as described [22] . RNAP enriched ( ER>2 ) genes were monitored for Topo I and gyrase enrichment ( ER>2 ) . Data depicted in Fig 2A indicate that out of 615 RNAP enriched genes , 75 . 6% and 89 . 0% genes were also enriched for Topo I and DNA gyrase respectively . Moreover , out of 488 Topo I enriched ( ER>2 ) genes , 89 . 5% were enriched for RNAP while 88 . 5% gyrase ( ER>2 ) enriched genes were also enriched with RNAP ( Fig 2A ) . The analysis indicated that a majority of the RNAP bound regions exhibit association with the Topo I/DNA Gyrase and vice versa . In order to delineate the global distribution profile of Topo I , RNAP and gyrase across the TUs , all the transcriptionally active TUs ( >1 kb length ) with identified TSS were selected and the mean read counts around the TSS of all TUs were calculated . The start site ( TSS ) was taken as a reference point and regions 500 bp upstream and 1500 bp downstream were included to generate the binding profile for Topo I , RNAP and DNA gyrase . The mean of the read counts was plotted to generate a single profile representing the global TU . The strongest peak of RNAP was located around the TSS . The position of the RNAP matches with the previously reported promoter proximal peak [22] . On the same TU , the occupancy signals of Topo I and DNA gyrase were found to be in a close proximity , overlapping to the RNAP peak ( Fig 2B ) . Importantly , around 500 bp downstream of the TSS stronger accumulation of DNA gyrase is apparent which is indicative of its role in removal of excessive positive supercoils generated downstream of the advancing RNAP . Further , analysis of the Topo I and DNA gyrase profile on the TUs depleted of RNAP signals ( ER<1 ) did not exhibit Topo I and gyrase peak around the TSS ( S5A Fig ) indicating that RNAP activity at TUs facilitates the recruitment of the two topoisomerases . The transcriptional activity of RNAP would induce torsional stress on the active TUs , which may facilitate the recruitment of topoisomerases . Co-localization of Topo I and gyrase signals genome-wide with RNAP peaks and TUs suggested that topoisomerases were predominantly recruited to active TUs . Investigation of Topo I and DNA gyrase signals on RNAP enriched genes ( Fig 2A ) further hinted their stronger association with transcriptionally active genes . To get the global view of the relationship between gene transcription and topoisomerase occupancy , TUs were segregated into two categories based on their gene expression profile ( based on Reads Per Kilobase of transcript per Million mapped reads i . e . RPKM values ) , high expression ( RPKM>3 ) and low expression ( RPKM<1 ) . A total of 342 protein coding TUs ( excluding rRNA and tRNA clusters ) from each category were taken and the average occupancy profile of Topo I , DNA gyrase and mock IP was generated ( Fig 3 and S5B Fig ) . From the averaged profiles , it is apparent that both Topo I and DNA gyrase associate strongly with highly expressed genes compared to the genes with low transcriptional activity ( Fig 3 ) . Negative control ( mock IP ) profile did not exhibit any appreciable differences in the enrichment on high and low expression genes ( S5B Fig ) indicating the specific preferential enrichment of topoisomerases on transcriptionally active genes . To obtain independent confirmation , the rv3852 gene [24] was cloned under an inducible acetamidase promoter [25] and introduced into M . smegmatis cells which lack the ortholog . The advantages of using this experimental set up were ( a ) the transcription of the TU can be induced strongly by the addition of acetamide , ( b ) the higher level of transcription achieved in this system would lead to accumulation of supercoils , ( c ) the absence of the orthologous gene allows monitoring of the occupancy of the topoisomerases and RNAP on the transcriptionally active supercoiled template without any interference from the genomic copy . Exponential phase cells were induced to activate transcription of rv3852 ( Fig 4A ) and the occupancy of RNAP , Topo I and DNA gyrase was monitored by ChIP-qPCR . From Fig 4B , it is evident that upon activation of transcription , topoisomerases were recruited to the rv3852 TU . To confirm the recruitment of topoisomerases to transcriptionally active DNA template , we have selected two TUs rrS and Rv1303 TUs , which were found to be transcriptionally active . On these TUs functional gyrase association was monitored by trapping the active gyrase molecules with moxifloxacin followed by monitoring its enrichment around promoter region as well on gene body . At both the regions moxifloxacin trapped gyrase-DNA cleavage complexes were observed ( Fig 4C ) . To establish that the functional gyrase-DNA complex formation is governed by transcription , Mtb cells were treated with sub-lethal concentrations of Rifampicin to inhibit transcription . Inhibition of transcription led to the reduction in ChIP-enrichment of DNA gyrase validating the model that transcription governs the recruitment of topoisomerases ( Fig 4D ) . In vitro , validation of the hypothesis was carried out by monitoring the recruitment of topoisomerases on the transcriptionally active DNA template . Recruitment of both DNA gyrase and Topo I on plasmid DNA was significantly higher when transcription was activated by the addition of RNAP and NTPs compared to the reaction where transcription did not occur ( S6 Fig ) . Overall , ChIP-Seq analysis , in vitro transcription and in vivo gene expression studies confirm the recruitment of topoisomerases to transcriptionally active DNA template . To study the interaction of topoisomerases with supercoiled domains , the occupancy of topoisomerases was determined on the intergenic regions of the convergent and divergent gene pairs . In the convergent gene pairs , the head-on movement of RNAP would induce positive supercoils downstream of both genes while movement of RNAP in the opposite direction would result in the accumulation of negative supercoils upstream of divergent gene pairs . Thus , the intergenic regions between convergent and divergent gene pairs would harbor distinct supercoiled domains generating the sites for the recruitment of respective topoisomerases . Convergent and divergent gene pairs were segregated and the occupancy of Topo I and DNA gyrase was monitored in the intergenic regions ( Fig 5A and 5B ) . In every gene pair at least one of the TU was transcriptionally active ( RPKM>1 ) . The analysis revealed higher occupancy of DNA gyrase in the intergenic regions between convergent genes ( Fig 5A ) whereas a higher density of Topo I was seen in the intergenic regions between the divergent genes ( Fig 5B ) . The twin-supercoiled domain model also predicts the accumulation of positive supercoils downstream of the region where transcription termination occurs and RNAP dissociates [26] implying that DNA gyrase should be recruited to the end of TUs . In order to test this hypothesis , employing RNA-Seq data [22] we have identified the putative transcription termination site ( TTS ) of the highly expressed and lowly expressed TUs . Around the TTS , 200 bp upstream and downstream regions were scanned for the occupancy of Topo I or DNA gyrase . From the data presented in Fig 5C , it is apparent that the occupancy of DNA gyrase is high downstream to TTS . Notably , the higher enrichment of DNA gyrase on the highly expressed genes compared to the lowly expressed genes suggested that the enrichment resulted from the accumulation of torsional stress generated due to active transcription . In contrast , Topo I enrichment peak was seen upstream of the DNA gyrase peak around TTS . The pattern of Topo I and DNA gyrase occupancy at the end of TUs establishes the existence of twin-supercoiled domain in vivo . Thus , overall , it is evident that Topo I and DNA gyrase are distributed on active TUs being recruited to negatively and positively supercoiled domains , respectively . In terms of transcriptional activity , the bacterial genome is non-uniform since some regions of the genome have higher transcriptional activity compared to others [22] . This may influence genome-wide supercoiling and thus the distribution of DNA topoisomerases across the genome . In order to explore their genomic distribution , the Mtb genome was divided into 9 bins ( ~ 500 Kb ) and the associated read counts were averaged for each bin ( Fig 6A ) . This revealed the non-random occupancy profile of topoisomerases across the genome . The association of DNA gyrase and Topo I was found to be higher at the OriC region of the genome compared to the other parts indicating enhanced torsional stress there ( Fig 6B and 6C ) . To establish whether higher torsional stress at OriC and thus higher occupancy of the topoisomerases is an attribute of transcription , transcript abundance data were plotted ( Fig 6D ) . The pattern of transcript abundance was found to correlate with the occupancy profile of the topoisomerases thus reconfirming that higher transcriptional activity across the genome recruits topoisomerases . Comparison of the occupancy profile of Mtb DNA gyrase with that of Topo I indicated a stronger occupancy of DNA gyrase near the Ter region of the genome although at OriC both the Topo I and DNA gyrase were abundant ( Fig 6E ) . At the end of DNA replication , daughter DNA molecules are entangled at the Ter domain . These entangled DNA molecules are resolved by the decatenation activity of Type II topoisomerases [27 , 28] . Mycobacterial DNA gyrase has evolved a strong decatenation ability [20 , 29 , 30] in addition to its characteristic supercoiling function . In contrast to Mtb , E . coli DNA gyrase has poor decatenation activity and the decatenation function is carried out by Topo IV [4] . Studies with E . coli [18 , 31] have shown the norfloxacin-mediated entrapment of catalytically active Topo IV at the genomic dif site which is required for the de-catenation function in vivo [18] . The comparative analysis of the binding profile of Mtb and E . coli DNA gyrase [17] showed the recruitment of gyrase from both species to the respective OriC regions ( Fig 6E and 6F ) . However , only Mtb gyrase was specifically recruited at the Ter region ( Fig 6E ) . Sequence analysis of the Mtb gyrase peak at Ter domain indicated the presence of GC rich sequence ( Fig 6G ) . Comparison of this putative dif site with the E . coli dif sequence and other dif sequences [32] did not indicate any conservation with the mycobacterial sequence ( Fig 6G ) . From these data it is apparent that the binding of Mtb DNA gyrase at the Ter region could be associated with its decatenation activity in vivo .
In 1987 , Liu and Wang proposed that the active transcription machine generates two supercoiled domains in the DNA template during transcription [1] . The proposal was based on the rationale that movement of the bulky transcriptional complex of RNAP along a DNA duplex would lead to the rotation of the DNA rather than of RNAP . This model predicted the generation of waves of positive supercoils downstream and accumulation of negative supercoils upstream of the advancing transcription apparatus . Several in vitro studies showed that transcription alters DNA template topology thus supporting the twin-supercoiled domain model [33–35] . Next , various DNA sensors of supercoiling were used to assess the operation of twin-supercoiled domain model in vivo [26 , 36] . The B-to-Z structural transitions of ( CG ) tracts , and the Tn3 and γδ resolution assays were used as supercoiling probes to examine transcription induced local topological changes [26 , 36] . That the transcription induced supercoils are acted upon by Topo I and DNA gyrase to permit the continuation of the process was indicated by the inhibition of DNA gyrase , which led to the accumulation of positively supercoiled DNA [37] , and by mutation of Topo I , which generated negatively supercoiled DNA [38] . Based on these studies , it was proposed that Topo I acts upstream while DNA gyrase functions downstream of the RNAP machinery . However , direct evidence for the interaction of topoisomerases with the twin-supercoiled domains was not available until the present study . The genome-wide occupancy profile of topoisomerases along with RNAP in Mtb described here is the first demonstration of the interaction of topoisomerases with the twin-supercoiled domains in vivo ( Fig 7 ) . Our genome-wide distribution of Topo I and DNA gyrase data and their binding profiles correlate well with that of RNAP indicating that there is a functional association of RNAP and topoisomerases . The interaction of topoisomerases with the transcription machinery may recruit them to the site of transcription where topoisomerase activity is needed to foster transcription . In E . coli , physical association of RNAP with Topo I was revealed [39] . Similarly , in mycobacteria , DNA gyrase was found in a complex with RNAP [40] . Thus , the co-localization of genome-wide peaks of two topoisomerases that carry out two opposing topological reactions , along with RNAP , implicates their in vivo role in the stimulation of transcription . Returning to the twin-supercoiled domain model , it is imperative that the transcriptional activity of RNAP generates supercoils in the vicinity of RNAP , which then has to be removed by the action of topoisomerases suggesting that these three players should work together . Indeed , the co-occupancy of topoisomerases and RNAP on the TUs indicated the co-existence of these proteins . Promoter proximal peaks of RNAP reflect the transcriptional complex , which assembles and pauses at the beginning of genes [22 , 41–43] . Once RNAP has entered the elongation phase , it will have generated negative supercoils upstream and positive supercoils downstream . From the binding profiles , it is apparent that the Topo I occupancy is higher upstream of RNAP while the gyrase occupancy peak was found to be present either along with RNAP or downstream of it within the ORF . Thus , the interaction of topoisomerases with the twin-supercoiled domains on a genome-wide scale is evident . Topoisomerase recruitment was found to be enhanced by transcription induction while the inhibition of transcription resulted in decreased gyrase ( present study ) and Topo IV activity [18] suggesting that the rate of transcription is the major factor governing the chromosomal supercoiling levels . An earlier study in E . coli and Salmonella [11] also indicated the role of RNAP elongation rate in generation of supercoils on the highly expressed genes . Overall , the enhanced association of topoisomerases with the TUs driving active transcription ( which would generate high torsional stress ) further validates the twin-supercoiled domain model . Transcriptome analysis of various yeast topoisomerase mutants demonstrated that the transcript abundance of genes with higher transcriptional activity was specifically reduced suggesting that activities of topoisomerases facilitate the transcription of genes [44] . Since DNA topoisomerases maintain the topological homeostasis inside the cell , perturbation of their activity would lead to changes in genome supercoiling and gene expression thereby highlighting their global regulatory role [14–16] . To carry out the global regulatory functions , topoisomerases interact with DNA and remove the excessive supercoiling thus ensuring balanced supercoiling levels in the cell . The genome-wide distribution of the recognition motifs of topoisomerases implies their interaction throughout the genome . However , the binding profile of Topo I and DNA gyrase indicated their non-random distribution , which is likely a consequence of differential accumulation of torsional stress across the genome . In addition to transcription and replication , various processes involving DNA , such as recombination or chromosome segregation , contribute to the differential distribution of torsional stress . One such epicenter of high torsional stress is the OriC region . The distribution profile of topoisomerases indicates their higher occupancy at OriC compared to the Ter domain . At OriC region , the replication machinery binds , melts the DNA and then initiates replication , which further enhances the torsional stress [45] . In every organism studied so far , higher transcriptional activity occurs at the OriC region [46] . Thus , a combination of the initiation of DNA replication and the higher transcriptional activity at OriC subjects this region to a greater torsional stress requiring the higher activity of topoisomerases . A contrasting landscape emerges at the Ter region when the distribution of topoisomerase there is compared to that of the TUs and OriC . Instead of a near symmetric distribution i . e . Topo I and gyrase occupancy on either side of the replication or transcription complex , only a gyrase peak is found . At the completion of DNA replication , the entangled daughter DNA molecules have to be segregated at or near the Ter domain of the genome [47 , 48] . In E . coli , resolution of daughter catenated molecules is carried out by the action of Topo IV , a dedicated decatenase while E . coli DNA gyrase carries out primarily supercoiling function [7 , 49] . The enhanced binding of Mtb DNA gyrase at the Ter domain seen in the present analysis appears to be a characteristic of Mtb; E . coli DNA gyrase does not show binding to the Ter domain . Our earlier studies revealed that mycobacterial DNA gyrase is a dual function enzyme with efficient decatenation activity apart from its natural supercoiling function [20] . Thus , these data provide the first in vivo evidence for the action of DNA gyrase at the Ter region most likely to carry out its efficient decatenation function required for daughter chromosome segregation . In conclusion , genome-wide ChIP-seq analysis of topoisomerases and RNAP demonstrates the interaction of Topo I and DNA gyrase with the twin-supercoiled domain in vivo ( Fig 7 ) . The global binding profiles of these players demonstrate their co-existence on the TUs with a non-random , organized distribution , correlating with the organization of TUs along the genome and extent of transcriptional activity . Additional experiments to probe the supercoiling of the genome [50] and topoisomerase occupancy by genetic or chemical perturbation of topology would provide deeper insights into the dynamics of supercoiling and the action of DNA topoisomerases in conjunction with the molecular machines engaged in replication and transcription .
Mtb H37Rv or Mtb H37Ra were grown in Dubos Broth ( Difco ) supplemented with Middlebrook albumin dextrose catalase ( ADC ) enrichment and 0 . 05% Tween 80 or on solid Middlebrook 7H11 medium ( Difco ) supplemented with oleic acid-albumin-dextrose-catalase ( OADC ) . Mtb cultures ( 50 ml ) grown to OD600 nm = 0 . 4–0 . 6 were treated with formaldehyde ( final concentration 1% ) and incubated for 10 min at 37°C . Cross-linking was quenched by addition of glycine ( final concentration 125 mM ) . Cells were then collected by centrifugation , washed twice with Tris-buffered saline ( 20 mM Tris-HCl pH 7 . 5 , 150 mM NaCl ) and stored at -80°C . Pellets were re-suspended in 4 ml IP buffer ( 50 mM HEPES-KOH pH 7 . 5 , 150 mM NaCl , 1 mM EDTA , 1% Triton X-100 , 0 . 1% sodium deoxycholate , 0 . 1% SDS , Roche Antiprotease mini ) and sonicated in a water bath sonicator ( Bioruptor , Diagenode ) to shear DNA to an average size of 150–700 bp . Cell debris was removed by centrifugation and the supernatant was used as input sample in subsequent IP experiments . The samples were incubated overnight at 4°C on a rotating wheel with antibodies ( polyclonal anti-GyrA or monoclonal anti-Topo I [51] . An IP experiment without antibody served as negative control ( Mock IP ) . Protein–DNA complexes were immuno-precipitated with 50 µl of Dynabeads Sheep IgG ( Dynal Biotech ) for 2 h at 4°C . The magnetic beads were collected and washed twice with IP buffer , once with IP buffer containing 500 mM NaCl , once with wash buffer III ( 10 mM Tris-HCl pH 8 , 250 mM LiCl , 1 mM EDTA , 0 . 5% Nonidet-P40 , 0 . 5% sodium deoxycholate ) , and once with Tris-EDTA buffer ( pH 7 . 5 ) . IP complexes were eluted from the beads by treatment with 100 µl elution buffer ( 50 mM Tris-HCl pH 7 . 5 , 10 mM EDTA , 1% SDS ) at 65°C for 20 min . Samples were then treated with 2 µl RNase A ( 10 mg/ml ) and cross-links were reversed by incubation for 2 h at 56°C and 6 h at 65°C in 0 . 5x elution buffer containing 2 . 5 µl proteinase K ( 20 mg/ml ) . DNA was extracted twice with phenol chloroform , precipitated and re-suspended in 20 µl of water . ChIP-qPCR was carried out to determine the enrichment of target sites for Topo I and Gyrase using specific primers for the expected targets ( S1 Fig ) . After confirming target enrichment , samples were processed for sequencing . DNA fragments ( 150 to 250 bp ) were selected for library construction and sequencing libraries were prepared using the ChIP-Seq Sample Preparation Kit ( Illumina; San Diego , California , USA; Cat . No . IP-102-1001 ) according to the protocol supplied with the reagents . Prior and post library construction , ChIP products were quantified using a Qubit fluorometer ( Invitrogen; Carlsbad , California , USA ) . One lane of each library was sequenced on the Illumina Genome Analyzer IIx using the Single-Read Cluster Generation Kit v4 and 36 Cycle Sequencing Kit v4 . Data were processed using the Illumina Pipeline Software v1 . 60 . ChIP-Seq analysis was performed using the HTS station pipeline ( http://htsstation . epfl . ch/ ) . Single-end sequence reads generated from ChIP-Seq experiments were aligned to the Mtb H37Rv genome ( NCBI accession NC_000962 . 2 ) using Bowtie [52] with default option in HTS station [53] . Different BAM files were normalized for sequencing depth . At each base position , number of reads mapping to that base was calculated and normalized to the total number of mapped reads [22] . Peaks were analyzed using MACS v . 1 . 4 [54] with parameters ‘‘-bw 200 -m 10100” . Alignment files were converted to bigWig format for visualization in the UCSC genome browser Mtb H37Rv 06/20/1998 assembly [55] . To determine the level of ChIP-Seq enrichment for genes , the enrichment ratio ( ER ) was calculated by dividing the read counts for the ChIP-Seq sample by the read counts of the Mock IP sample . Topo I and Gyrase binding site motifs were searched for using the MEME Suite ( http://meme . nbcr . net/meme/ ) in regions 300 bp upstream and 300 bp downstream of the predicted peak summit and a consensus motif sequence was deduced . ChIP-Seq data for RNAP was taken from NCBI [22] and analyzed as described above . For generating the occupancy profile , transcription start site ( TTS ) co-ordinates were obtained from Cortes et al . , 2013 [56] . Mtb H37Rv genome annotation was taken from the TubercuList database ( http://tuberculist . epfl . ch/ ) , which contains 4019 protein coding sequence ( CDS ) , 73 genes encoding for stable RNAs , small RNAs and tRNAs . Mycobacterium smegmatis mc2 155 cells were electroporated with the pJAM2-Rv3852 construct [24] . The exponentially grown culture was treated with 2% acetamide for 6 h to induce transcription of the rv3852 gene cloned under the acetamidase promoter . Treated and untreated cultures were processed for ChIP using RpoB , Topo I and GyrA-specific ( against GyrA subunit of DNA gyrase ) antibodies . From the purified ChIP DNA , qPCR analysis was carried out to monitor the occupancy of Topo I , DNA gyrase and RNAP on rv3852 . The data were normalized with the mock control ( no antibody ) and presented as enrichment over the untreated samples ( induced/uninduced ) . A non-specific ( unrelated ) IgG antibody was used as a negative control . For Moxifloxacin ( Moxi ) -gyrase ChIP ( Moxi-Gyrase ChIP ) , Mtb cells were treated with 1 . 25 μg/ml for 8 h , collected by centrifugation , washed ( as described above ) and re-suspended in a buffer containing 50 mM HEPES-KOH pH 7 . 5 , 5000 mM NaCl , 1 mM EDTA , 1% Triton X-100 , 0 . 1% sodium deoxycholate , 0 . 1% SDS , Roche Antiprotease mini ) and further processed as described above . The transcription inhibition was carried out by treating Mtb cells with 0 . 25 μg/ml Rifampicin ( Rif ) for 6 h prior to the treatment with Moxi . To assess the occupancy of the topoisomerases on DNA engaged in active transcription , in vitro transcription reactions were carried out . The reactions were carried out in transcription buffer ( 50 mM Tris HCl , ( pH 8 . 0 at 25°C ) , 10 mM magnesium acetate , 100 μM EDTA , 100 μM DTT , 50 mM KCl , 50 μg/ml BSA , 5% glycerol ) using 30 nM supercoiled plasmid ( MtbGyr1542-pARN104-harboring the 867 bp region including the gyrase promoter elements and 675 bp region upstream and downstream to the transcription start site of gyrB respectively ) supplemented with 100 μM of NTPs and with or without 200 nM of σA-enriched RNAP ( M . smegmatis ) [57] . Reaction mixtures were incubated at 37°C for 5 min and supplemented with 30 nM of M . smegmatis Topo I , and 60 nM of reconstituted gyrase from Mtb , then incubated further at 37°C for 5 min and cross-linked using a UV trans-illuminator ( 365 nm for 30 min ) followed by ethanol precipitation . The pellet was dissolved in 40 µl of distilled water and treated with DNase I ( New England Biolabs , UK ) . Bound gyrase and Topo I were transferred to nitrocellulose membrane by slot blotting and analyzed by their respective anti-GyrA and anti-Topo I antibodies .
|
The generation of DNA topological constraint is intrinsic to transcription . Although in vitro studies indicated DNA gyrase and topoisomerase I action in the removal of excess supercoils , ahead and behind the transcribing RNA polymerase , in vivo recruitment and interaction of both topoisomerases with the genome has not been investigated . Using advanced sequencing , we have mapped the genome-wide footprints of topoisomerase I and DNA gyrase along with RNAP in deadly pathogen , Mycobacterium tuberculosis . We show that in vivo distribution of topoisomerases is guided by active transcription and both the enzymes co-occupy active transcription units . We establish their interaction with the regions of genome having propensity to accumulate negative and positive supercoiled domains , validating their role in managing the twin supercoiled domains .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"bacteriology",
"microbiology",
"dna",
"transcription",
"bacterial",
"genetics",
"sequence",
"motif",
"analysis",
"dna",
"microbial",
"genetics",
"bacteria",
"microbial",
"genomics",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"bacterial",
"genomics",
"genomic",
"signal",
"processing",
"genomics",
"bioinformatics",
"gene",
"expression",
"comparative",
"genomics",
"actinobacteria",
"biochemistry",
"signal",
"transduction",
"cell",
"biology",
"nucleic",
"acids",
"mycobacterium",
"tuberculosis",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"cell",
"signaling",
"organisms"
] |
2017
|
Transcription facilitated genome-wide recruitment of topoisomerase I and DNA gyrase
|
Trypanosoma cruzi is the causative agent of Chagas' disease . Novel chemotherapy with the drug K11777 targets the major cysteine protease cruzain and disrupts amastigote intracellular development . Nevertheless , the biological role of the protease in infection and pathogenesis remains unclear as cruzain gene knockout failed due to genetic redundancy . A role for the T . cruzi cysteine protease cruzain in immune evasion was elucidated in a comparative study of parental wild type- and cruzain-deficient parasites . Wild type T . cruzi did not activate host macrophages during early infection ( <60 min ) and no increase in ∼P iκB was detected . The signaling factor NF-κB P65 colocalized with cruzain on the cell surface of intracellular wild type parasites , and was proteolytically cleaved . No significant IL-12 expression occurred in macrophages infected with wild type T . cruzi and treated with LPS and BFA , confirming impairment of macrophage activation pathways . In contrast , cruzain-deficient parasites induced macrophage activation , detectable iκB phosphorylation , and nuclear NF-κB P65 localization . These parasites were unable to develop intracellularly and survive within macrophages . IL 12 expression levels in macrophages infected with cruzain-deficient T . cruzi were comparable to LPS activated controls . Thus cruzain hinders macrophage activation during the early ( <60 min ) stages of infection , by interruption of the NF-κB P65 mediated signaling pathway . These early events allow T . cruzi survival and replication , and may lead to the spread of infection in acute Chagas' disease .
Trypanosoma cruzi is the parasitic agent of Chagas' disease that affects approximately 12 million people throughout Latin America ( WHO ) . Current chemotherapy with nifurtimox and benznidazole is unsatisfactory due to severe side effects that require medical supervision [1]–[3] . T . cruzi infection is classically transmitted by an insect vector , the reduviid bug [4] . Parasites disseminate from the insect bite site and , in the most common clinical course of Chagas' disease , infect cardiac myocytes leading to acute myocarditis or chronic infection with relentless cardiac failure . Fulminant disease is commonly seen with HIV infection or immunosuppression . In patients with immunodeficieny , parasites may be found in many organs , and a highly fatal meningoencephalitis often ensues . These observations led us to hypothesize that T . cruzi successfully evades the host immune response , and may in fact utilize unresponsive macrophages as a means of egress from the insect bite site prior to dissemination to other cell types [5] . While the specific mechanisms of immune evasion by T . cruzi remain largely unknown , several reports have suggested that the major protease of T . cruzi , cruzain , ( a . k . a . cruzipain , GP57/51 ) is a key factor [6] , [7] . Cruzain plays a biological role in cell remodeling during transformation of the insect epimastigote stage of T . cruzi to infectious metacyclic [8]–[11] . Scharfstein and collaborators found cruzain involvement during trypomastigote infection by the proteolytic release of kinin from host cells surfaces and activation of bradykinin receptors [12]–[14] . More recent reports suggested that cruzain mediates anti-apoptotic mechanisms in T . cruzi-infected myocardiocytes in vitro [15] . The biological role of cruzain in the intracellular amastigote stage of T . cruzi and in Chagas' disease pathogenesis remains nevertheless largely unknown . Assessments of the role of cruzain in T . cruzi pathogenicity have been hampered by genetic redundancy and the failure of gene deletion attempts [9] , [12] , [16] , [17] . Successful knockout of the cruzain gene cluster has not been achieved to date leading to speculation that deletion of the cruzain coding genes might be lethal . Chemical knockout with cysteine protease inhibitors was also lethal for T . cruzi [18] . Although the role of auto-proteolysis in cruzain activation was confirmed , these studies did not allow elucidation of the biological role of this protease in human disease . Duschak et al . [19] , [20] observed lower cruzain activity and protease sequence alterations in attenuated strains as compared to their virulent parental T . cruzi , while other authors found infectivity is not dependent on cruzain expression or activity [21] . An alternative approach to study the biological role of this vital gene product is to examine the phenotype of protease-deficient organisms . We generated a protease deficient T . cruzi that retained less than 1% of cruzain activity of the wild type parental clone [22] . We now report that cruzain deficient T . cruzi rapidly activate host macrophages via NF-κB P65 and are unable to survive intracellularly within macrophages . In contrast , infection with wild type parasites appears to induce cruzain-mediated proteolysis of NF-κB P65 leading to unresponsiveness of the host macrophage during early ( <60 minutes ) infection . This immune evasion mechanism may be critical for T . cruzi survival during early natural infection with a low number of trypmastigotes .
Wild type CA-I/72 T . cruzi was isolated from an Argentinean chronic chagasic patient and cloned [23] . Cruzain-deficient , and cysteine protease inhibitor ( K11777 ) -resistant ( CA-I/KR; KR ) T . cruzi were derived from parental CA-I/72 parasites [22] . The cysteine protease inhibitor K11777 ( K777 , N-Methyl-Pip-F-hF-VSΦ ) was kindly provided by J . Palmer , Celera , CA [18] , [24] . Epimastigotes were maintained at 26°C in Brain-heart tryptose ( BHT ) medium [25] with 10% heat inactivated fetal calf serum ( FCS ) , and with the addition of 20-fold the lethal dose ( 200 µM ) of K11777 for phenotypic cruzain-deficient CA-I/KR . Bovine embryo skeletal muscle cells ( BESM ) cells were a kind gift of J . A . Dvorak , NIH . Mammalian stages of wild type ( WT ) CA-I/72 were maintained in BESM cells with RPMI-1640 medium and 5% heat inactivated horse serum ( HS ) at 37°C [26] . Intracellular cruzain-deficient CA-I/KR T . cruzi were maintained serially in BESM as above but with the addition of 5% heat inactivated FCS , 5% heat inactivated horse serum , and 10 µM K11777 . Long-term culture at higher concentrations of the inhibitor was toxic for BESM cells . For some experiments , the mammalian stages of cruzain-deficient CAI/KR were also cultured for 4 passages ( 2 months ) in the absence of inhibitor . The approximate duration of the intracellular cycle of wild type CA-I/72 ( 4 . 5 days ) [23] and CA-I/KR ( 18 days ) was estimated in infected BESM cells by contrast phase microscopy and by fluorescence microscopy of propidium iodine stained slides . Cruzain-deficient parasites were unable to survive and develop within macrophages . J774 mouse macrophages were from the UCSF cell culture facility . J774 cells were cultured as above and in some experiments irradiated ( 1000 Rad ) 24 h prior to use to arrest cell division . Normal peritoneal macrophages were collected from C3H mice ( The Jackson Lab ) . This study was carried out in strict accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol was approved by the Institutional Animal Care and Use Committee of the University of California , San Francisco ( AN079928-02A ) . WT and cruzain-deficient epimastigotes were washed three times in PBS , counted in a Coulter Counter Multisizer 3 ( Beckman ) and taken to equal numbers of parasites/ml , lysed by 5 cycles of freeze-thawing , centrifuged at 14 , 000 xg for 1 h at 4°C [22] , and frozen at −70°C until used . Fractions ( 105 epimastigotes/lane ) were reacted with the iodinated inhibitor DCG04 [27] , [28] and electrophoresed . Resolved molecular species were then isolated and analyzed by nano LC/MS , CID spectra , and/or fractionation by MALDI-TOF at the Protein Facility , UCSF . Protein fragments ( 9-17 aa long ) were analyzed by NCBI blast . BESM cells infected with intracellular amastigotes of WT or cruzain-deficient T . cruzi were processed for IEM with a specific anti-cruzain polyclonal antibody as previously described [24] . Briefly , T . cruzi-infected BESM cells were collected by centrifugation , washed twice with PBS and fixed for 2 hours at 4°C with 2% paraformaldehyde-0 . 05% glutaraldehyde in 0 . 1 M phosphate buffer , pH 7 . 4 . Cells were then cryoprotected , frozen , sectioned , and immunolabeled sequentially with rabbit polyclonal anti-cruzain Ab and goat anti-rabbit IgG-10 nm gold-labeled Ab ( dilutions 1∶250 and 1∶500 , respectively ) . Thin sections were observed in a Tecnai 10 ( FEI Co . ) electron microscope . Gold-labeled cruzain localizing to the amastigote cell surface was quantified in micrographs of intracellular parasites in three independent experiments ( n = 55 wild type amastigotes; n = 55 cruzain-deficient amastigotes ) . The cell perimeter of sectioned intracellular amastigotes was measured with Openlab software ( Improvision ) , and cruzain density was expressed as the number of gold labeled cruzain particles/µm of parasitic cell membrane . Results were statistically analyzed ( t test ) . J774 macrophages were seeded onto 12 well tissue culture plates for 24 h prior to infection with WT or cruzain-deficient T . cruzi as appropriate . Infection was performed at a low ratio of 0 . 5 parasites/macrophage . Some infected cultures were treated with cysteine protease inhibitor . Controls consisted of J774 macrophage cultures that were uninfected , treated with purified LPS ( 150 ng/ml LPS , Sigma ) [29] , or treated with K11777 . Samples were collected at 1 min , 30 min , 60 min , and 150 min , and at 48 h post-infection . Monolayers were washed twice with cold PBS , solubilized in 1 ml cold PBS with1% Triton X-100 , scraped , aliquoted , immediately snap-frozen , and stored at −70°C until used . Samples were centrifuged at 14 , 000 xg for 30 min at 4°C and supernatants taken to identical protein concentration . Samples ( 1 mg protein/lane ) were heated at 70°C for 5 min and resolved by Nupage electrophoresis ( 10% BisTris gels at 200 V , Mops Buffer ) ( Novex , Invitrogen ) prior to transferring to PVDF membranes ( 35 V for 2 h ) . Blots were blocked overnight with 1% BSA ( Sigma ) . Samples were blotted with rabbit anti–NF-κB P65 Ab , rabbit anti- iκB Ab , mouse anti-phosphorylated ( ∼P ) iκB Ab ( Santa Cruz Biotech , CA ) and anti actin Ab ( Cal Biochem ) . All methods were according to manufacturer's instructions ( Santa Cruz Biotech , CA ) . Additional controls were similarly infected with Leishmania mexicana , a kind gift of J . Mottram ( University of Glasgow , UK ) , for 48 hours . Results were confirmed in duplicate experiments according to methods described by Ma and colleagues [30]–[32] . Briefly , 5×106 cells per point ( 1 , 10 , 30 , and 60 min ) were infected with T . cruzi trypomastigotes as above ( ratio 0 . 5 parasites/cell ) . Non-infected macrophages either untreated ( MΦ ) or treated with LPS , and/or K11777 as indicated ( 10 , 20 , 40 and 60 min ) were also used as controls . Macrophages were washed twice with PBS , scraped , transferred to eppendorf tubes , and centrifuged prior to lysis with NP40 lysis buffer ( 0 . 1% NP40 , 1 mM Na vanadate and 5 mM Na fluoride−50 mM HEPES ) . Nuclei were then pelleted by centrifugation ( 20 min at 14 , 000 xg , 4°C ) and discarded , and the supernatants were boiled in Laemli buffer for 10 min immediately prior to SDS-PAGE/WB ( Cell Signalling Technology ) [30] , [31] . Membranes were developed sequentially with anti ∼PiκB Ab ( Cell Signalling ) , anti-iκB Ab ( Cell Signalling ) , and anti-actin Ab ( Southern Biotech ) [30] . Duplicates of samples described above for WB and appropriate controls were simultaneously seeded onto 12 well tissue culture plates containing sterile round cover glasses . At 1 min , 30 min , 60 min , and 150 min post-infection , cells were washed , fixed with fresh 4% paraformaldehyde in PBS , rinsed in PBS , treated with 0 . 1% Triton X-100 in PBS for 5 min , rinsed , and processed for confocal microscopy with anti -mouse NF-κB P65Ab ( Dil 1/100; Santa Cruz Biotech , CA ) followed by Alexa 488 labeled secondary Ab and propidium iodine ( Molecular Probes ) . Results are from 3–7 independent experiments ( n = 2–3 slides per treatment ) . Confocal images were acquired with a Leica Laser Confocal microscope TCS-NP , using Leica software and identical parameters for all samples , namely , a 100X objective with a numerical aperture of 1 . 4 in a 1024×1024 format , a pinhole of 0 . 7 airy , and a Z section of 0 . 2 µm . For colocalization studies of cruzain and NF-κB P65 , macrophages infected for 30 min with WT T . cruzi and treated or not with K11777 , or similarly infected with cruzain-deficient T . cruzi and treated or not with K11777 , and non-infected controls were fixed and stained sequentially with rabbit anti-NF-κB P65 Ab ( Dil 1/100; Santa Cruz Biotech ) and Alexa 488-labeled secondary Ab , followed by rabbit polyclonal anti-cruzain antibody ( Dil 1/100 ) [24] previously labeled with Alexa 594 according to manufacturer's instructions ( Molecular Probes ) . For confocal studies of IL12 expression , macrophages treated or not with K11777 were infected with WT or cruzain-deficient T . cruzi for 1 h ( ratio 0 . 5 trypomastigotes/macrophage ) followed by overnight incubation with fresh RPMI medium at 37°C . Cultures were then exposed to 150 ng/ml purified LPS for 2 h ( Sigma ) , followed by 5 µg/ml BFA ( Molecular Probes ) for 1 h at 37°C . Appropriate controls were included in three independent experiments ( n = 2 per sample ) . Slides were then fixed as above , and simultaneously stained with anti-IL12 antibody ( 1/100 ) ( R&D Systems ) [33] for 1 h followed by secondary Cy2-labeled secondary Ab ( dil 1/1000 ) ( Biomed ) and propidium iodine ( PI ) ( Dil 1/5000 ) ( Molecular Probes ) . Coverslips were mounted with Vecta-Shield ( Vector Lab ) . All slides were observed and documented on the same day in a Leica microscope , model TCS-NP , using Leica software and identical confocal parameters as described above . The microscopist analyzed all slides blindly . Fluorescence was then analyzed quantitatively with Openlab software ( Improvision ) . Fluorescence intensity was recorded at random ( n = 12 per sample ) and statistically analyzed ( t test ) . For other additional studies , macrophages infected with WT and cruzain-deficient CA-I/KR , and uninfected controls were stained with a macrosialin ( mouse CD68 homolog ) specific Ab ( Serotec ) ( Dil 1/200 ) followed by a CY2 affinity-pure donkey anti-rat IgG ( Jackson Immuno-Research lab ) secondary Ab and PI ( Molecular Probes ) . Human recombinant NF-κB P65 ( rNF-κB P65 ) ( Active Motif , Japan ) expressed in E . coli from a full-length cDNA clone has a 14 aa deletion at the C-term . One µl of rNF-κB P65 was diluted to 150 nM in 10 µl of 1x buffer ( with 5 µl 1M DTT/ml or 5 mM DTT ) and reacted or not with recombinant cruzain at 150 nM ( 1/1 ) and serial dilutions ( 1/10 to 1/1000 ) in 100 mM Na Acetate Buffer pH 5 . 5 for 2 h at 37°C . Human rNF-κB P65 was also reacted with sonicates from 2 . 5×105 wild type CA-I/72 or cruzain deficient CA-I/KR parasites as above . Samples were then resolved by WB as above with anti- NF-κBP65 Ab . To better understand the effect of macrophage activation on parasite intracellular development , macrophages were infected with WT T . cruzi ( ratio 0 . 5 trypomastigotes/cell ) for 1 h at 37°C . Cultures included untreated controls , and macrophages treated with 150 nM purified LPS [29] as follows: LPS was added 1 h prior to T . cruzi infection , concomitantly with infection , or 1 h after infection . Cultures were treated with LPS for up to 48 h . Duplicate cultures were fixed with 4% paraformaldehyde at 1 h ( to ) , 24 h , and 48 h post-infection . Cells were stained with PI and the mean number of intracellular parasites/cell was estimated in 200 cells per slide ( n = 3 slides/treatment ) in two independent experiments [26] . Results were analyzed statistically ( t test ) . Cruzain activity was determined in extracts of WT and cruzain-deficient CA-I/KR epimastigotes as previously described [22] . For L-arginase determinations [7] , [34] , J774 uninfected controls and macrophages infected with WT or cruzain-deficient trypomastigotes were used . Cells were collected 48 h post-infection and samples fixed in 4% paraformaldehyde for counting in a Coulter Counter Multisizer 3 ( Beckman ) . Aliquots corresponding to 105cells/50 µl were prepared by duplicate for L-arginase assays as described by Stempin et al . [7] Briefly , cells were lysed in 0 . 1% Triton X-100 buffer with protease inhibitors , centrifuged at 14 , 000 xg for 30 min at 4°C , and supernatants stored at −70°C . L-arginase activity in supernatants was determined as described [7] , [34] . Urea ( µg/ml ) was measured at 540 nm . Results from three independent experiments ( n = 3 per sample ) were analyzed statistically with Prism 4 software .
The cysteine protease inhibitor K11777 is cidal for WT T . cruzi . By exposing CA-I/72 T . cruzi to gradual step-wise micromolar increases in K11777 concentration over a period of 2 years , we generated a K11777-resistant and cruzain-deficient T . cruzi that retains negligible ( <1% ) protease activity as detected by a fluorescent protease substrate [22] . Both parental WT and cruzain-deficient parasites are clonal populations [22] . Cruzain-deficient T . cruzi have remained resistant to K11777 for 14 years even after passage through animals . The mechanism of drug resistance has been described in epimastigotes and is due to secretion of inactive , unprocessed cruzain [22] . The IC50 for CA-I/72 epimastigotes is 3–5 µM K11777 while cruzain-deficient epimastigotes are routinely maintained in 20-fold the lethal dose of K11777 ( 200 µM ) . IC50 values for intracellular wild type and cruzain-deficient amastigotes are 0 . 8–1 µM and 12 µM K11777 , respectively . Differing from parental CA-I/72 T . cruzi that are lethal at doses of ≥103 trypomastigotes , cruzain-deficient parasites are unable to establish infection in normal mice even at doses of 106 trypomastigotes and are only lethal in a severely immunodeficient Rag 1 −/− mouse model of Chagas' disease ( Doyle , unpublished data ) . These results will be independently submitted for publication . To identify and compare proteases in WT and cruzain-deficient T . cruzi , we used a functional proteomic method developed to profile protease targets in crude cellular extracts [27] , [28] . The iodinated probe DCG04 , allowed the specific identification of mature , active cruzain only in parental WT epimastigotes following SDS-PAGE ( Figure 1 ) . Cruzain-deficient parasites expressed unprocessed inactive cruzain with the prodomain attached as previously shown by WB [22] but no active mature protease . The identity of cruzain was confirmed by MS/MALDI and NCBI blast . To confirm decreased cruzain expression also in the intracellular pathogenic amastigote stage we performed comparative , quantitative IEM . A statistically significant ( p<0 . 1 ) three-fold decrease in expression of membrane bound cruzain was confirmed for intracellular cruzain-deficient amastigotes as compared to WT controls ( Figure 2 ) . It has been proposed that proteases of L . mexicana cleave NF-κB to facilitate immune evasion [35] . We therefore assayed NF-κB activation in cells recently ( ≤60 min ) infected by cruzain deficient parasites . We compared macrophage activation signaling pathways [36]–[39] as a consequence of early infection with WT versus cruzain-deficient T . cruzi . The infection of a cell population by T . cruzi follows a binomial distribution , with few cells very heavily infected while most cells remain non-infected . Moreover , infection with a high ratio of parasites per cell results in premature rupture of heavily infected macrophages . To prevent the premature release of intracellular parasites that may secondarily activate macrophages in the population , we used a very low infection ratio of 0 . 5 parasites per cell . To detect macrophage activation , we performed WB targeting an NF-κB family member and its specific inhibitor [40]–[45] . Similar results were obtained with two methods and Abs from the two different vendors . WB analyses with anti-NF-κB P65 Ab ( Figure 3A ) showed expression in macrophages infected with WT T . cruzi at 1 min and 30 min post-infection ( lanes 1 , 4 ) was similar to uninfected controls ( lane 3 ) , and lower ( lane 8 ) at 60 min post-infection . Phosphorylation of iκB is indicative of cell activation [40] , [45] . Samples for WB were either heated to 70°C for 5 min ( Santa Cruz Biotech . methods ) or boiled for 10 minutes immediately prior to SDS-PAGE to dissociate complexes [30]–[32] . Samples blotted with specific anti ∼PiκB Abs from different commercial sources and using different protocols confirmed negligible ∼P iκB indicative of unresponsiveness in macrophages infected with WT T . cruzi for up to 60 min and uninfected controls ( Figure 3B ) . In contrast , ∼P iκB indicative of activation was confirmed in macrophages infected with cruzain-deficient T . cruzi as early as 1 min post-infection . Additional macrophage controls infected with cruzain-deficient T . cruzi cultured for 2 months without K11777 showed intermediate results as these parasites only regained partial cruzain activity , and were not investigated further [22] . Additional controls were uninfected macrophages treated or not with LPS and/or K11777 ( K77 ) ( Figure 3D ) . In L . mexicana infected controls [32] , NF-κB P65 degradation was still observed 48 h post-infection ( data not shown ) . To further investigate macrophage responses to WT versus cruzain deficient T . cruzi-infection , we performed confocal studies . NF-κB P65 localization was cytoplasmic in macrophage controls ( Figure 4E ) . NF-κB P65 localization was also cytoplasmic in macrophages infected with WT T . cruzi for 1 , 30 , and 60 min post-infection confirming unresponsiveness to T . cruzi infection . Interestingly , NF-κB P65 covered the surface of WT amastigotes as early as 30 min post-infection ( Figure 4A–B ) . Similar results were observed in BESM cells and thioglycolate elicited peritoneal mouse macrophages ( data not shown ) . In contrast , intense NF-κB P65 label indicative of cell activation localized to the nucleus of macrophages infected with cruzain-deficient T . cruzi ( Figure 4C–D ) ( yellow fluorescence ) , and no NF-κB P65 was seen bound to parasites for up to 60 min . Activation also occurred in macrophage controls infected with WT T . cruzi in the presence of 10 µM K11777 ( data not shown ) . Colocalization of NF-κB P65 ( green fluorescence ) and cruzain ( red fluorescence ) on the cell surface of intracellular parasites occurred in macrophages infected with WT T . cruzi even during trypomastigote to amastigote transformation ( Figure 5A–C ) . Intracellular WT T . cruzi treated with the trypanocidal inhibitor K11777 for 30 min showed marked cruzain accumulation presumably in the parasitic Golgi compartment [24] and no recruitment of NF-κB P65 to the parasite surface ( Figure 5D , arrow ) . Human rNF-κB P65 was proteolytically cleaved into two major fragments by native cruzain expressed by WT ( Figure 6 ) but not by proteases expressed by cruzain-deficient T . cruzi . Similarly , recombinant cruzain degraded human r-P65 at equimolar concentrations under the experimental conditions used but not in the presence of 10 µM K11777 . To understand the effect of macrophage activation [37] , [42] on T . cruzi intracellular survival and development , we performed quantitative in vitro assays [26] . T . cruzi divided normally in control macrophages and the mean number of parasites per cell ( P/cell ) increased from ∼0 . 5 at to to 2 . 8 at 48 h ( Figure 7 ) [23] . Activation of macrophages with purified LPS 1h prior to or concomitantly with infection resulted in parasite death . T . cruzi developed well in macrophages treated with LPS 1h post-infection; the lower mean P/cell probably results from death of parasites still trapped within the parasitophorus vacuole when activation occurred . Thus macrophage unresponsiveness in early infection ( <60 min ) is crucial for parasite survival and intracellular development . To confirm macrophage unresponsiveness to infection with WT T . cruzi infection , we next investigated IL12 expression by confocal microscopy followed by fluorescence quantification with Improvision software . Macrophages infected with WT T . cruzi , and subsequently activated with LPS followed by BFA treatment , showed negligible cytoplasmic IL12 ( Figure 8A ) . A significant increase ( P<0 . 1 ) in IL12 accumulation occurred in macrophages infected with wild type T . cruzi and treated with the inhibitor K11777 that prevents cruzain activity ( Figure 8B ) . Similarly , activation and significant IL12 accumulation occurred in macrophages infected with cruzain-deficient T . cruzi ( Figure 8C ) . Other investigators have reported increased L-arginase activity induced by cruzain treatment of cells [7] . Only WT parasites induced 2-5 fold higher L-arginase activity in host macrophages while cruzain-deficient parasites failed to increase enzymatic activity ( P<0 . 01 ) . Representative values from one of 3 independent experiments are 3 . 13±0 . 02 µg urea/ml for macrophages infected with WT parasites , 1 . 56±0 . 02 µg urea/ml for macrophages infected with cruzain-deficient parasites , and 1 . 5±0 . 01 µg urea/ml for uninfected macrophages .
A complex and dynamic scenario of host cell-parasite interactions is becoming apparent as pathogens use different mechanisms to subvert host-cell signaling pathways or modulate their kinetics during intracellular development . Several reports show that T . cruzi modulates signaling pathways in mammalian cells [15] , [43] , [44] . Some authors propose that activation of NF-κB proteins may modulate tissue specificity , as muscle cells that preferentially harbor T . cruzi do not become activated [37] while others detect muscle and endothelial cell activation post-infection [45] , [46] . Our interest was to investigate the role of the protease cruzain during early infection ( <60 min ) of macrophages by T . cruzi , and in particular on the nuclear factor NF-κB P65 , using the inhibitor K11777 and phenotypic cruzain knockouts [22] . A major role of the NF-κB family is the regulation of aspects of the innate and adaptive immune responses . NF-κB members control the transcription of genes encoding cytokines and antimicrobial molecules , as well as genes regulating cell differentiation , survival , proliferation and apoptosis [38] . The complex response depends on the cell type and on the nature , duration and intensity of the activating signal . NF-κB complexes are normally inactive in the cytoplasm and upon activation they enter the nucleus to modulate gene expression . Membrane receptors ( e . g . TRL ) and other signaling regulators are also involved [38]–[42] , [43]–[52] . To compare infection with WT versus cruzain-deficient T . cruzi , we performed WB of NF-κB P65 and ∼P iκB ( Figure 3 ) [39] , [47] that showed unresponsiveness of macrophages infected with WT parasites . Confocal studies showed that WT parasites rapidly sequestered NF-κB P65 onto their cell surface ( Figure 4 , A–B ) . Similar events occurred during early infection of peritoneal mouse macrophages and BESM cells ( data not shown ) . Cruzain localized to the cell surface of WT intracellular amastigotes ( Figure 2 ) [53] and colocalized with NF-κB P65 ( Figure 5 ) . Treatment with K11777 caused accumulation of unprocessed and inactive protease in the Golgi [24] and abrogated NF-κB P65 sequestration by the parasite ( Figure 5D ) . Degradation of NF-κB P65 was noted 60 min after infection with wild-type T . cruzi ( Figure 3 ) and native cruzain degraded recombinant nuclear factor ( Figure 6 ) . Moreover , inhibition of IL12 confirmed macrophage unresponsiveness in early WT T . cruzi infection ( Figure 8A ) . In contrast , macrophages infected with cruzain-deficient T . cruzi became rapidly activated via NF-κB P65 ( Figure 3 , KR ) and intense nuclear localization of P65 occurred shortly after infection ( Figure 4C–D ) highlighting a role for cruzain in the modulation of host cell signaling pathways . As anticipated , infection with cruzain-deficient parasites induced IL12 accumulation ( Figure 8C ) . Preventing macrophage activation is crucial for a successful infection . Indeed , wild type T . cruzi were killed when macrophages were activated with LPS prior to or during infection ( Figure 7 ) , presumably while parasites are still contained within the parasitophorus vacuole [54]–[57] and are susceptible to cidal macrophage products . Other macrophage metabolic pathways were also modulated by T . cruzi . Infection with WT but not cruzain-deficient T . cruzi significantly increased L-arginase activity [6] , [7] . A similar phenomenon occurs during Leishmania infection [58] . We also observed down-modulation of macrosialin , the murine analog of human CD68 , in cells infected with WT T . cruzi ( not shown ) . Further studies may identify other host cell molecules modulated by T . cruzi [59]–[63] . Experimental and clinical evidence shows a correlation between the severity of Chagas' disease and the persistence of T . cruzi within tissues [64]–[65] . NF-κB complexes are important for the development and function of both innate and adaptive immune responses triggered by pathogens [38] , [51] . Metabolic pathways of the innate immune system induced in early infection have important consequences in the evolution of the disease as they play a role in the control of T . cruzi replication , tissue distribution and degree of parasitism . As infection progresses , an adaptive immune response develops with production of high levels of inflammatory cytokines , IFNγ and NO radicals [38] , [59] , [66]–[71] . In the case of T . cruzi , TLR2 and TLR9 receptors mediate cellular activation [43] , [59] , [66] , [72]–[73] . Our results show that macrophages infected with WT T . cruzi remained unresponsive in early infection , and support the hypothesis that intracellular pathogens are uniquely protected from macrophage activation [38] . Confirming our results , expression profiling of bone marrow macrophages infected with T . cruzi revealed very few transcriptional changes during the first 12 hours . At 24 hours post infection , both macrophages and fibroblasts express some interferon-regulated genes [62] , [74] , [75] . The L . mexicana cruzain homologue , “cpb” , likewise degrades NF-κB P65 , IκBα , and IκBβpreventing macrophage activation [35] . Similar strategies to suppress immune responses have been found in the coccidian parasite Theileria [76] , and in poliovirus and other picornavirus as viral proteases cleave NF-κB P65 generating inactive products [77] . We then hypothesize that cruzain plays a key role by degrading NF-κB P65 and hindering activation of innate phagocytes recruited to the bite site [5] . Macrophage unresponsiveness would favor parasite survival in early infection and delay the onset of the immune response . A successful macrophage infection results in the production of hundreds of infectious T . cruzi . For example a single CA-I/72 parasite originates ∼130 infectious trypomastigotes in just 4 . 5 days [23] . Once the host immune response is triggered on [65]–[67] , parasite development would be restricted to permissive tissues such as muscle cells [37] , [71] . Thus cruzain may function during the early events of macrophage infection favoring immune evasion by T . cruzi .
|
Trypanosoma cruzi ( T . cruzi ) is the unicellular parasite that causes Chagas' disease , a devastating health burden throughout Latin America now also affecting developed countries . Macrophages are the first cells that become infected by T . cruzi and disseminate the infection to other tissues . The parasite then preferentially infects and multiplies within heart muscle cells causing severe heart disease and often death . The new drug K11777 targets a vital parasite enzyme , the protease cruzain . Consequently , it is important to understand what the enzyme is doing during infection . To elucidate the role of the protease , we compared infection of macrophages with parental wild type parasites and with protease deficient T . cruzi . We now report a role for the parasitic protease in immune evasion . The protease prevents macrophage activation thus allowing T . cruzi survival and replication , and favoring the spread of infection .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"biology"
] |
2011
|
The Trypanosoma cruzi Protease Cruzain Mediates Immune Evasion
|
Aβ peptide accumulation is thought to be the primary event in the pathogenesis of Alzheimer's disease ( AD ) , with downstream neurotoxic effects including the hyperphosphorylation of tau protein . Glycogen synthase kinase-3 ( GSK-3 ) is increasingly implicated as playing a pivotal role in this amyloid cascade . We have developed an adult-onset Drosophila model of AD , using an inducible gene expression system to express Arctic mutant Aβ42 specifically in adult neurons , to avoid developmental effects . Aβ42 accumulated with age in these flies and they displayed increased mortality together with progressive neuronal dysfunction , but in the apparent absence of neuronal loss . This fly model can thus be used to examine the role of events during adulthood and early AD aetiology . Expression of Aβ42 in adult neurons increased GSK-3 activity , and inhibition of GSK-3 ( either genetically or pharmacologically by lithium treatment ) rescued Aβ42 toxicity . Aβ42 pathogenesis was also reduced by removal of endogenous fly tau; but , within the limits of detection of available methods , tau phosphorylation did not appear to be altered in flies expressing Aβ42 . The GSK-3–mediated effects on Aβ42 toxicity appear to be at least in part mediated by tau-independent mechanisms , because the protective effect of lithium alone was greater than that of the removal of tau alone . Finally , Aβ42 levels were reduced upon GSK-3 inhibition , pointing to a direct role of GSK-3 in the regulation of Aβ42 peptide level , in the absence of APP processing . Our study points to the need both to identify the mechanisms by which GSK-3 modulates Aβ42 levels in the fly and to determine if similar mechanisms are present in mammals , and it supports the potential therapeutic use of GSK-3 inhibitors in AD .
Alzheimer's disease ( AD ) is the leading cause of dementia in the ageing population . Symptoms include , but are not limited to , memory loss , cognitive decline , and deterioration of language skills . The pathological hallmarks of AD are the presence of plaques and neurofibrillary tangles [1] . The tangles are composed of hyperphosphorylated tau protein while the plaques are comprised of amyloid beta ( Aβ ) peptides , various species of which are derived from the amyloid precursor protein ( APP ) , the most abundant being Aβ40 and Aβ42 [2] . AD-causing mutations either increase the level of Aβ42 or the ratio of Aβ42/Aβ40 , indicating that this is the more toxic form of the peptide [2] . The leading candidate explanation for the molecular basis of AD pathology is the amyloid cascade hypothesis . This states that the Aβ protein initiates the disease process , activating downstream neurotoxic mechanisms including the dysregulation of tau . Perhaps the strongest support for the amyloid cascade hypothesis is that all of the mutations implicated in early-onset , familial AD , such as the Aβ Arctic mutation , increase the aggregation or production of Aβ [1] . Although tau mutations exist , none have been linked to familial AD , but rather to fronto-temporal dementia , in which Aβ plaques are absent [3] , [4] . The amyloid cascade has also been tested experimentally in various ways . For example , a double transgenic mouse model expressing APP-V7171 and Tau-P301L , develops amyloid pathology similarly to mice transgenic for APP-V7171 alone , whereas tauopathy is dramatically enhanced in the double transgenic compared to mice transgenic for Tau-P301L alone . This implies that Aβ pathology affects tauopathy but not vice versa [5] . Also , clearance of Aβ using Aβ-specific antibodies reduced early tau burden , while elevating tau burden in transgenic mice had no effect on Aβ accumulation [6] , [7] . Furthermore , a reduction in tau levels rescued learning and memory impairment induced by Aβ in a mouse model expressing human APP [8] . Aβ increases the phosphorylation of tau protein and concomitantly activates glycogen synthase kinase , GSK-3 [9] , [10] . GSK-3 is a multi-functional kinase involved in regulating various cellular processes , including growth and differentiation [9] , [11] . There are two isoforms of the protein , GSK-3α and GSK-3β . They share 98% identity within their kinase domain , but are not functionally identical , although both have been suggested to be involved in AD pathogenesis [11] . GSK-3α has been implicated in the amyloidogenic processing of APP to yield Aβ peptides [12] , while GSK-3β has been implicated in the tau-related pathogenesis of AD , by colocalizing with tau tangles and phosphorylating tau [9] . As yet , the exact role of GSK-3 in the generation of Aβ peptides is not known . GSK-3 itself is also regulated by phosphorylation . Phosphorylation at Ser9 of GSK-3β and the equivalent Ser21 of GSK-3α inhibits activity , while phosphorylation at Tyr216/Tyr279 of GSK-3β and GSK-3α , respectively , is thought to increase activity [13] . Remarkable similarities are seen between double transgenic mice expressing tau either with APP or with GSK-3β . This finding is consistent with the hypothesis that amyloid acts via activation of GSK-3 to modulate tau function [5] , [9] , [14] . Lithium chloride , which is used as a mood stabilizing agent in patients with bipolar disorders , inhibits GSK-3 activity , either by competing with magnesium ions [15] or by increasing Ser9 phosphorylation [16] . Lithium reduces amyloid production by altering the role of GSK-3α in APP processing/cleavage; selective inhibition of GSK-3 by siRNA or expressing a dominant negative form of GSK-3 also decreases Aβ production in cultured cells and mice [12] , [17] . Furthermore , lithium reduces both tau phosphorylation at several GSK-3 epitopes and tauopathy in a mouse model expressing mutant human tau [12] , [18] . However , in another study , lithium was seen to reduce tau phosphorylation but not to affect Aβ load in a triple mutant mouse expressing human APPswe , human tauP301L and with mutant presenilin 1 PS1M146V knock-in . The differing observations might be due to variations in the age at which the mice were treated with lithium , as suggested by the authors [19] . Fruit flies , Drosophila melanogaster , can provide useful invertebrate models of neurodegeneration because of their complex brains , short lifespans and relative ease of genetic manipulation . Several fly models of aspects of AD biology have been made , including ones that over-express either Drosophila or human tau , and show neuronal dysfunction phenotypes [20]–[22] . Co-expression of human tau protein with Shaggy ( Sgg ) , the Drosophila homologue of GSK-3 [23] , exacerbates these neurotoxic phenotypes and leads to the appearance of neurofibrillary tangles [22] , [24] , [25] . Fly models expressing Aβ peptides have also been generated , and show neurodegeneration and amyloid deposits [26] , [28] . Although an APP orthologue exists in flies , the Aβ sequence is not conserved , and Drosophila models directly expressing Aβ allow study of Aβ toxicity in the absence of any endogenous amyloid production [29] . In this study we have generated a fly model that expresses Arctic mutant Aβ42 peptide in the nervous system of adult flies , using an inducible system for gene expression , because we wished to understand the underlying mechanism of disease progression of AD in adults , without complications from developmental effects . We first characterised this model , and then used it to investigate the amyloid cascade hypothesis , by modulating the levels of endogenous fly tau and examining the effects on phenotypes consequent upon expression of Aβ . We also investigated the requirement for GSK-3 in Aβ pathology and its role in direct regulation of Aβ peptides .
To generate an adult-onset fly model of Alzheimer's disease , we expressed Arctic mutant Aβ42 peptides using an inducible pan-neuronal driver . An elav GeneSwitch ( elavGS ) driver line [30]–[31] that has been used previously to develop an adult-onset Drosophila model of spinocerebellar ataxia ( SCA ) [32] was used to direct expression of a UAS-Arctic Aβ42 transgene [33] both spatially and temporally , to neurons of the adult fly ( Figure 1A ) . A UAS-Aβ40 line [33] was used as a control for over-expression of non-toxic forms of Aβ in fly neurons , since this form of the peptide has previously been shown to have no detrimental effect in flies [26]–[28] . We measured expression of Aβ peptides in adult neurons when we treated elavGS;UAS-Arctic Aβ42 flies with the activator mifepristone ( RU486;RU ) from two days post-eclosion , by measuring RNA and protein levels at 4 and 21 days into treatment ( Figure 1B and 1C ) . Aβ transcripts were clearly elevated in RU-treated elavGS;UAS-Arctic Aβ42 flies in comparison with untreated ( −RU ) flies at both time-points ( Figure 1B ) . Moreover , an Aβ42-specific ELISA confirmed that Aβ42 protein was elevated in elavGS;UAS-Arctic Aβ42 ( +RU ) flies compared to untreated ( −RU ) flies and that the level of protein increased with age ( Figure 1C ) . Since RNA transcript level decreased with age , this age-dependent accumulation of Aβ42 protein is most likely to be attributable to an increased rate of translation of the protein relative to the rate of protein degradation . Aggregation of Aβ has been shown to be of critical importance for its pathogenicity [34] . Therefore , we assessed the state of aggregation of Aβ42 in the mutant flies by separating soluble and insoluble protein fractions from fly brain extracts . At day 15 , when the first signs of pathology were observed in the Arctic Aβ42 flies ( see below ) , we found that most of the Aβ42 protein had accumulated into an insoluble , fibrillar form ( Figure 2 ) , consistent with the aggregation-promoting effects of the Arctic mutation [35] . Overall these results confirm that the elavGS-UAS system used in this study is sufficient to induce over-expression of Aβ peptides specifically in the adult fly nervous system , and that Arctic mutant Aβ42 protein accumulates with age . Previously published studies have shown that constitutive expression of Arctic Aβ42 peptide in fly neurons significantly shortens lifespan , induces behavioural impairments and causes neuronal death [26]–[28] . To determine whether adult-onset expression of Arctic Aβ42 peptide in neurons causes similar phenotypes , we examined survival , neuronal and behavioural dysfunction in our inducible Drosophila model of AD . First , we measured the effects of Arctic Aβ42 expression on lifespan , by treating elavGS;UAS-Arctic Aβ42 and elavGS;UAS-Aβ40 flies with RU from two days post-eclosion and recording their survival . Expression of Arctic Aβ42 in adult neurons shortened median lifespan by about 50% and maximum lifespan by about 45% in comparison to non-RU-treated flies , and to Aβ40 +RU and −RU control flies ( Figure 3 ) , demonstrating a specific lifespan-shortening effect of Arctic Aβ42 compared to the Aβ40 form of the peptide . Next , we determined whether adult-onset expression of Arctic Aβ42 peptide in fly neurons caused neuronal toxicity , by analysing neuronal function . As a direct measure of physiological activity , we examined the electrophysiological responses of the Drosophila giant fibre system ( GFS; Figure 4A ) . Adult elavGS;UAS-Arctic Aβ42 flies were fed + or − RU486 media from two days post-eclosion , and GFS activity measured at day 16 and day 28 into treatment . Giant fibres ( GF ) were stimulated via electrodes inserted inside the compound eye , and post-synaptic potentials recorded in the tergotrochanteral muscle ( TTM ) and the dorsal longitudinal flight muscle ( DLM ) ( Figure 4A ) ; parameters measured were the latency from GF stimulation to muscle response and the stability of the response to high frequency stimulation . At day 16 , response latencies in the TTM , DLM and the TTM to high frequency stimulation were comparable between elavGS;UAS-Arctic Aβ42 flies on + and − RU486 food ( Figure 4B and Figure S1 ) . However , at day 28 , expression of Arctic Aβ42 peptide significantly increased the response latency measured in both the TTM and DLM , and inhibited the stability of the TTM response to high frequency stimulation ( at 100 , 200 and 250 Hz ) in comparison to untreated control flies ( Figure 4C and Figure S1 ) . This indicates a progressive neuronal dysfunction following adult-onset induction of Arctic Aβ42 , with young flies exhibiting no dysfunction in the GFS , while older flies showed obvious defects in response to both a single stimulus and to high frequency stimuli . As a behavioural measure of neuronal dysfunction in our inducible model , locomotor activity was assessed using a negative geotaxis ( climbing ) assay that has been used extensively to characterise fly models of neurodegenerative diseases [33] , [36] . Drosophila display an age-related decline in climbing behaviour , and this was apparent in the non-RU-treated and elavGS;UAS-Aβ40 +RU control flies used in the current study ( Figure 5 ) . We found that flies expressing Arctic Aβ42 displayed a reduced negative geotaxis in comparison to their −RU control flies and the Aβ40 +RU and −RU flies ( Figure 5 ) . The climbing behaviour of the Arctic Aβ42 flies had declined to a level by day 15 that was reached by the control flies only by day 28 . Finally , we quantified neuronal loss , as measured by the number of cell bodies in one hemisphere , in flies over-expressing Arctic Aβ42 peptide in adult neurons compared to non-expressing controls . No neuronal loss was evident in the brains of these flies ( Figure S2 ) . Collectively , these data demonstrate that expression of Arctic Aβ42 specifically in the neurons of the adult fly leads to early death and progressive neuronal dysfunction , in the apparent absence of neuronal loss . Hence we have successfully developed an inducible Drosophila model of AD that will provide a useful system in which to further investigate the potential mechanisms underlying pathogenesis in Alzheimer's disease , without any confounding effects on neuronal development . Because of the described role of GSK-3 in Alzheimer's disease , we investigated the activity of the fly orthologue of GSK-3 , Sgg , in the Aβ42-expressing flies . Phosphorylation at Ser9 of Sgg is important in suppressing its kinase activity . We found that expression of Arctic Aβ42 in the adult nervous system decreased the Ser9 phosphorylation level of Sgg , indicating an up-regulation of the activity of the kinase ( see Figure 6 ) . This increase in Sgg activity could have contributed to the toxicity we observed in our Aβ42 expressing flies . Lithium is a GSK-3 inhibitor , and we therefore tested its effect on Ser9 phosphorylation in the Aβ42 flies and , indeed , we found an increase in Ser9 phosphorylation compared to untreated controls ( Figure 6 ) . We next investigated if the increase in Sgg activity that we observed in the Artic Aβ42-expressing flies contributed to Aβ42 toxicity . To do this , we co-expressed in adult neurons a dominant negative form of Sgg , the S9E mutant , which mimics an inhibited state of the kinase [37] , [38] and renders it inactive . Expression of this dominant-negative Sgg increased the median and maximum lifespan of flies expressing Arctic Aβ42 . Flies co-expressing Arctic Aβ42 and the dominant negative mutant S9E lived significantly longer than control flies co-expressing Arctic Aβ42 and GFP ( Figure 7 ) , to control for any titration effect of GAL4 in the presence of a second UAS-transgene . Furthermore , inactivation of Sgg , either by expressing the dominant negative mutant S9E or feeding the flies lithium in adulthood , significantly suppressed the climbing deficit of the Arctic Aβ42-expressing flies ( Figure 8 ) . Two different doses of lithium ( 30mM and 100mM ) both rescued the climbing deficit of the Aβ42-expressing flies . These data demonstrate that inhibiting the activity of Sgg in neurons in adults suppresses the adult onset Arctic Aβ42 induced toxicity , and demonstrate experimentally a functional role of GSK-3 in mediating Aβ42 toxicity . Since GSK-3 is a well-established tau kinase , and tau is abnormally phosphorylated in AD , we next investigated whether the protective effect of Sgg inhibition on Aβ42 toxicity in our fly model is mediated via alterations in tau phosphorylation . Hence , we examined the phosphorylation of Drosophila tau in flies over-expressing Arctic Aβ42 peptide in the absence or presence of lithium-treatment ( Figure 9 ) . We analysed tau phosphorylation using Phos-tag acrylamide gels , a technique for separating phosphorylated protein isoforms [39] , which has been employed previously to investigate the phosphorylation of fly proteins [40] . Phos-tag is a phosphate-binding compound which , when incorporated into polyacrylamide gels , can result in an exaggerated mobility shift for phosphorylated proteins , dependent on the degree of phosphorylation . When heat-stable fly head homogenates were run on Phos-tag gels , several prominent tau bands were detected , implying that endogenous tau is phosphorylated at multiple sites in WT tissue ( Figure 9A ) . De-phosphorylation using λ-protein phosphatase confirmed that the high molecular weight bands were due to tau phosphorylation , and that non-phosphorylated fly tau runs as a doublet ( Figure 9A ) . This method is thus capable of detecting at least some phosphorylation changes on the endogenous fly tau . Using this method , no difference in the level of tau phosphorylation was observed in flies over-expressing Arctic Aβ42 compared to non-expressing controls or to Arctic Aβ42 flies treated with lithium chloride ( Figure 9A and Figure S3A ) . The phosphorylation sites on fly tau have not yet been extensively characterized . Drosophila-specific phosphorylation-dependent antibodies are hence not available for the examination of specific sites . However , several GSK-3 specific sites [41] , [42] , and sites reported to be altered by Aβ42 peptide [43]–[45] , on human tau appear to be conserved in the Drosophila tau sequence ( Figure S4 ) . Of these , Ser262 , and Ser356 phosphorylation-dependent human tau antibodies were found to detect fly tau protein specifically ( Figure 9B and Figure S3B ) . We therefore used these antibodies to examine the effects on tau phosphorylation at these sites . Arctic Aβ42 over-expression , or treatment of Aβ42-expressing flies with lithium , did not modify phosphorylation at the Ser262 or Ser356 homologous tau epitopes ( Figure 9B ) . This suggests that neither Aβ42 nor GSK-3 are predominant in vivo regulators of the phosphorylation of these sites on fly tau . Although we could not uncover a role for tau phosphorylation in Aβ42 toxicity , we investigated whether the presence of tau modulates Aβ42 pathology . We found that loss of tau reduced the Arctic Aβ42 climbing dysfunction . UAS-ArcAβ42/+;elavGS tau EP3203/tau deficiency ( dfc ) flies , which express Arctic Aβ42 in a genetic background homozygous mutant for tau ( see Figure 10A for tau expression levels; tau antibody has previously been described [46] ) , had improved locomotor ability compared to UAS-ArcAβ42/+;elavGS tau EP2303/TM6 flies , which express a much greater level of tau ( P<0 . 0001 , two-way ANOVA; Figure 10A ) and UAS-ArcAβ42/+;elavGS/+ flies , which express wild-type tau levels ( P = 0 . 0002 , two-way ANOVA; Figure 10B ) , on +RU food . It is important to note , however , that tau loss of function flies themselves displayed some locomotor dysfunction compared to controls ( Figure 10B; P<0 . 0001 comparing UAS-ArcAβ42/+;elavGS/+ to UAS-ArcAβ42/+;elavGS tau EP3203/tau dfc on −RU486 , two-way ANOVA ) , thus potentially reducing the apparent protective effect of removing tau on Aβ42 pathology . These data parallel observations in mammals , where loss of murine Tau rescued Aβ-induced behavioural deficits in a mouse AD model [8] . We further investigated interactions between Drosophila tau and GSK-3 in the protection against Aβ42 pathology , to determine if they might act in the same biochemical pathway ( Figure 10B ) . Lithium treatment alone had a greater protective effect against Aβ42 toxicity than did loss of tau function alone , although this may have been confounded by the reduced climbing ability of flies with reduced tau . Moreover , lithium treatment rescued Aβ42-induced climbing dysfunction to the same extent in the presence or absence of tau , ( P = 0 . 692 comparing UAS-ArcAβ42/+;elavGS/+ and UAS-ArcAβ42/+;elavGS tau EP3203/tau dfc flies on +RU , + lithium food , two-way ANOVA ) suggesting that endogenous tau is not required for the lithium effect . This suggests that a large proportion of the protective effect of GSK-3 inhibition on Aβ42 toxicity is mediated via non-tau-dependent mechanisms . We also show that tau is required for the manifestation of Aβ42 effects . Our experimental design does not address , neither excludes , the possibility that a direct interaction of GSK-3 and tau may also affect Aβ42 toxicity . Because the amelioration of Aβ42 toxicity by reduced GSK-3 activity did not appear to be mediated mainly through tau , we next examined the direct effect of GSK-3 inhibition on Aβ42 levels . Interestingly , we found by ELISA analysis that Aβ peptide was significantly reduced in flies expressing Arctic Aβ42 when Sgg activity was reduced . Adult flies that co-expressed Arctic Aβ42 and the inactive dominant negative mutant Sgg S9E , or adult flies that expressed Arctic Aβ42 and were fed lithium , showed a major reduction in total Aβ42 levels in comparison to flies expressing Arctic Aβ42 alone and reared on food without lithium ( Figure 11A ) . The transcript levels of Aβ42 in the presence of functional or inhibited Sgg activity were not significantly different ( Figure 11B ) , suggesting that Sgg does not affect transgene expression , but rather acts directly or indirectly on Aβ degradation/sequestration . These data demonstrate for the first time a role of GSK-3 in determining the level of Aβ42 peptide , in the absence of effects on APP processing . Furthermore , this reduction in the levels of Aβ42 peptide by GSK-3 inhibition is most likely not mediated by tau , since loss of tau function partially rescued Aβ toxicity , but did not affect Aβ42 levels ( Figure 10A and Figure S5 ) . These data again suggest that GSK-3 can modify Aβ42 toxicity via tau-independent mechanisms .
Glycogen synthase kinase-3 is increasingly thought to play a pivotal role in the pathogenesis of Alzheimer's disease , both as a regulator of the accumulation of Aβ peptide [12] , [17] , [47] and through its well-established role as a tau kinase [48]–[52] . Although previous studies in mice have suggested that GSK-3 alters Aβ levels via modulation of APP processing [12] , [17] , the direct effects of the enzyme on Aβ toxicity , and in the adult nervous system , have not been examined . We therefore performed a more direct analysis of the specific role of GSK-3 in regulating Aβ42 toxicity in adult neurons in vivo , by modulating its activity in an adult-onset Drosophila model of Alzheimer's disease . Our study shows for the first time that GSK-3 inhibition ameliorates Aβ42 toxicity in adult flies , and also highlights a novel mechanism of protection by which GSK-3 directly regulates Aβ42 levels in the absence of any effects on APP processing . We have generated an inducible Drosophila model of Alzheimer's disease . Over-expression of the Arctic Aβ42 peptide in adult fly neurons led to shortened lifespan , neuronal dysfunction and behavioural impairments . However , no neuronal loss was evident in the brains of these flies . This finding contrasts with previous reports that flies over-expressing Arctic Aβ42 , using a constitutive neuronal driver , develop vacuoles [28] , [33] , These contrasting results could be reconciled if neuronal loss upon Aβ42 over-expression is a consequence of developmental abnormalities in the Drosophila brain or if neuronal loss represents an end-stage event in response to Aβ42 toxicity , since vacuolation has been reported only under the most extreme conditions of expression , age and temperature , while neuronal toxicity in these models is already apparent under less stringent conditions [33] . Importantly , our findings agree with those of other studies demonstrating that neuronal loss is generally not evident in murine models of amyloidosis , such as in mice transgenic for the amyloid precursor protein [53] . Moreover , our study has provided direct evidence of neuronal dysfunction in response to Aβ42 , by electrophysiological methods , at a timepoint more suitable to understanding the early events that lead to neuronal decline in AD . GSK-3 activity was increased in our flies upon Arctic Aβ42 over-expression , as measured by reductions in phosphorylation of endogenous Sgg at the inhibitiory Ser9 site . This is consistent with previous observations showing that Aβ42 alters GSK-3 phosphorylation in cells [54] and in mice [14] . Contrary to these findings , however , one other study has shown that WT Aβ42 expression does not alter phosphorylation of Sgg in flies [55] . These differences may reflect varying mechanisms by which Arctic mutant Aβ42 and WT Aβ42 peptides modulate GSK-3 activity , or a distinction in the effect of expressing Aβ42 throughout development compared to adult-only expression . Hence , our observed increase in Sgg activity may reflect an age-dependent effect . We aimed , therefore , to further investigate the functional role of this kinase in mediating Aβ toxicity in our adult-onset AD model . GSK-3 plays an important role in neuronal development [56] , [57] . Previous analyses of the role of GSK-3 in amyloid toxicity using constitutive expression systems may , therefore , represent abnormal neuronal development in addition to the response to AD pathology in the adult period . Hence , we confined GSK-3 inhibition to the adult neurons of Arctic Aβ42-expressing flies , by over-expressing a dominant negative form of the Drosophila orthologue Sgg , using our inducible expression system , or we treated whole adult flies with the GSK-3 inhibitor , lithium . GSK-3 inhibition extended lifespan of Arctic Aβ42-expressing flies and suppressed the locomotor dysfunction caused by expression of the peptide in adult neurons . This is an important finding because it demonstrates a definitive role for GSK-3 in Aβ42 pathogenesis in adult flies . Moreover , we found that inhibiting Sgg in adulthood had no adverse effect on wild type flies . Hence , our study provides further support for the therapeutic potential of GSK-3 inhibitors in treating Alzheimer's disease . Further studies are required to test this potential in mammalian models , firstly to confirm the relevance of GSK-3 for direct regulation of Aβ42 toxicity and secondly to establish a therapeutic index for GSK-3 inhibition in the treatment of AD . Aβ42 has been shown previously to increase tau phosphorylation in cells [45] and in mice [5] , [43] . Correlative evidence has suggested that GSK-3 may mediate the effects of Aβ on tau phosphorylation , since Aβ increases GSK-3 activity [10] . Furthermore , GSK-3 and APP cause similar increases in tau phosphorylation and aggregation in tau over-expressing mice [14] . Because both tau and GSK-3 appeared to play a causal role in Arctic Aβ42 toxicity in our flies , we examined whether the protective effect of GSK-3 inhibition on Aβ toxicity might be mediated via alterations in phosphorylation of endogenous Drosophila tau . We found that Arctic Aβ42 over-expression and lithium treatment of Aβ42-expressing flies did not have an observable effect on overall tau phosphorylation , as revealed both by generic measures , and at two specific sites by examination of phosphorylation of fly tau at Ser262 and Ser356 epitopes . However , because the Ser 262 and 356 sites are also predominant in vivo substrates for MARK ( microtubule-associated protein ( MAP ) -microtubule affinity regulating kinase ) [58] , [59] , further investigation of more specific GSK-3 sites on Drosophila tau , using mass spectrometric methods , would provide a more definitive analysis of the role of tau phosphorylation by GSK-3 in mediating Aβ42 toxicity in the fly . Our data suggest that tau phosphorylation may not be the only mechanism by which Aβ42 exerts its toxic effect in Drosophila , since phosphorylation sites thought to be important for mediating Aβ42 effects on tau ( Thr212/Ser214 [43] , Thr231 [45] , Ser422 , Ser262 [44] ) are predominantly not conserved in the fly , or are not altered by Aβ42 over-expression in our AD model . These findings seem to contradict previous studies showing that Aβ42 increases phosphorylation of human tau in Drosophila [55] , [60] , and indicating that particular sites , such as Ser262 , are important in mediating Aβ toxicity in the fly [60] . This disparity could reflect a lack of conservation of the mechanisms through which Aβ42 regulates human and Drosophila tau proteins . However , the endogenous fly tau does appear to be an important downstream mediator of Aβ42 toxicity , since loss of tau function partially reduced Aβ42 pathology in our study . This protective effect of tau against Aβ42 toxicity , however , may have been masked by the locomotor dysfunction from reducing levels of tau itself in flies that do not over-express the Aβ42 peptide . As a previous study has reported similar tau-dependent neuropathological phenotypes in an APP-overexpressing mouse model of AD [8] , our findings suggest that the role of tau in amyloid toxicity is conserved over large evolutionary distances . We further examined the epistatic interaction between GSK-3 inhibition and tau loss of function in protecting against Aβ42 toxicity in our fly model . Lithium alone had a greater protective effect against Aβ42 toxicity than loss of tau function alone , and lithium could prevent Aβ42-induced dysfunction in the presence or absence of tau . This suggests that a large proportion of the protective effect of GSK-3 inhibition on Aβ42 toxicity is mediated via non-tau-dependent mechanisms and that tau is not required for this effect . We observed no additive effect of combining lithium treatment and tau loss of function in protection against Aβ42-induced pathology; however , this could have been a consequence either of toxicity of loss of tau in the absence of Aβ42 or of the level of protection afforded by lithium , which could have produced a ceiling effect . Our findings agree with other recently published studies , showing that loss of tau only partially protects against GSK-3-induced neuronal degeneration in adult mice [61] , suggesting that other non-tau-dependent mechanisms of GSK-3 neuro-toxicity exist . Previous studies have shown a reduction in Aβ load in mice as a result of GSK-3 inhibition , but this has been explained mainly by dysregulation of APP processing , either by increasing γ-secretase activity [12] or by increasing the phosphorylation of APP and therefore directing its subcellular location to sites of secretase activity [17] . Our inducible Drosophila model , however , expresses the Arctic Aβ42 peptide directly , thus circumventing the requirement for APP processing . We found that inhibition of GSK-3 caused a reduction in the level of Aβ42 peptide , but not in RNA transcript levels . Hence , although previous studies have indicated that GSK-3 does not affect Aβ degradation [17] , our findings demonstrate a novel effect of GSK-3 in Aβ metabolism , irrespective of APP processing , in the adult nervous system . Furthermore , this observation may , partially , explain the non-tau-dependent effect of GSK-3 in protecting against Aβ toxicity in our flies . Our study , therefore , implies that GSK-3 may increase Aβ degradation or clearance . Candidate in vivo Aβ degrading enzymes include neprilysin ( NEP ) [62] , insulin degrading enzyme ( IDE ) [63]–[64] and to a lesser extent endothelin converting enzymes ( ECE-1 , 2 ) [65] and plasmin [66] . Mice deficient in IDE [63] , [64] or NEP [62] display increased levels of Aβ peptides in the brain . Conversely , increasing expression and activity of NEP or IDE reduces the cerebral amyloid plaque burden observed in APP over-expressing mice [67] , further implying that these are predominant in vivo Aβ degrading enzymes . Direct interactions between GSK-3 and IDE or neprilysin activities in relation to Aβ degradation , however , have not been extensively investigated . Studies examining the effects of reduced insulin signalling on amyloid toxicity in APP over-expressing mice have reported either increased GSK-3 activity , reduced IDE activity and increased amyloidosis [68] or decreased GSK-3 activity , increased IDE expression and reduced amyloidosis [69] in the brain , suggesting an inverse correlation between GSK-3 and IDE activities in relation to Aβ metabolism . Other studies have reported no correlation between inhibition of GSK-3 activity and neprilysin levels in relation to reduced Aβ load in mice [17] , but NEP activity was not measured and may provide a more accurate indication of the role of this enzyme in Aβ metabolism downstream of GSK-3 . As most information from mouse models is correlative , however , further work is required to determine whether these degrading enzymes are direct mediators of Aβ degradation in response to GSK-3 inhibition , by modulating their activities in our inducible Drosophila model . Drosophila homologues of both IDE and NEP exist , and over-expression of both NEP [26] , [70] and IDE [71] have been shown to reduce Aβ induced neurotoxicity in flies . This suggests that these degradation mechanisms may be generally conserved , and that the fly is a valuable model for the direct analysis of these genetic interactions with respect to the role of GSK-3 in AD . The more general proteosome degradation pathway could also play a role in regulating Aβ degradation or sequestration . Heat shock protein 90 ( Hsp90 ) , a protein chaperone involved in the proteosome degradation pathway , is thought to phosphorylate GSK-3 and regulate its activity [13] . In addition , an increase in levels of GSK-3 down-regulates the transcriptional activity of Heat shock factor-1 ( HSF-1 ) and Hsp70 [72]; thus a decrease in GSK-3 activity could lead to an increase in levels/activity of these chaperone molecules and augment the levels of Aβ . In a Caenorhabdits elegans worm model of AD , the aggregation-mediated Aβ42 toxicity was regulated by modulating the levels of hsf-1; a reduction in hsf-1 increased paralysis in these worms , suggesting a role of hsf-1 in the dis-aggregation of Aβ toxic oligomers [73] . Thus , any of these pathways/molecules could play a role in affecting Aβ load in our fly model and will require detailed exploration in the future . Our data highlight that this fly model is suitable for the study of AD pathology , since we observe neuronal dysfunction and toxicity that are particular to the expression of Aβ42 peptide . Furthermore , we have been able to test the amyloid cascade hypothesis in part , to show that tau is acting downstream of Aβ pathology . This inducible model of AD will also open the way to understanding the role of events at different ages and of the ageing process itself in the biological pathway leading to this ageing related disease . We have shown the involvement of GSK-3 , particularly in adulthood , in AD pathogenesis , and also uncover a novel mechanism by which GSK-3 could be acting directly or indirectly on Aβ , by reducing Aβ load . These results raise new potential therapeutic benefits of GSK-3 in AD pathology .
All fly stocks were maintained at 25°C or 27°C on a 12∶12-h light∶dark cycle at constant humidity on a standard sugar-yeast ( SY ) medium ( 15gl−1 agar , 50 gl−1 sugar , 100 gl−1 autolysed yeast , 100gl−1 nipagin and 3ml l−1 propionic acid ) . Adult-onset neuronal-specific expression of Arctic mutant Aβ42 peptide or constitutively active Sgg was achieved by using the elav GeneSwitch ( elavGS ) -UAS system [GAL4-dependant upstream activator sequence; [30]] . ElavGS was derived from the original elavGS 301 . 2 line [30] and obtained as a generous gift from Dr H . Tricoire ( CNRS , France ) . UAS-ArcAβ42 and UAS-SggS9E were obtained from Dr D . Crowther ( University of Cambridge , UK ) and the Bloomington Drosophila Stock Centre respectively . Tau dfc ( 9530 ) and EP line 3203 ( 17098 ) were received from Bloomington Drosophila stock centre . The EP line orientation is opposite to tau expression and causes a reduction in tau expression [46] . elavGS and UAS-lines used in all experiments were backcrossed six times into the w1118 genetic background . Male flies expressing UAS-constructs were crossed to female flies expressing elavGS , and adult-onset neuronal expression induced in female progeny by treatment with mifepristone ( RU486;200mM ) added to the standard SY medium . Lithium Chloride was made at 1M concentration and added to 200mM RU486 standard SY medium at a final concentration of 30mM or 100mM . For all experiments , flies were raised at a standard density on standard SY medium in 200 mL bottles . Two days after eclosion once-mated females were transferred to experimental vials containing SY medium with or without RU486 ( 200mM ) at a density of 10 flies per vial . Deaths were scored almost every other day and flies were transferred to fresh food three times a week . Statistical analyses were performed using JMP ( version 7 . 0 ) software ( SAS Institute , Cary , NC , USA ) . Data are presented as survival curves and analysis was performed using log-rank tests to compare between groups . To characterise the adult-onset behavioural effects of Arctic Aβ42 peptide on neuronal function , climbing assays were initially performed at 25°C according to previously published methods [75] . Climbing ability was analysed every 2–3 days post-RU486 treatment . Briefly , 15 adult flies were placed in a vertical column ( 25cm long , 1 . 5cm diameter ) with a conic bottom end , tapped to the bottom of the column , and their subsequent climb to the top of the column was analysed . Flies reaching the top and flies remaining at the bottom of the column after a 45 sec period were counted separately , and three trials were performed at 1 min intervals for each experiment . Scores recorded were the mean number of flies at the top ( ntop ) , the mean number of flies at the bottom ( nbottom ) and the total number of flies assessed ( ntot ) . A performance index ( PI ) defined as ½ ( ntot+ntop−nbottom ) /ntot ) was calculated . Data are presented as the mean PI ± SEM obtained in three independent experiments for each group , and analysis of variances ( ANOVA ) and post hoc analyses were performed using JMP 7 . 0 software . To assess various modifiers of the adult-onset neuronal dysfunction in Arctic Aβ42 over-expressing flies , a less stringent climbing assay was performed at 27°C . Thirty flies expressing Arctic Aβ42 with or without co-expressing modifiers in the neurons ( elav-GAL4GS ) were used for the climbing assay , adapted from Ganetzky et al . [76] . Climbing performance ( ability to climb past a 5cm mark in 18s ) was assessed after eclosion post RU486 treatment . Recordings from the GFS of adult flies were made essentially as described in [77]; a method based on those described by Tanouye and Wyman ( 1980 ) [78] and Gorczyca and Hall ( 1984 ) [79] . Flies were anaesthetized by cooling on ice and secured in wax , ventral side down , with the wings held outwards in the wax . A tungsten earth wire ( ground electrode ) was placed into abdomen; tungsten electrodes were pushed through the eyes and into the brain to deliver a 40V pulse for 0 . 03ms using a Grass S48 stimulator . Recordings were made from the TTM and contralateral DLM muscle using glass microelectrodes ( resistance: 40–60 MΩ ) . The electrodes were filled with 3M KCl and placed into the muscles through the cuticle . Responses were amplified using Getting 5A amplifiers ( Getting Instruments , USA ) and the data digitized using an analogue-digital Digidata 1320 and Axoscope 9 . 0 software ( Axon Instruments , USA ) . For response latency recordings , at least 5 single stimuli were given with a 5s rest period between each stimulus; trains of 10 stimuli , at either 100Hz , 200 Hz or 250Hz , were given a 5s rest interval between each train . For neuronal cell loss , adult heads were fixed , dehydrated and transverse sections were stained with Toluidine blue . Cell bodies were then counted blind for each genotype . Total RNA was extracted from 20–25 fly heads using Trizol ( GIBCO ) according to the manufacturers' instructions . The concentration of total RNA purified for each sample was measured using an eppendorf biophotometer . 1µg of total RNA was then subjected to DNA digestion using DNAse I ( Ambion ) , immediately followed by reverse transcription using the Superscript II system ( Invitrogen ) with oligo ( dT ) primers . Quantitative PCR was performed using the PRISM 7000 sequence-detection system ( Applied Biosystems ) , SYBR Green ( Molecular Probes ) , ROX Reference Dye ( Invitrogen ) , and Hot Star Taq ( Qiagen , Valencia , CA ) by following manufacturers' instructions . Each sample was analysed in triplicate with both target gene ( Arctic Aβ42 ) and control gene ( RP49 ) primers in parallel . The primers for the Aβ transgenes were directed to the 5′ end and 3′ end of the Aβ coding sequence: forward GATCCTTCTCCTGCTAACC; reverse CACCATCAAGCCAATAATCG . The RP49 primers were as follows: forward ATGACCATCCGCCCAGCATCAGG; reverse ATCTCGCCGCAGTAAACG . To extract total Aβ42 , five fly heads were homogenised in 50 µl GnHCl extraction buffer ( 5 M Guanidinium HCl , 50 mM Hepes pH 7 . 3 , protease inhibitor cocktail ( Sigma , P8340 ) and 5mM EDTA ) , centrifuged at 21 , 000 g for 5 min at 4°C , and cleared supernatant retained as the total fly Aβ42 sample . Alternatively , soluble and insoluble pools of Aβ42 were extracted using a protocol adapted from previously published methods [80]: 200 fly heads were homogenised in 200 µl tissue homogenisation buffer ( 250mM sucrose , 20mM Tris base , 1mM EDTA , 1mM EGTA , protease inhibitor cocktail ( Sigma ) then mixed further with 200µl DEA buffer ( 0 . 4% DEA , 100mM NaCl and protease inhibitor cocktail ) . Samples were centrifuged at 135 , 000 g for 1 hour at 4°C ( Beckman Optima Max centrifuge , TLA120 . 1 rotor ) , and supernatant retained as the cytosolic , soluble Aβ42 fraction . Pellets were further resuspended in 400µl ice-cold formic acid ( 70% ) , and sonicated for 2×30 sec on ice . Samples were re-centrifuged at 135 , 000 g for 1 hour at 4°C , then 210 µl of supernatant diluted with 4ml FA neutralisation buffer ( 1M Tris base , 0 . 5M Na2HPO4 , 0 . 05% NaN3 ) and retained as the insoluble , formic acid-extractable Aβ42 fraction . Total , soluble or insoluble Aβ42 content was measured in Arctic mutant Aβ42 flies and controls using the hAmyloid β42 ELISA kit ( HS ) ( The Genetics Company , Switzerland ) . Total Aβ42 samples were diluted 1∶100 , and soluble versus insoluble Aβ42 samples diluted 1∶10 in sample/standard dilution buffer , then ELISA performed according to the manufacturers' instructions . Protein extracts were quantified using the Bradford protein assay ( Bio-Rad protein assay reagent; Bio-Rad laboratories Ltd ( UK ) ) and the amount of Aβ42 in each sample expressed as a ratio of the total protein content ( pmoles/g total protein ) . Data are expressed as the mean ± SEM obtained in three independent experiments for each genotype . ANOVAs and Tukey's-HSD post-hoc analyses were performed using JMP 7 . 0 software . Fly heads were homogenised in Mes buffer ( 100 mM Mes , pH 6 . 5 , 1 M NaCl , 0 . 5 mM MgCl2 , 1 mM EGTA , 10 mM NaF , Protease inhibitor cocktail [Sigma , P8340] and Phosphatase inhibitor cocktail 2 [Sigma , P5726] ) , centrifuged at 20 , 000 g for 30 minutes at 4°C , then supernatants adjusted to 0 . 5% β-mercaptoethanol , boiled for 5 minutes and re-centrifuged . Cleared supernatants were then retained as heat-stable soluble tau . Control samples were dephosphorylated by incubating with λ-protein phosphatase ( NEB , P0753 ) for 3 hours at 30°C . To detect phosphorylated and non-phosphorylated tau , samples were separated by SDS-PAGE , using Phos-tag AAL-107 ( FMS laboratory ) according to the manufacturers' instructions . Western blotting was then performed using a non-phosphorylation dependent Drosophila rabbit anti-tau antibody ( 1∶5000 ) . Quantitation was performed using Image J software ( National Institutes of Health ) . Phosphorylated and non-phosphorylated tau was expressed as a percentage of the total amount of tau present in each sample . ANOVA and Tukey's-HSD post-hoc analyses were performed using JMP 7 . 0 software . Drosophila heads were pooled and homogenised in 2×Laemmli sample buffer containing β-mercaptoethanol and boiled for 10 mins . Proteins were separated on SDS polyacrylamide gels and blotted onto nitrocellulose . Membranes were incubated in a blocking solution containing 5% milk proteins either in TBST for 1hr at room temperature for Tau blot or in PBST for 20 min for Sgg blot , then probed with primary antibodies overnight at 4°C . Mouse anti-actin antibody ( Santacruz ) was used at a 1 in 1000 dilution . Tau antibody ( rabbit ) was a kind gift from Nick Lowe , and used at a 1 in 2000 dilution . Rabbit anti-phospho S262 and S356 ( Abcam ) were used at a dilution of 1∶1000 in 5% BSA TBST . Quantitation was performed using Image J software . Rabbit anti-phospho S9/S21 GSK-3 antibody ( AB 9331 Cell Signaling ) at a 1 in 250 dilution was detected using an anti-rabbit 800 nm flourophore conjugate ( Rockland , USA ) . Monoclonal antibody , pan Sgg mouse monoclonal ( 4G1G ) at a 1 in 2000 dilution was a kind gift from Marc Bourouis , which was subsequently detected using a 680 nm anti-mouse flourophore conjugate ( Invitrogen , UK ) . Membranes were sequentially scanned at 700 and 800 nm using a Licor Odyssey infrared scanner . Densitometric measurements were taken in both wavelengths . Relative phospho-serine 9 Sgg levels were determined by dividing the signal at 800 nm by that obtained at 700 nm . Details of secondary antibodies and Odyssey analysis have been previously described in [69] .
|
Alzheimer's disease ( AD ) is the leading cause of dementia in the ageing population . Symptoms include memory loss and decline in understanding and reasoning . Alois Alzheimer , who reported the first case of AD , observed plaques and tangles in the brains of patients . The plaques are made up of amyloid protein , while the tangles are of tau protein . One of the main scientific ideas about AD is that it starts with build-up of amyloid , which then alters tau protein , causing the disease . Another protein , called GSK-3 , also seems to play a part . Simple invertebrates such as flies are useful for understanding human diseases . We have created an AD model in the fruit fly Drosophila where amyloid protein is present in the nerve cells of the adult fly; this caused the flies to be impaired in their survival , nerve function , and behavior . We found that amyloid increased the activity of GSK-3 , and so we experimentally turned down its activity and found that this improved the survival and behavior of the flies . Importantly , turning down the activity of GSK-3 in flies that did not have amyloid did not seem to harm them . GSK-3 could therefore be a good target for drugs against AD .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/disease",
"models",
"geriatrics/dementia",
"genetics",
"and",
"genomics/animal",
"genetics",
"neurological",
"disorders/alzheimer",
"disease"
] |
2010
|
Inhibition of GSK-3 Ameliorates Aβ Pathology in an Adult-Onset Drosophila Model of Alzheimer's Disease
|
The metabolism of microorganisms is regulated through two main mechanisms: changes of enzyme capacities as a consequence of gene expression modulation ( “hierarchical control” ) and changes of enzyme activities through metabolite-enzyme interactions . An increasing body of evidence indicates that hierarchical control is insufficient to explain metabolic behaviors , but the system-wide impact of metabolic regulation remains largely uncharacterized . To clarify its role , we developed and validated a detailed kinetic model of Escherichia coli central metabolism that links growth to environment . Metabolic control analyses confirm that the control is widely distributed across the network and highlight strong interconnections between all the pathways . Exploration of the model solution space reveals that several robust properties emerge from metabolic regulation , from the molecular level ( e . g . homeostasis of total metabolite pool ) to the overall cellular physiology ( e . g . coordination of carbon uptake , catabolism , energy and redox production , and growth ) , while allowing a large degree of flexibility at most individual metabolic steps . These properties have important physiological implications for E . coli and significantly expand the self-regulating capacities of its metabolism .
Metabolism is a fundamental biochemical process that converts nutrients into energy and biomass precursors , thus enabling cells to maintain their structures , grow , and respond to their environment . While the topology of metabolic networks is fairly well known , understanding how metabolic behaviours emerge from the dynamic interactions of their molecular components remains one of the main challenges faced by systems biology and is crucial for the development of synthetic biology [1] . The operation of metabolic networks , i . e . the metabolic fluxes , represents the ultimate output of several regulatory mechanisms . Metabolic fluxes are functions of enzyme activities and of the concentrations of reactants , products , and other effectors . While the enzyme activities are the ultimate outcome of gene expression through the hierarchy of transcriptional , post-transcriptional , translational and post-translational regulatory mechanisms , the reactant and effector concentrations are directly regulated at the metabolic level by enzyme activities themselves . Hierarchical regulation , and in particular transcriptional regulation , has attracted much attention because of mature experimental methods , but also because of early examples of flux increase with enzyme induction [2 , 3] . These studies suggested an intuitive picture where fluxes mainly depend on enzyme concentrations , themselves mainly dependent on the level of transcript—a view that puts genes and their regulation at the top of a hierarchy of control and that regards metabolism as mostly a consequence of gene expression . An increasing body of evidence , however , indicates that this view of a hierarchical ( or “dictatorial” ) regulation of metabolism by gene expression is too simplistic . Large-scale 13C-flux analyses revealed that flux distributions in Saccharomyces cerevisiae and Escherichia coli are incredibly robust to the deletion of global transcriptional regulators [4 , 5] . Integration of transcript and enzyme abundances with fluxes measured under different environmental conditions indicated that hierarchical regulation is insufficient to explain most of the flux reorganizations [6–9] . Therefore metabolism can no longer be seen as a passive process primarily regulated at the hierarchical level , but rather that it plays an active role in the control of its own operation via a dense network of metabolite-enzyme interactions . However , because hundreds of these interactions simultaneously regulate fluxes , which in turn affect metabolite levels , the system-wide role of metabolic regulation is hard to dissect and so far remains largely uncharacterized . Mathematical frameworks such as metabolic control analysis [10 , 11] were developed to analyze such complexity and improve our understanding of the role of each interaction on the metabolic network . These frameworks are particularly useful when applied to ( validated ) kinetic models that quantitatively describe the mechanistic interactions between the molecular species and their dynamics . However , developing models that are truly representative of real cell metabolism requires large amounts of experimental data to establish complex rate laws and identify parameters for each interaction [12] . Most models have focused on single pathways or on small sub-systems [e . g . 13 , 14–18] . Such models accurately predict the response of those pathways to perturbations and reveal insights on the role of particular regulatory interactions on the metabolic operation [19–21] . Great progress has recently been made to develop larger scale kinetic models using top-down approaches [22–29] , hence paving the way towards comprehensive understanding of the role of metabolic regulation at the whole cell level . These large-scale kinetic models highlighted the system-wide impact of local properties on the functioning of metabolic networks , such as an improved metabolic flexibility caused by enzyme saturation [26] . However , these large scale models are typically constructed from whole-genome metabolic reconstructions using generic rate laws , and contain a low level of mechanistic details ( in particular are mostly devoid of allosteric regulation ) . An alternative approach constructed a highly detailed model of an entire cell of Mycoplasma genitalium [30] , but unfortunately while this model represents a considerably high level of mechanistic detail in many cellular processes , it entirely lacks metabolic regulation ( as it uses dynamic flux balance analysis rather than a mechanistic kinetic model ) . Hence , many of the properties that emerge from metabolic regulation are not captured by current large-scale models . In this study , we aim at investigating the role of metabolic regulation on the central metabolic network of E . coli , which constitutes the backbone of its metabolism by providing macromolecular precursors , reducing equivalents , and energy for growth and maintenance . While previous studies typically focused on the role of particular regulatory interactions , we attempt to determine whether more global and generic properties arise from the interplay of the many regulatory interactions that compose metabolic regulation . To accomplish this , a kinetic model of E . coli central carbon and energy metabolism was developed and validated against a large set of existing experimental data . This model includes more mechanistic details than previous ones , and the impact of metabolic regulation on this system was analyzed using local and global methods .
The kinetic model developed in this study represents the central metabolism of Escherichia coli cultivated on glucose under aerobic conditions ( Fig 1 ) . This model contains 3 compartments ( environment , periplasm and cytoplasm ) , 62 metabolites , and 68 reactions which represent the main central carbon and energy pathways of E . coli , namely: glucose phosphotransferase system ( PTS ) , glycolysis and gluconeogenesis ( EMP ) , pentose phosphate ( PPP ) and Entner-Doudoroff ( EDP ) pathways , anaplerotic reactions ( AR ) , tricarboxylic acid cycle ( TCA ) , glyoxylate shunt ( GS ) , acetate metabolism ( AC ) , nucleotide interconversion reactions ( NC ) and oxidative phosphorylation ( OP ) . A reaction was also included to account for the consumption of metabolic precursors , reducing equivalents , and energy , and thus linking metabolism to cell proliferation . To account for metabolic regulation , a total of 255 metabolite-enzyme interactions ( i . e . where metabolites modulate the reaction rates through thermodynamic or kinetic regulation , such as being substrates , products , allosteric modulators , or other type of inhibitors or activators ) were included in the model , amongst which 34 are long-range regulatory interactions ( i . e . where certain metabolites , which are not reactants , modulate the rates of these reactions ) . Previously published kinetic models of E . coli metabolism were used as scaffolds to construct this model [18 , 31 , 32] . Both the number of pathways and the level of mechanistic detail were increased in the present model ( S1 Table ) . For instance , this model now couples carbon metabolism with a detailed representation of oxidative phosphorylation , which makes it possible to balance the concentrations of cofactors ( ATP/ADP/AMP , NAD ( P ) ( H ) and FAD ( H2 ) ) and simulate energy and redox metabolism . Previous models accounted for the consumption of metabolic precursors for growth in a decoupled way . This may be enough from the point of view of mass balance , but results in artifacts if used for an understanding of dynamics and regulation . In contrast the present model includes a single reaction to model growth , which ensures that the building blocks are consumed in stoichiometric proportions fixed by the cell composition , and not independently from each other . The rate of this reaction is a function of the intracellular concentrations of all the building blocks . This represents a significant improvement by satisfying the following growth rate properties: i ) it monotonically increases with the availability of each building block , ii ) it is asymptotically independent of each pool above a saturating concentration , and iii ) it approaches zero if any pool approaches zero [21] . These properties were not reflected in the previous models [33] . The present model was calibrated to represent the metabolic state of E . coli cultivated under carbon limitation , a condition frequently experienced by this bacterium in laboratories , in industrial bioprocesses , and likely in its natural environment . To the extent possible , values of the biochemical parameters were taken from experimental determinations available in the literature . Parameters not available in the literature were estimated to reproduce steady-state and time-course experimental data obtained from a unique E . coli strain ( the model strain K-12 MG1655 ) grown under a unique reference condition ( M9 minimal medium with glucose as sole carbon source , dilution rate = 0 . 1 h-1 , temperature = 37°C , pH = 7 . 0 , pO2 > 20% ) [13 , 34–37] . This step is critical since both metabolite concentrations and fluxes depend on environmental conditions and differ between strains [38–40] . While results described below are largely in agreement with other experimental observations , the model was not forced to reproduce them , providing an important validation of the model . Detailed information on the construction and validation of the model is given in the Methods section and Supporting Information ( S1 Text ) . The model is included in Supporting Information ( S1 Model ) formatted in SBML [41] and COPASI [42] formats , and is available from the BioModels database [43] with accession number MODEL1505110000 . The control properties of E . coli central metabolism in the reference state ( see above ) were investigated under the metabolic control analysis framework [10 , 11] . Flux ( CEJ ) and concentration ( CEM ) control coefficients quantify the impact of a small change in the rate of each reaction ( e . g . through change in the enzyme concentration E ) , on each flux ( J ) and each metabolite concentration ( M ) . Since each metabolic step affects all fluxes and concentrations to some extent , we calculate a metric of its overall control on fluxes and concentrations as the L2 norm of all its flux- and concentration-control coefficients ( see Methods ) , respectively . The overall flux- and concentration-control by each step in the network is displayed in Fig 2 . The main control point is the glucose inflow reaction with a control of 8 . 7 on fluxes and 5 . 3 on concentrations , using this summary metric . The system is therefore sensitive to its environment , as expected . Note that this is a direct sensitivity of metabolism to the environment , not through the ( hierarchical ) action of signal transduction and gene expression , which is not represented in this model; if it were its effect would thus be overlaid ( likely with a delay ) on the direct effect displayed in our model . Reactions that were identified by previous models as exerting a strong flux control under similar environmental conditions , such as the glucose phosphotransferase reactions or phosphofructokinase [13 , 32 , 44 , 45] , showed low control in our model ( respectively 0 . 1 and 0 . 8 ) . Rather , consistently with experimental evidence ( see for example [46 , 47–49] ) , the flux control was predicted to be shared between enzymes of all the pathways , amongst which cytochrome bo oxidase ( reaction CYTBO , with 4 . 7 overall flux control ) , glucose-6-phosphate dehydrogenase ( ZWF , 3 . 9 ) , glyceraldehyde-3-phosphate dehydrogenase ( GDH , 2 . 6 ) , citrate synthase ( GLT , 2 . 9 ) and the anabolic machinery ( GROWTH , 1 . 4 ) are the ones with the largest share . A similar situation was observed for the control of concentrations , which is widely distributed across the network , and with the environment as the strongest control . In fact , a significant correlation ( Pearson R = 0 . 86 , P-value = 10−15 ) can be observed between the overall flux- and concentration-control exerted by each step ( Fig 2C ) , indicating that , in general , enzymes which exert the strongest control on fluxes also exert the strongest control on concentration . A global sensitivity analysis [50] shows that these conclusions are robust with regard to parameter uncertainties ( Fig 2A and 2B ) . Further analysis confirmed the wide distribution of flux control across all the enzymes , with 97% of the individual control coefficients between -0 . 3 and 0 . 3 ( Fig 2D ) . These observations , in agreement with the view that “rate-limitation” is distributed across the network and is variable [51] , explain why fluxes are robust to moderate or even large changes of enzyme levels . The fluxes in this reference state are more sensitive to the environment , with 75% of the control coefficients exerted by the glucose supply reaction higher than 0 . 3 ( Fig 2E ) . Similar conclusions were reached regarding the control of metabolite concentrations , which is distributed across the network ( Fig 2B ) , with 97% of the control coefficients exerted by enzymes between -0 . 3 and 0 . 3 ( Fig 2F ) and 49% of the control coefficients exerted by the glucose supply reaction higher than 0 . 3 ( Fig 2G ) . Despite the low control exerted by enzymes over fluxes and concentrations at the network level , a detailed analysis of flux control coefficients reveals generic regulatory patterns between most of the pathways ( Fig 3 ) . A general observation is that the control of each pathway resides largely outside of itself . For example , the control of the partition of carbon through competing pathways is shared between enzymes of each pathway . The glycolytic phosphofructokinase ( PFK ) exerts a ( small ) negative control on the PPP and ED fluxes ( CPFKZWF=−0 . 15 ) and a positive control on the glycolytic flux ( CPFKPGI=0 . 05 ) , while the glucose-6-phosphate dehydrogenase ( ZWF ) of the PPP and ED pathways exerts a strong positive control on its own flux ( CZWFZWF=0 . 75 ) and a negative control on the glycolytic flux ( CZWFPFK=−0 . 12 ) . Similar behavior is observed at the main metabolic branch nodes , e . g . between the TCA cycle and the glyoxylate shunt or between the pentose phosphate and Entner-Doudoroff pathways . It is important to note that the fraction of flux diverted to each branch does not depend only on the local enzyme kinetics , contrary to what is sometimes suggested [6] , but on several enzymes of each of the competing pathways . Several feedforward and feedback interactions are also observed between the pathways . For instance , the pyruvate kinase ( PYK ) controls fluxes through the TCA cycle ( CPYKSDH=0 . 07 ) , and is controlled by some TCA reactions ( CLPDPYK=0 . 06 , CSDHPYK=0 . 19 ) . Similarly , the ATP demand ( that can be represented by the ATP utilization for maintenance , ATP_NGAM ) is activated by glycolysis ( CGDHATP_NGAM=0 . 26 ) and exerts in turn a positive feedback control on this pathway ( CATP_NGAMPGI=0 . 12 ) and a negative control on growth ( CATP_NGAMGROWTH=−0 . 09 ) , as observed in vivo [72] . Interestingly , biomass synthesis ( GROWTH ) is strongly controlled by the upstream glucose supply ( CGLC_INFLOWGROWTH=1 . 5 ) , with all other control coefficients lower than 0 . 13 . In turn , biomass synthesis exerts a small but global feedback control on most catabolic fluxes . Those several , intertwined feedback and feedforward interactions stress the high degree of functional organization of the central carbon and energy metabolism . This may be an important feature to maintain the coordination between the different pathways at the cellular level: if the rate of a particular reaction—be it upstream or downstream—is affected by a perturbation , this information will be transmitted from this “sensor reaction” to the entire system , resulting in a global response . Note that this response is sensed in a very short time scale , rather than the slower response that happens after signal transduction and consequent changes in gene expression . To get a broader picture of the role of metabolic regulation on the coordination of E . coli metabolism , the solution space of this network was explored with and without considering metabolic regulation . Two versions of the model were used: the kinetic version which accounts for metabolic regulation , and a stoichiometric version of the same model which contains only stoichiometric constraints ( and is thus similar to a flux balance analysis model ) . The solution space of each model was explored using a random sampling approach: 600 , 000 flux distributions were uniformly sampled from the solution space using the stoichiometric model , and steady-states were simulated for 600 , 000 sets of random enzyme levels using the kinetic model . For each set , enzyme levels ( i . e . Vmax ) were sampled from a log uniform distribution ( between 0 . 1 and 10 times the enzyme levels of the initial model ) to ensure each order of magnitude to be sampled in similar proportions . It is important to mention that cells do not express enzymes levels according to the distribution generated , therefore the distribution of the variables is not expected to provide any information on the probability for a cell to reach a specific state in vivo [50] . Rather , uniformity is used to clearly grasp the functional implications of applying metabolic regulation to the network . We first investigated the relationship between supply ( glucose uptake ) and demand ( growth ) , which provides information on the allocation of resources by the metabolic network [52] . Direct sampling of the solution space ( Fig 4A ) revealed that most of the metabolic states are not efficient in term of resource allocation: most of them correspond to a high glucose uptake rate , but with a low growth rate , because this situation significantly increases the attainable intracellular flux states . Interestingly , the opposite picture is observed when metabolic regulation is applied on this network ( Fig 4B ) : a smaller region of the solution space is reached , where the growth rate is now coupled to the glucose uptake rate . To evaluate this prediction quantitatively , we gathered from the literature experimental data obtained from 254 growth experiments carried out under similar environmental conditions ( glucose as sole carbon source in aerobic conditions ) [5 , 27 , 34 , 53–67] ( S2 Dataset ) . These data were collected from “wild-type” and mutant strains obtained by deletion or overexpression of central metabolic enzyme genes or global regulators of gene expression , and cultivated under a wide range of experimental conditions . Therefore , these 254 data represent a very broad range of the metabolic states that can be expressed by E . coli growing on glucose . Importantly , these data were not used for parameter estimation , thereby they constitute an independent validation and provide a robust assessment of the predictive ability of the model . These experimental observations correspond to the region of the solution space less frequently sampled using the stoichiometric model , but they closely match the region sampled by the kinetic model ( Fig 4B ) . This means that the observed physiology of E . coli is closer to the metabolic model that is regulated by metabolite-enzyme interactions ( the kinetic model ) than it is to a metabolic model that would be regulated by gene expression alone ( the stoichiometric model ) . Hence , metabolic regulation alone , without needing to invoke coordinated expression of genes , seems to be sufficient to explain the emergence of a coupling between anabolic ( growth ) and catabolic ( glucose uptake ) fluxes , and thereby appears to be a major determinant of the overall cellular physiology by ensuring an efficient and robust allocation of nutrients towards growth . We extended the above analysis to determine whether additional couplings emerge from metabolic regulation . Several variables representative of the physiological state of E . coli were calculated for each steady-state reached by the kinetic model , namely: growth and glucose uptake rates , ATP , NADH and NADPH production rates , sum of all intracellular fluxes , sum of all intracellular metabolite concentrations and cost of enzymes ( defined as the product of enzyme concentration and number of amino acids of the corresponding enzyme , summed over all reactions , as detailed in Methods ) . Additional variables derived thereof were also computed: biomass , ATP , NADH and NADPH yields , enzyme cost and ATP production rate per sum of fluxes , and sum of fluxes per glucose consumed . Pairwise relationships between systemic variables and ( absolute and relative ) fluxes through the main pathways were examined using Spearman correlation and mutual information . The outcome is a correlation matrix which maps the degree of functional coupling between all the variables ( Fig 5A ) . The same patterns were highlighted by both methods , which indicate that these couplings are monotonic since mutual information , but not Spearman correlation , would identify non-monotonic relations . A systematic positive correlation was predicted between some anabolic and catabolic fluxes and yields: glucose uptake rate , growth rate , NADPH production rate , biomass yield and NADPH yield ( with ρ > 0 . 85 ) , indicating a considerable degree of coordination in the metabolic operation . These variables are negatively correlated with the energetic ( ATP and NADH ) yields ( ρ < -0 . 85 ) , which is consistent with the fact that the single carbon source , glucose , is used by two competing metabolic processes: energy production and biomass synthesis , reflected in an increase in YATP and increase in biomass yield , respectively . The oxygen uptake rate also correlated positively with ATP and NADH production ( with ρ > 0 . 90 ) , reflecting the important role of oxidative phosphorylation in energy production under aerobic conditions . The low correlation coefficient between the sum of fluxes and the cost of enzymes ( ρ = 0 . 11 ) indicates that the sum of fluxes cannot be considered as a proxy for enzyme investment per se . The outcome of predictive analyses based on this assumption ( such as the minimization of the sum of fluxes in FBA according to the hypothesis that cells minimize their enzyme levels ) should therefore be interpreted with caution . In general , systemic variables correlated poorly with relative and absolute fluxes from most of the pathways . This is interesting as it shows that while there is coordination between several processes , there is nevertheless a significant degree of flexibility in the intracellular flux distribution . A notable exception was observed for the TCA cycle: its absolute flux is positively correlated with catabolic and energy fluxes ( vO2_uptake , vGlc_uptake , vNADH_production , vATP_production , with ρ > 0 . 84 ) , and the relative contribution of this pathway negatively correlated with anabolic rates and yields ( growth and NADPH production rates , ρ < -0 . 92 ) and positively correlated with energy yields ( ρ > 0 . 90 for ATP and NADH yields ) . Thus , the partition of carbon between energy production ( ATP and NADPH ) and growth ( via the synthesis of many anabolic precursors ) is predicted to be realized primarily at the level of the TCA cycle and appears to be largely controlled at the metabolic level . To evaluate these model predictions , additional experimental data on extracellular and intracellular fluxes ( growth rate , glucose and oxygen uptake rates , and TCA cycle fluxes through the citrate synthase ) were collected from the literature [5 , 27 , 34 , 53 , 54 , 56 , 58–60 , 62–65 , 67] ( S2 Dataset ) . These data , which were not used to calibrate the model , covered the particular regions highlighted by the kinetic model ( Fig 5B–5D ) . The excellent agreement between the spread of simulated and experimental data strongly supports the existence of the functional couplings predicted by the model . It is important to mention that these couplings are not caused by stoichiometric constraints since they are not observed when the solution space is uniformly sampled using the stoichiometric model ( Fig 5E–5G ) . The results also show that the coordination of gene expression by hierarchical regulatory mechanisms is not an important factor in these couplings since they are still maintained when enzyme levels are changed randomly . In contrast , metabolic regulation brought about by metabolite-enzyme interactions is sufficient to explain their emergence; therefore they represent intrinsic properties of the central metabolism of E . coli . Interestingly , additional couplings predicted by the model were recently observed in vivo in both prokaryotic ( E . coli ) and eukaryotic ( S . cerevisiae ) microorganisms [73]: between the ATP and NADH production rates ( ρ = 0 . 93 , Fig 6A ) , between the sum of fluxes per glucose uptake rate and the ATP yield ( ρ = 0 . 71 , Fig 6B ) , and between the growth rate per sum of fluxes and the sum of fluxes per glucose uptake rate ( ρ = -0 . 85 , Fig 6C ) . Since the central metabolic networks of E . coli and S . cerevisiae are highly conserved , the present results may explain why similar properties are observed in both microorganisms , though this hypothesis requires further investigation . The results presented above support the view that metabolic regulation reduces the solution space defined by the stoichiometric constraints , as previously suggested [68 , 69] . However , the very low probability regions of the solution space might not be captured by random sampling approaches [50] . To test further if metabolic regulation actually shrinks the solution space of E . coli central metabolism , its boundaries were determined with and without considering metabolic regulation by using the kinetic and the stoichiometric models , respectively . Unexpectedly , the boundaries were similar for both models ( Fig 7 ) . This indicates that metabolic regulation does not shrink the solution space of the system—and thus does not restrict the metabolic capabilities of E . coli– , at least for the variables considered here . Thus we have to conclude that evolution must have “discovered” this region of parameter space which gives selective advantage to the organism . It has been shown that metabolic regulation plays an important role in metabolite homeostasis , which prevents osmotic stress and disadvantageous spontaneous reactions by avoiding large changes in metabolite concentrations ( for example see [20 , 70] ) . Interestingly , we noticed that , for 86% of the steady-states reached by the kinetic model ( when the enzyme levels were chosen at random ) , changes in total concentration of metabolites are lower than three-fold relative to the calibrated state ( 47 mM ) ( Fig 8 ) , but the changes in fluxes are several orders of magnitude higher . This narrow range of predicted intracellular concentrations is physiologically relevant [63 , 71] . Since no constraints on metabolite concentrations were included in the model , we conclude that metabolic regulation alone may explain global metabolite homeostasis , while still allowing significant changes in fluxes .
In this study , we investigated the contribution of metabolic regulation on the operation of the central metabolism of E . coli , which provides building blocks , cofactors , and energy for growth and maintenance . We developed , to our knowledge , the first detailed kinetic model of this system that links metabolism to environment and cell proliferation through intracellular metabolites levels . This model , validated by 778 independent flux data from some 266 experiments , allowed the identification of several properties which emerge from metabolic regulation and explain many experimental observations of E . coli’s physiology . The intrinsic , self-regulating capacities of E . coli central metabolism appear to be far more significant than previously expected . The results presented here imply that gene regulation is not required to explain these properties . Metabolic control analysis showed that the flux and concentration control exerted by single enzymes is low and largely distributed across the network , confirming again the insights of Kacser and Burns [51] . This significantly contrasts with the outcome of previous kinetic models [13 , 32 , 44 , 45] , where a few enzymes were predicted to exert most of the flux control , but is in line with much experimental evidence [7 , 21 , 46 , 47 , 72] . Our results therefore support the view that the concept of “rate-limiting” steps does not apply to E . coli metabolism , and likely not to the metabolism of other organisms . Its persistence in the literature is a major handicap to understanding metabolism . In fact , the central metabolism is not even self-contained in terms of control due to a large portion of control being exerted by the environment , making E . coli responsive to environmental changes . One of the most striking examples of this phenomenon is manifested in growth controlling most fluxes but being controlled virtually by glucose availability alone . The low control exerted by single enzymes on the system makes the metabolic operation of E . coli robust to fluctuations of enzyme levels that may arise from noise in gene expression or other factors . Moreover , the majority of control resides not within but outside the controlled pathways . The dense , yet highly organized , interactions between pathways allow a rapid and coordinated response of the entire system to perturbations . Exploration of the solution space indicated that metabolic regulation does not significantly restrict the metabolic capabilities of E . coli , as was previously believed [68 , 69] . While the observed behavior of many different E . coli strains and mutants are confined to a small region of the solution space , this is not due to kinetic constraints as it is possible to simulate other behaviors simply by changing parameter values . This apparent paradox can be resolved , of course , if the action of natural selection had favored these behaviors . The systematic mapping of the relationships between various systemic variables revealed that metabolic regulation is sufficient to explain the emergence of several functional couplings , which are independent from gene regulation ( since they are conserved when enzyme levels are changed randomly by orders of magnitude ) and cannot be explained by stoichiometric constraints . An important finding is that metabolic regulation alone may be responsible for the coordination of major catabolic , energetic and anabolic processes at the cellular level to optimize growth . Metabolic regulation thus appears to be sufficient to maintain multi-dimensional optimality of E . coli metabolism [65] . Despite this overall coordination , there is a large degree of flexibility at most individual metabolic steps . The role of metabolic regulation in maintaining global homeostasis of intracellular metabolite pools under a broad range of flux states was also verified by the present model . The modeling results were in excellent agreement with experimental data , even quantitatively . E . coli metabolism displays remarkably robust yet simple emergent properties , and these properties have major implications on its overall cellular physiology , e . g . by preventing unnecessary osmotic stress , maintaining the coordination between key processes , and optimizing the allocation of resources towards particular functions such as growth . The self-regulating capabilities of E . coli central metabolism reflect the evolutionary selection that has been exerted on the ensemble of enzymes ( in terms of kinetic and regulatory properties , but not necessarily of expression levels ) to realize a network with these properties . Since central metabolism is essential in most organisms and is highly conserved across the three domains of life , it is tempting to speculate that metabolic regulation is responsible for the very similar operation principles observed in different organisms [73] . Of course we do not suggest that hierarchical regulation does not play an important role in the metabolic operation of E . coli , but it is in addition to the properties observed here , since these can operate without it . For instance , the robustness of the flux partition to the deletion of global transcriptional regulators was interpreted as a low control of this partition at the hierarchical level [5] , and our results confirm that this robustness lies , to some extent , in metabolic regulation , given the low control exerted by enzymes . However , this conclusion is valid only for moderate changes of enzyme levels ( with the notable exception of the flux through the TCA cycle ) , and other mechanisms ( such as hierarchical regulation ) are required to explain the robust flux partition . Expanding the kinetic model to incorporate regulation of gene expression will be needed ultimately to understand the interplay between these two regulatory levels [19 , 74 , 75 , 88] .
The kinetic model of the central carbon and energy metabolism of Escherichia coli K-12 MG1655 ( Fig 1 ) was developed with the software COPASI ( build 45 ) [42] . This model is briefly described in this section , and additional information can be found in Supporting Information ( S1 Text ) . The model is available in SBML and COPASI formats in Supporting Information ( S1 Model ) , as well as from the Biomodels database [43] with identifier MODEL1505110000 . Values of 56% of the parameters ( 253 on a total of 449 ) were directly taken from the literature . Parameters not available in the literature , which do not have a real biochemical estimate ( e . g . Michaelis constants of the biomass function ) , or for which biochemical measurements are generally not representative of intracellular conditions ( e . g . Vmax ) were estimated to reproduce in the best possible way 276 experimental data obtained from E . coli K-12 MG1655 grown on glucose , under aerobic condition , at a dilution rate of 0 . 1 h-1 . These data were steady state fluxes and metabolite concentrations [13 , 34 , 36 , 37 , 82] and time-course concentrations of intracellular metabolites in response to a glucose pulse [35] ( S1 Dataset ) . Parameter estimation was formulated as a constrained optimization problem: minimize f ( p ) subject to g ( p ) ≥c where p is the parameter vector , f is the objective function which evaluates the deviation between the simulated and measured data , g ( p ) is the constraint function , and c is the constraint vector . The objective function f was defined as the sum of squared weighted errors: f ( p ) =∑i ( xi−yi ( p ) σi ) 2 where xi is the experimental value of the data point i , with experimental standard deviation σi , and yi ( p ) is the corresponding simulated value . Constraints were defined on estimated parameters ( 10−4 mM ≤ KM ≤ 103 mM; 10−2 mM/s ≤ Vmax ≤ 103 mM/s; 10−4 ≤ Keq ≤ 106 ) to ensure they are kept within a biologically reasonable range . The objective function was minimized with the Particle Swarm Optimization algorithm [83] , using the Condor-COPASI system [84] on a pool of 2500 CPU cores . The experimental and fitted data are provided in Supporting Information ( S1 Dataset ) . Values of all the parameters ( and the corresponding references for those values taken from the literature ) are given in Supporting Information ( S1 Text ) . Analyses described below were performed using R ( v3 . 0 , www . r-project . org ) after converting the model into Fortran . All the scripts are provided in Supporting Information ( S1 Code ) .
|
Metabolism is a fundamental biochemical process that enables cells to operate and grow by converting nutrients into ‘building blocks’ and energy . Metabolism happens through the work of enzymes , which are encoded by genes . Thus , genes and their regulation are often thought of controlling metabolism , somewhat at the top of a hierarchical control system . However , an increasing body of evidence indicates that metabolism plays an active role in the control of its own operation via a dense network of metabolite-enzyme interactions . The system-wide role of metabolic regulation is hard to dissect and so far remains largely uncharacterized . To better understand its role , we constructed a detailed kinetic model of the carbon and energy metabolism of the bacterium Escherichia coli , a model organism in Systems and Synthetic biology . Model simulations indicate that kinetic considerations of metabolism alone can explain data from hundreds of experiments , without needing to invoke regulation of gene expression . In particular , metabolic regulation is sufficient to coordinate carbon utilization , redox and energy production , and growth , while maintaining local flexibility at individual metabolic steps . These findings indicate that the self-regulating capacities of E . coli metabolism are far more significant than previously expected , and improve our understanding on how cells work .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"cell",
"physiology",
"carbohydrate",
"metabolism",
"chemical",
"compounds",
"enzymes",
"metabolic",
"networks",
"enzymology",
"carbohydrates",
"glucose",
"metabolism",
"organic",
"compounds",
"glucose",
"cell",
"metabolism",
"network",
"analysis",
"enzyme",
"metabolism",
"enzyme",
"chemistry",
"computer",
"and",
"information",
"sciences",
"proteins",
"enzyme",
"regulation",
"gene",
"expression",
"chemistry",
"biochemistry",
"organic",
"chemistry",
"cell",
"biology",
"genetics",
"monosaccharides",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"metabolism"
] |
2017
|
Metabolic regulation is sufficient for global and robust coordination of glucose uptake, catabolism, energy production and growth in Escherichia coli
|
Mutations are a critical driver of cancer initiation . While extensive studies have focused on exposure-induced mutations , few studies have explored the importance of tissue physiology as a modulator of mutation susceptibility in vivo . Of particular interest is inflammation , a known cancer risk factor relevant to chronic inflammatory diseases and pathogen-induced inflammation . Here , we used the fluorescent yellow direct repeat ( FYDR ) mice that harbor a reporter to detect misalignments during homologous recombination ( HR ) , an important class of mutations . FYDR mice were exposed to cerulein , a potent inducer of pancreatic inflammation . We show that inflammation induces DSBs ( γH2AX foci ) and that several days later there is an increase in cell proliferation . While isolated bouts of inflammation did not induce HR , overlap between inflammation-induced DNA damage and inflammation-induced cell proliferation induced HR significantly . To study exogenously-induced DNA damage , animals were exposed to methylnitrosourea , a model alkylating agent that creates DNA lesions relevant to both environmental exposures and cancer chemotherapy . We found that exposure to alkylation damage induces HR , and importantly , that inflammation-induced cell proliferation and alkylation induce HR in a synergistic fashion . Taken together , these results show that , during an acute bout of inflammation , there is a kinetic barrier separating DNA damage from cell proliferation that protects against mutations , and that inflammation-induced cell proliferation greatly potentiates exposure-induced mutations . These studies demonstrate a fundamental mechanism by which inflammation can act synergistically with DNA damage to induce mutations that drive cancer and cancer recurrence .
Effective strategies for preventing and treating cancer depend not only upon understanding genetic and exposure-induced factors , but also physiological factors that drive disease . DNA damage , caused by endogenous metabolites and exogenous agents , promotes mutations , a key driver of phenotypic changes that potentiate metastasis and enable recurrence after treatment [1] . While significant progress has been made in terms of understanding how genes and exposures modulate the risk of mutations , relatively little is known about the potential role of tissue physiology in modulating the risk of mutations in vivo . Of particular interest is the inflammatory state , a critical cancer risk factor that is associated with sweeping changes in tissue architecture due to immune cell infiltration and associated changes in the levels of cytokines and reactive oxygen and nitrogen species ( RONS ) [2–4] . Inflammation is a well-established tumor promoter that contributes to cancer growth , angiogenesis , and resistance to apoptosis [2 , 5] . In addition to the role of inflammation in cancer progression , it is increasingly recognized that inflammation-induced DNA damage may also drive mutations that contribute to both initiation and progression [3 , 6] . With recent advances that enable analysis of key factors that impact the risk of mutation [7] , here , we set out to determine how interactions between DNA damage and inflammation-induced physiological changes impact the risk of mutations in vivo . It has long been thought that it is the convergence of conditions that induce DNA damage and cell division simultaneously that is a key driver of inflammation-induced mutations [8–11] . Nevertheless , studies that directly query the combined effect of RONS-induced DNA damage and cell division are lacking , both in vitro and in vivo . Importantly , the same proposed mechanism for synergy between cell division and endogenous RONS applies to exogenous DNA damaging agents . In the clinic , virtually all cancer patients are exposed to high levels of DNA damage when treated with radiation and/or chemotherapy , for which DNA damage is often critical to the mode of action . It is well established that an increase in the mutation rate contributes to cancer promotion and drug resistance [12–15] . Therefore , understanding physiological factors that modulate susceptibility to therapy-induced mutations could open doors to strategies to reduce disease recurrence . Pancreatic inflammation is a key risk factor for pancreatic cancer [11 , 16] , one of the most deadly cancers; most patients who initially respond to radio-chemotherapy suffer relapse , such that only ∼5% of patients survive more than 5 years after diagnosis [17] . Inflammation-induced DNA damage potentially plays an important role in driving mutations that enable pancreatic cancer initiation and recurrence . During inflammation there are high levels of RONS , which can induce cytotoxic and mutagenic DNA lesions , including abasic sites , oxidized bases ( e . g . , 8oxoG ) , deaminated bases ( e . g . , uracil and hypoxanthine ) and ethenoadenine ( eA ) [18 , 19] . In addition to base damage , RONS also induce DNA double strand breaks ( DSBs ) . DSBs are among the most toxic of DNA lesions and they can also be potently mutagenic due to the potential loss of vast stretches of chromosomes if not accurately repaired [1 , 20] . Homologous recombination ( HR ) plays a critical role in preventing DSB-induced cytotoxicity by repairing DSBs during S/G2 [21] . To initiate repair , the DNA is resected by MRE11 and EXO1 to generate 3’ single-stranded overhangs [22–25] . BRCA2 then loads RAD51 onto the single-stranded DNA to form a nucleoprotein filament capable of homology searching and strand invasion [26–30] . The resulting D-loop enables the copying of sequence information that can then be processed by downstream proteins to complete the repair process [21] . While HR is effective for repair of two-ended DSBs , its most important role is in the repair of one-ended DSBs that arise when replication forks break down . Unlike two-ended DSBs , which can be repaired by alternative mechanisms , one-ended DSBs require HR for accurate sequence realignment and reinsertion of the broken DNA end . Inflammation induces single strand breaks and replication-blocking lesions , both of which promote replication fork breakdown . Furthermore , mutations in BRCA2 are genetic risk factor for pancreatic cancer [31] , indicating that HR is indeed active in the pancreas [32] . Thus , RONS are predicted to create DSBs during pancreatitis , and HR can potentially repair inflammation-induced DSBs in the pancreas . Ironically , while HR prevents cytotoxicity and is mostly accurate , HR carries a risk of sequence changes . Misalignments during HR promote large scale sequence rearrangements , such as deletions , duplications and translocations [33–35] , and these HR-driven events have been observed in cancers [36 , 37] . Furthermore , HR between homologous chromosomes can also lead to loss of heterozygosity ( LOH ) , a major mechanism for the inactivation of tumor suppressor genes . Indeed , studies with cultured cells have demonstrated that HR is the underlying cause of 30 to 70% of LOH events [38–40] , and the importance of HR-driven LOH has also been demonstrated in tumors [41 , 42] . Finally , it has recently been shown that HR also promotes point mutations in mammalian cells , due to misincorporation during repair synthesis [43–48] . Taken together , it is now clear that virtually all cancers harbor one or more HR-driven sequence changes that promote initiation and progression . Given the importance of HR , we created a mouse model that enables the detection of HR in vivo ( see ref . 7 ) . The fluorescent yellow direct repeat ( FYDR ) mice harbor an integrated direct repeat comprised of two non-functional EYFP expression cassettes , wherein transfer of sequence information by HR from one cassette to the other can reconstitute full length sequence and give rise to fluorescence ( Fig . 1A–D ) [49] . The FYDR recombination substrate is designed to detect the major classes of HR events , including gene conversion ( wherein sequence information is transferred from one duplex to the other ) , sister chromatid exchange ( e . g . , gene conversion with crossover ) and replication fork repair ( S1 Fig . ) [50] . Importantly , FYDR fluorescence after replication fork repair indicates misalignment and transfer of sequence information during HR , and in some cases the gain of one repeat unit in the FYDR substrate ( Fig . 1A ) . Given that all cells that are positive for fluorescence result from sequence misalignment and harbor a change in sequence information , the FYDR readout is indicative of mutation events . The FYDR mouse model thus affords key advances in studies of mutagenesis , since it became possible for the first time to visualize mutant cells that arise within intact tissues of adult animals [7] . Here , we have integrated approaches for visualization and quantification of DNA damage , cell proliferation , and mutation within intact tissues in order to learn about their interrelationships in the context of inflammation . We found that following controlled induction of acute inflammation , the timing for inflammation-induced DSBs is separate from the timing for cell proliferation , creating a protective kinetic barrier against potential synergy between DNA damage and cell division . Breaking this barrier by creating overlap between peak cell proliferation and the acute phase of inflammation causes a synergistic increase in HR-driven mutations . Furthermore , under conditions of inflammation-induced cell proliferation , there is a dramatic increase in susceptibility to mutations induced by exposure to an exogenous DNA damaging agent of a class that is present in environmental contaminants and also commonly used in the clinic . This work reveals the critical role that tissue physiology plays in mutation susceptibility and opens doors to new avenues of cancer prevention and treatment .
In the FYDR mice , HR-induced misalignments between two copies of an expression cassette for EYFP are detectable as fluorescent foci within intact pancreatic tissue ( Fig . 1A , B ) . In some cases , foci are comprised of more than one fluorescent recombinant cell , indicative of a recombination event in a single cell that has subsequently undergone clonal expansion ( Fig . 1C ) [51] . Analysis of tissue histology shows that in the pancreas , acinar cells undergo HR ( Fig . 1D ) , and previous studies show that acinar cells comprise virtually all of the recombinant cells in the FYDR pancreas [51] . The FYDR mice enable studies of exposure-induced HR in the pancreas of adult animals . Of particular interest are alkylating agents , an important class of DNA damaging agents that are present in food and in our environment , some of which have been shown to cause cancer [52–54] . Ironically , alkylating agents are used to treat cancer when given at high doses [55] . Temozolomide , a methylating agent that is used in cancer chemotherapy , kills tumor cells by creating DNA lesions that either directly or indirectly inhibit DNA replication , causing cytotoxicity [55] . Cells that do not die from exposure to temozolomide potentially run the risk of harboring chemotherapy-induced mutations , including HR events . To determine if alkylation damage induces HR in the pancreas , FYDR mice were exposed to the model methylating agent MNU , which creates the same types of base lesions as temozolomide . Results show that MNU causes a significant increase in the frequency of fluorescent foci ( Fig . 1E ) , indicating that the FYDR mouse model is effective for studies of DNA damage-induced HR . In order to study the interactions between DNA damage and inflammation , we exploited cerulein , a cholecystokinin analog that is well established as an inducer of pancreatic inflammation [56 , 57] . Animals exposed to cerulein by 6 hourly intraperitoneal injections showed pancreatic edema and infiltration by inflammatory cells , chiefly neutrophils ( Fig . 2A , B ) . The extent of features of pancreatitis was found to be statistically significantly increased when quantified by a trained pathologist ( Fig . 2C ) . In studies of long term exposure to cerulein , we observed severe tissue atrophy and metaplasia in wild type mice ( Fig . 2D , E ) , and precancerous lesions in K-Ras mice ( S2 Fig . ) , indicating that cerulein exposure serves as a relevant model for pancreatitis-induced cancer . During inflammation , an increase in the levels of macrophages and neutrophils leads to increased levels of RONS [18] . RONS in turn induce base lesions including eA , 8oxoG and Hx , which have been observed at sites of inflammation [18 , 19] . Many RONS-induced DNA lesions have the potential to cause recombinogenic DSBs through chemical cleavage , by enzymatic processing , or as a result of replication fork breakdown [58–60] . To learn if pancreatic inflammation induces DSBs in vivo , we analyzed the frequency of DSB repair foci by quantifying cells with five or more γH2AX foci ( H2AX becomes phosphorylated to form γH2AX in the vicinity of DSBs ) [61] . Immunohistochemical ( IHC ) analysis of pancreatic tissue reveals a clear induction of DSBs after exposure to cerulein ( Fig . 3A , B ) . As HR has been shown to be induced by DSBs in vitro [62 , 63] , we next asked if DSBs associated with acute inflammation induce HR in vivo . To increase the sensitivity of our approach , animals were exposed to three bouts of acute pancreatitis . Analysis of the frequency of HR events in control animals shows that there is variation in the frequencies of foci/cm2 , ranging from ∼15 to ∼100 ( Fig . 3C ) , consistent with previous studies [7 , 64 , 65] . ( It is noteworthy that variation in mutation frequency among normal animals has similarly been shown in several other mouse models for mutation detection [66–69] ) . Unexpectedly , in animals that were subjected to three bouts of inflammation , we did not detect any increase in the frequency of recombination events ( indicated by fluorescent foci; Fig . 3C ) . Analysis of the frequency of fluorescent recombinant cells similarly did not reveal any increase in HR in the animals exposed to three bouts of inflammation ( Fig . 3D ) . HR is active during S/G2 , whereas most cells in healthy pancreatic tissue are non-dividing cells in G0/G1 [70] , raising the possibility that HR was not active in RONS-exposed cells during the three bouts of inflammation . To learn about the extent of cell division during the course of inflammation , we quantified dividing cells when tissue is healthy ( Fig . 4A ) , subject to acute inflammation ( Fig . 4B ) or recovering ( Fig . 4C; five days after cerulein exposure , when features of inflammation have cleared ) . Cell proliferation during the course of the inflammatory response was evaluated by staining for Ki-67 , a marker of cell proliferation [71] . Results show that there are very few Ki-67 positive cells in control and acutely inflamed tissue ( Fig . 4D , E ) . In contrast , the frequency of Ki-67 positive cells is significantly induced during tissue recovery ( Fig . 4F ) and when quantified using image analysis software ( see Materials and Methods ) ( Fig . 4G ) . As an alternative approach , animals were treated with BrdU , a thymidine analog that becomes integrated into the DNA of dividing cells and can be detected using immunohistochemistry . Pancreatic tissue was disaggregated , and the frequency of BrdU positive cells was analyzed by flow cytometry . Consistent with the Ki-67 analysis , results show a clear increase in the frequency of dividing cells several days after acute inflammation ( Fig . 4H ) . Thus , with both methods , we found that acute phase inflammation is separate from a subsequent proliferative phase . As HR is active primarily during S/G2 , we hypothesized that the lack of HR induction following three independent bouts of inflammation might be due to the kinetic separation between acute inflammation-induced DSBs and recovery-induced cell proliferation . We therefore asked if inflammation might induce HR if the timing were adjusted to create overlap between inflammation-induced DSBs and cell proliferation . For ‘protocol 1’ described above , animals were exposed to three independent bouts of inflammation , each two weeks apart ( Fig . 5A ) . Here , for ‘protocol 2’ , animals were also exposed to three bouts of inflammation , however bouts of inflammation were 4–5 days apart ( Fig . 5B ) . For ‘protocol 1’ , we observed that exposure to cerulein induces acute inflammation , as can be seen by the edema and infiltration under inflamed conditions ( compare Fig . 5C and 5D ) . At the time of acute inflammation , the frequency of dividing cells is unchanged compared to untreated animals ( Fig . 5F , G ) . However , cells with high numbers of γH2AX foci are apparent ( Fig . 5J ) , which is consistent with DNA damage formed by RONS that are associated with the acute phase of inflammation . We also observed acute inflammation using ‘protocol 2’ ( Fig . 5E ) . Unlike protocol 1 , we also observed concomitant induction of cell division , consistent with the proliferative phase of the first bout of inflammation ( Fig . 5H ) . Cells with high frequencies of γH2AX foci are evident ( Fig . 5K ) . To learn more about the impact of overlap between bouts of inflammation , the extent of inflammation was assessed by a trained pathologist , the extent of cell proliferation was quantified by automated image analysis , and the frequency of γH2AX positive cells was measured by counting cells with >5 γH2AX foci . Results show that the severity of the acute phase of inflammation is similar regardless of whether bouts of inflammation occur independently or in an overlapping fashion ( Fig . 6A , B ) . In contrast , cell proliferation is dramatically increased under conditions where the response to the first bout of inflammation overlaps with initiation of the second bout of inflammation ( Fig . 6C , D ) . The frequency of DSBs is increased in both independent and overlapping bouts of inflammation , and the increase is greater under conditions of overlap between the acute phase of inflammation and the proliferative phase ( compare Fig . 6E and Fig . 6F ) . Similar results were observed for the third bout of inflammation ( S3–S4 Fig . ) , although the frequencies of γH2AX were reduced during the third bout of inflammation relative to the second bout under conditions of overlap . It is unclear why the third bout of inflammation is apparently less damaging , however one possibility is that HR proficiency increased during the course of the exposure protocol , leading to more rapid clearance of DSBs . It is noteworthy that clearance of potentially toxic DSBs is advantageous to cell survival , but carries the risk of mutations due to HR misalignments . To learn about the impact of inflammatory response kinetics on susceptibility to HR , recombination events were quantified within intact pancreatic tissue , and the frequency of recombinant cells was evaluated in disaggregated pancreatic tissue by flow cytometry . Under conditions of overlapping bouts of inflammation ( protocol 2 ) , there is a significant increase in the frequency of recombination events , which is both visually apparent ( Fig . 7A ) and quantitatively significant ( Fig . 7B ) . In addition , there is a significant increase in the frequency of fluorescent recombinant cells under conditions of overlap ( Fig . 7C ) , but not when animals are exposed to three independent bouts of inflammation ( Fig . 3D ) . The observation that overlapping bouts of inflammation induce HR is consistent with a model wherein inflammation-induced cell proliferation sensitizes tissue to HR induced by endogenously-produced DNA damage . We next asked about the potential for inflammation-induced cell proliferation to cause increased sensitivity to HR induced by an exogenous DNA damaging agent , specifically the model cancer chemotherapeutic , MNU . Experiments were designed with the objective of finding the time when inflammation-induced cell proliferation is high , and then exposing animals to MNU ( Fig . 8A ) . To quantify the extent of inflammation-induced proliferation , pancreatic tissue was analyzed for Ki-67 positive cells . There is a significant increase in cell proliferation at the time of the MNU exposure ( Fig . 8B ) . MNU on its own causes a visually apparent ( Fig . 8C ) and statistically significant increase in the frequency of HR events in healthy animals ( Fig . 8D ) ( note that the data from Fig . 1E have been replotted to facilitate comparisons among cohorts ) . The effect of MNU on HR was dose dependent: at 25 mg/kg , there was a statistically significant increase in the number of fluorescent foci ( Fig . 8D ) , whereas there was not a significant increase in HR after treatment with 7 . 5 mg/kg MNU ( S5 Fig . ) . We also found that a single bout of inflammation does not induce HR ( Fig . 8C , D ) , which is consistent with results shown above ( Fig . 3 ) . Importantly , when animals were exposed to MNU at a time when inflammation-induced proliferation is high , there was a dramatic increase in the frequency of HR ( Fig . 8C , D ) , revealing that physiological changes associated with inflammation and exposure to an exogenous DNA damaging agent act synergistically to induce HR . These results call attention to the importance of inflammation as a modulator of DNA damage-induced sequence rearrangements induced by exposure to an alkylating agent that serves as a model for environmental and clinical DNA damaging agents .
Pancreatic cancer is one of the most deadly cancers , yet relatively few studies have explored factors that govern susceptibility to mutations that initiate pancreatic cancer . Furthermore , while radiation and chemotherapy can be effective initially , recurrence is virtually inevitable [72] , and mutations are a key driver of recurrence since they enable evolution into drug resistant and more aggressive phenotypes [12–15] . Thus , there is a need for a deeper understanding of the mechanisms of DNA damage-induced mutations in the pancreas . Furthermore , while it is well established that pancreatitis is a key risk factor for pancreatic cancer [11 , 16] , studies had not previously been done to explore how physiological changes associated with inflammation modulate the risk of mutations in vivo . Here , we show that pancreatic inflammation leads to DNA double strand breaks , and that pancreatitis is associated with hyperproliferation . By creating conditions where there is overlap between bouts of inflammation , we show that DSBs and hyperproliferation act synergistically to induce sequence rearrangements in vivo ( Fig . 9 ) , which both demonstrates a correlation between DSBs and HR in vivo and provides insights into the underlying mechanisms that make pancreatitis a risk factor for cancer . Furthermore , we show that inflammation-induced proliferation acts synergistically with a DNA alkylating agent to induce sequence rearrangements in vivo , providing new understanding into factors that modulate the risk of sequence changes that promote cancer . For decades , it has been known that inflammation is a risk factor for cancer [11 , 16] , and it has long been postulated that it is the combination of inflammation-induced DNA damage and inflammation-induced cell proliferation that plays a key role in promoting mutagenesis [8–11] . Nevertheless , direct evidence for this model was lacking . Here , we show that , unexpectedly , several bouts of acute inflammation on their own are not sufficient to drive sequence rearrangements , and that separation of the acute phase of inflammation ( associated with RONS and DNA damage ) and the proliferative stage of inflammation provides a barrier to DNA damage-induced sequence rearrangements . Consequently , conditions that lead to chronic inflammation may be more likely to potentiate tumorigenic mutations compared to isolated bouts of inflammation , which is consistent with epidemiological studies [73 , 74] . Here , we observed that approximately half of the animals exposed to overlapping bouts of inflammation have frequencies of recombinant cells that are ∼100–200% higher than the untreated control animals . Given that the mutation rate can be rate limiting in tumor promotion [14] , a doubling of the mutation frequency could potentially double the probability of cancer recurrence . An increased risk of mutations has relevance to many medical conditions that are associated with chronic inflammation [4] . Inflammatory bowel diseases such as ulcerative colitis and Crohn’s disease involve chronic inflammation in the colon , while chronic esophagitis and pancreatitis affect the upper gastrointestinal tract and the pancreas respectively . In addition , chronic infections with bacteria , viruses and parasites can lead to chronic inflammation at multiple sites . Importantly , chronic inflammatory conditions typically last for an extended period of time . Thus , a relatively small increase in susceptibility to mutations in people is anticipated to become very significant given the accumulation over a period of years . RONS create a wide array of DNA lesions that includes dozens of different types of base lesions as well as abasic sites and strand breaks [75–77] . There is a wealth of information about the mutagenicity of RONS-induced DNA damage derived from studies in vitro [78 , 79] . Many elegant studies have revealed the mutagenic potential of specific RONS-induced base lesions using site-specific lesion technology [80] , and many others have described the ability of inflammatory chemicals to induce mutations in RONS-exposed cells in vitro [81 , 82] . Using these and other approaches , we now know quite a lot about the molecular and biochemical mechanisms of RONS-induced mutagenesis . For example , 8oxoG readily mispairs with thymine when bypassed by translesion polymerases [83 , 84] , and that cells prevent TLS-driven mutagenesis by removing 8oxoG [85–87] . Cells also have additional strategies for preventing RONS-induced mutations , including removal of damaged bases from the nucleotide pool ( e . g . , Mth1 ) [88 , 89] , and removing the misincorporated base opposite the lesion post replication ( e . g . , removal of adenine across from 8oxoG by Mutyh ) [89 , 90] . While the literature describing RONS-induced base lesions in vitro is extensive ( we refer the reader to several excellent reviews [78 , 79 , 81 , 82] ) , relatively few studies have addressed RONS-induced mutagenesis in vivo . These studies showed that base excision repair is critical in suppressing RONS-induced mutations in vivo [91–94] , and that inflammation induces mutations in the affected tissues [95–97] . Interestingly , in one such study it was shown that H . pylori infection is associated with mutations [98] , however the frequency of mutations decreased when Ogg1 was knocked out , leaving unclear the mechanism of mutagenesis . In another study , Ogg1 was found to suppress mutations induced by oxidative damage [99] . The most direct evaluation of the relationships among inflammation , DNA damage , mutagenesis , and cancer was done recently in the laboratory of L . Samson . This study showed that a deficiency in the Aag glycosylase is associated with increased inflammation-induced cancer , and that tumors harbor mutations consistent with the predicted mutations that would result from an Aag deficiency [6] . Here , we have extended the in vivo studies of inflammation and mutagenesis to specifically query the inter-relationships among inflammation , cell proliferation , DSBs , and their consequences ( homologous recombination events ) , using tools that had not previously been applied to this problem . It is important to note that this study focuses on a specific class of mutation ( HR-driven sequence rearrangements ) , and that there are other classes of mutations that are not detected by the FYDR assay , such as base damage-induced point mutations ( which often arise during TLS ) , and small insertions/deletions ( which are sometimes associated with NHEJ ) . Nevertheless , inflammation-induced HR events are expected to arise contemporaneously with other classes of mutations . Specifically , both point mutations and HR events arise primarily as a consequence of DNA damage that is present during DNA replication . Thus , HR may serve as an indicator of a more general increase in mutagenesis . Indeed , an association between point mutations and HR events is consistent with observations showing that exposure-induced HR is an excellent predictor of carcinogenicity , which generally arises as the result of multiple classes of mutations [100] . RONS-induced DSBs are rarely caused by direct reaction with the DNA [18 , 19] , but instead are the result of enzymatic processing . Specifically , base excision repair of RONS-induced lesions is associated with gaps that form as repair intermediates [60] . These single strand breaks can become DSBs when repair patches are closely opposed [60 , 101] . Additionally , replication forks that encounter RONS-induced single strand breaks can break down [21] , creating a double strand break . We observed an increase in DSBs under both the conditions of isolated bouts of inflammation , and overlapping bouts of inflammation . Interestingly , under conditions where proliferation from the first bout of inflammation overlaps with acute inflammation from the second bout of inflammation , we observed that DSBs were greatly increased compared to conditions without overlap . This observation is consistent with the possibility that DSBs form in a replication-dependent manner as a result of replication fork breakdown . In the FYDR direct repeat substrate , full-length Eyfp sequence can be reconstituted by several HR mechanisms . For example , if there is a fork breakdown event during DNA replication , misinsertion of the double-strand end can restore full length Eyfp , leading to a gain of one repeat unit ( a rearrangement at the FYDR substrate , Fig . 1A ) . Importantly , the FYDR substrate is similar in size to Alu repeats ( ∼500 bp vs ∼300 bp ) , which make up almost 10% of the human genome and are frequent sites of HR-induced rearrangement formation [102] . HR between Alu repeats can yield deletions , duplications and translocations [102] . Alu-mediated rearrangements have been shown to activate oncogenes in cancer [103] and to inactivate tumor suppressor genes such as p53 [104] . Further , HR-driven rearrangements between Alu repeats have been shown to drive carcinogenesis in inflammation-associated cancers [36 , 105] . Thus , HR events that occur in FYDR mice after replication fork repair are related to genetic changes that are relevant for carcinogenesis in humans . Alkylating agents are abundant in our environment , endogenously produced in our cells , and used at high doses as cancer therapeutics . Understanding factors that modulate alkylation-induced mutations is therefore relevant both to cancer etiology and to cancer recurrence . We show here that inflammation-induced cell proliferation acts synergistically with alkylation damage to induce sequence rearrangements ( Fig . 9 ) . Thus , one potential factor when considering the underlying mechanisms by which chronic inflammation promotes cancer is that the inflammatory response sensitizes tissue to exposure to DNA damaging agents that are in our environment and in our food . Furthermore , as proliferation itself is sufficient to increase susceptibility to DNA damage induced sequence rearrangements [106] , careful consideration should be given to babies in utero and young children for whom high levels of cell proliferation are anticipated to greatly sensitize cells to exposure-induced mutations . Thus , when screening for potentially carcinogenic exposures , it will be important to consider the importance of a person’s physiological state when assessing risk , with regard to both chronic inflammatory conditions and stage in development . Recurrence is the single biggest hurdle in cancer treatment , and mutations are critical in eliciting phenotypic changes that initiate new secondary cancers , promote existing cancer cells , and potentiate drug resistance [1 , 12–15] . It has recently been demonstrated that mutation rate directly impacts the emergence of drug resistance [14] . While in some cases cancer cells are hypermutable [13] , many transformed cells have a normal mutation rate [12] , making exposure-induced mutations highly relevant . Tumors generally exist in a chronic pro-inflammatory environment . Associated increases in proliferation of both tumor and stromal cells are anticipated to increase susceptibility to RONS-induced and chemotherapy-induced HR events that can promote metastasis and recurrence ( Fig . 9 ) . Novel approaches for treating cancer are currently in development , including staged release of drugs from nanoparticles that increase cell killing by chemotherapeutic agents [107] . These approaches could help minimize treatment-induced mutations and thus slow the emergence of drug resistant or more aggressive cancers . The observation that there is synergy between conditions that induce hyperproliferation and conditions that cause DNA damage is relevant to millions of people who suffer from chronic inflammation and are thus at increased risk of mutations that drive cancer . In addition , the observation that inflammation sensitizes tissue to alkylation-induced HR is relevant to other exposures that create DNA lesions that inhibit replication , including constituents of food , cigarette smoke , and environmental carcinogens ( e . g . , aflatoxin , BaP , PhIP ) . Importantly , although the focus of this work is on HR at an integrated reporter , the FYDR model serves as a powerful tool to learn about more general increases in HR throughout the genome , with their accompanied increased risk of LOH , insertions , deletions , and point mutations , all of which drive cancer ( Fig . 9 ) . Through these studies of the dynamic physiological changes associated with inflammation , this work contributes to our fundamental understanding of how inflammation drives genetic changes that cause cancer and calls attention to new avenues to disease prevention and treatment .
All animal experiments were conducted according to the Guide for the Care and Use of Laboratory Animals , and were approved by the MIT Committee on Animal Care . Cerulein , methylnitrosourea ( MNU ) , BrdU , soybean trypsin inhibitor and collagenase were purchased from Sigma-Aldrich . Female C57Bl/6 pun FYDR mice ( [7] , 5 to 7 weeks old ) were used for measuring HR . Inflammation , proliferation and double-strand breaks were measured using female C57Bl/6 ( Taconic ) and C57Bl/6 pun FYDR mice ( 5 to 7 weeks old ) . Metaplastic and preneoplastic lesions were assayed using male wild type or K-Ras mutant mice ( gifts from T . Jacks , MIT ) on the FVB background ( 8 months old at analysis ) . Mice were housed in an AAALAC approved , specific pathogen free facility under a 12h light/dark cycle and were fed a standard rodent diet ( LabDiet RMH 3000 , Purina LabDiet ) and autoclaved water ad libitum . For measuring HR , litters were split between experimental groups . Mice were subjected to 3 episodes of acute pancreatitis on experimental days 0 , 4 and 9 , or on days 0 , 14 and 28 . Each episode was elicited by 6 hourly intraperitoneal injections of cerulein ( dissolved in PBS , 50 μg/kg for each injection ) . Control animals did not receive injections , as serial injections of PBS have no effect on HR ( S6 Fig . ) . To assess inflammation , Ki-67 expression , and double-strand breaks , mice were humanely euthanized 12 hours after the first cerulein injection and pancreata were harvested for histological analysis . To assess regenerative cell proliferation by BrdU labeling , mice were dosed with BrdU ( 75 mg/kg ) five days after the first bout of acute pancreatitis . Four hours after BrdU injection , mice were humanely euthanized and their pancreata were harvested and processed for BrdU detection by flow cytometry . To assess homologous recombination , mice were humanely euthanized 10 to 15 days after the last pancreatitis episode , and pancreata were harvested for the FYDR assay . Chronic pancreatic inflammation was elicited by cerulein injections ( 5 μg dissolved in saline , single intraperitoneal injection , 5 days a week ) for 6 months , as described in [108] . Control mice received saline injections . Mice were 2 months old at the beginning of treatment . At 8 months of age , mice were humanely euthanized and pancreata were harvested for histological analysis . Mice received 6 hourly intraperitoneal injections of cerulein ( dissolved in PBS , 50 μg/kg for each injection ) to induce acute pancreatitis . Control mice received 6 hourly injections of PBS . To assess regenerative proliferation by Ki-67 expression , mice were humanely euthanized five days after acute pancreatitis induction and their pancreata were harvested for histological analysis . To induce exogenous DNA damage during regenerative proliferation , mice were dosed with methylnitrosourea ( 25 mg/kg , dissolved in PBS , pH 4 ) five days after cerulein treatment . ( Note that the timing in this experiment is different from the timing in the repeated inflammation experiment , as MNU generates DNA damage directly and much faster than inflammation induced by cerulein . ) Control mice were dosed with PBS , pH 4 . Mice were humanely euthanized 3 to 4 weeks after methylnitrosourea injection and pancreata were harvested for the FYDR assay . Pancreata were disaggregated by mechanical chopping and collagenase V digestion at 37°C for 40 min , followed by gentle pipetting . Cells were collected by centrifugation and were stained with the APC Cell Proliferation Detection Kit ( BD Pharmingen ) according to the manufacturer’s instructions . Samples were analyzed on a FACSCalibur flow cytometer ( BD Biosciences ) using CellQuest Pro software . On average , 20 000 cells were analyzed per sample . Pancreata were fixed in 10% neutral buffered formalin , embedded in paraffin , and sectioned at 4 μm . After deparaffinization , heat-induced antigen retrieval was performed using a modified citrate buffer ( Dako ) . Ki-67 antibody ( rat anti-mouse Ki-67 , Dako ) was used at a dilution of 1/100 at room temperature for 1 hour . Secondary antibody ( biotinylated rabbit anti-rat Ig , Dako ) was used at a dilution of 1/100 at room temperature for 20 minutes , and detected using streptavidin-conjugated peroxidase and DAB . Sections were counter-stained with hematoxylin . In repeated inflammation experiments , the percentage of Ki-67 positive nuclei was determined in 20 randomly selected images ( ×20 ) using image analysis software ( Visiopharm , Hørsholm , Denmark ) . In the acute inflammation + MNU experiment , the number of Ki-67 positive nuclei was counted in 15 randomly selected image fields ( ×20 ) in a blinded fashion . Sections ( 4 μm ) of formalin-fixed , paraffin-embedded tissue were deparaffinized and antigen-retrieved using modified citrate buffer ( Dako ) . Sections were incubated with primary γH2AX antibody ( Millipore ) at a dilution of 1/100 at 4°C for 3 hours . Secondary antibody ( Alexa Fluor 488 Goat Anti-Mouse IgG , Invitrogen ) was used at a dilution of 1/500 at room temperature for 1 hour . Sections were counter-stained with DAPI before imaging . For each section , images of 20 randomly selected image fields were acquired at a magnification of ×40 using ImagePro Plus software ( Media Cybernetics ) . DAPI-stained nuclei were counted using ImagePro Plus , and nuclei containing more than 5 γH2AX foci were counted manually in a blinded fashion . In situ fluorescent imaging . Pancreata were immersed in ice cold soybean trypsin inhibitor solution ( 0 . 01% in PBS ) immediately after harvesting . Pancreata were pressed between glass coverslips separated by 0 . 5 mm spacers and imaged on a Nikon 80i epifluorescence microscope ( Nikon ) with a CCD camera ( CoolSNAP EZ , Photometrics ) using a ×1 objective at a fixed exposure time ( 2 s ) . EYFP was detected in the FITC channel . Multipoint images captured using an automated stage ( ProScan II , Prior Scientific ) and NIS Elements software ( Nikon ) were stitched automatically or manually in Adobe Photoshop ( Adobe Systems ) . Brightness and contrast were adjusted identically across images , and foci were manually counted in a blinded fashion . Areas of pancreata were determined using ImageJ software ( NIH ) by manually tracing the pancreas outline . Flow cytometry . Following imaging , pancreata were disaggregated into single-cell suspensions as described in [7] , with minor modifications . Briefly , pancreata were minced with scalpel blades , followed by digestion with collagenase V ( 2 mg/ml in Hanks’ Balanced Salt Solution ) for 40 min at 37°C . The resulting suspension was gently triturated to increase mechanical separation and filtered through a 70 μm cell strainer ( BD Falcon ) into an equal volume of media ( DMEM F12 HAM with 20% FBS ) . Cells were collected by centrifugation , resuspended in 350 μl OptiMEM ( Invitrogen ) and filtered through 35 μm filter caps into flow cytometry tubes ( Beckton Dickinson ) . Samples were analyzed on a FACScan cytometer ( Beckton Dickinson ) using CellQuest Pro software ( Beckton Dickinson ) . On average , 1 800 000 cells were analyzed per sample . Pancreata were fixed in 10% buffered formalin , embedded in paraffin , sectioned ( 4 μm ) and stained with hematoxylin and eosin . Pancreata were then examined and scored by a trained veterinary pathologist in a blinded fashion on a scale of 0 to 4 for the following individual features: inflammation , edema , hemorrhage , acinar degeneration/necrosis , acinar loss/atrophy , fat infiltration , fibrosis , acinar to ductal metaplasia ( ADM ) , acinar/ductal hyperplasia , acinar dysplasia/neoplasia and ductal dysplasia/hyperplasia . For the acute studies , only a few relevant subsets were analyzed and scored , whereas for the chronic studies , the full set of criteria was assessed . Inflammation , proliferation and double-strand break indices were compared with Student’s t-test . Numbers of recombinant foci , recombinant cell frequencies , and pathological scores do not follow a normal distribution and were compared with the Mann–Whitney U-test . All statistical analyses were performed in GraphPad Prism , Version 5 . 02 ( GraphPad Software ) . A P value less than 0 . 05 was considered statistically significant .
|
People with chronic inflammatory conditions have a markedly increased risk for cancer . In addition , many cancers have an inflammatory microenvironment that promotes tumor growth . Here , we show that inflammatory infiltration synergizes with tissue regeneration to induce DNA sequence rearrangements in vivo . Chronically inflamed issues that are continuously regenerating are thus at an increased risk for mutagenesis and malignant transformation . Further , rapidly dividing tumor cells in an inflammatory microenvironment can also acquire mutations , which have been shown to contribute to drug resistance and disease recurrence . Finally , inflammation-induced tissue regeneration sensitizes tissues to DNA damaging environmental exposures and chemotherapeutics . The work described here thus increases our understanding of how inflammation leads to genetic changes that drive cancer formation and recurrence .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Inflammation-Induced Cell Proliferation Potentiates DNA Damage-Induced Mutations In Vivo
|
Advances in proteomic technologies continue to substantially accelerate capability for generating experimental data on protein levels , states , and activities in biological samples . For example , studies on receptor tyrosine kinase signaling networks can now capture the phosphorylation state of hundreds to thousands of proteins across multiple conditions . However , little is known about the function of many of these protein modifications , or the enzymes responsible for modifying them . To address this challenge , we have developed an approach that enhances the power of clustering techniques to infer functional and regulatory meaning of protein states in cell signaling networks . We have created a new computational framework for applying clustering to biological data in order to overcome the typical dependence on specific a priori assumptions and expert knowledge concerning the technical aspects of clustering . Multiple clustering analysis methodology ( ‘MCAM’ ) employs an array of diverse data transformations , distance metrics , set sizes , and clustering algorithms , in a combinatorial fashion , to create a suite of clustering sets . These sets are then evaluated based on their ability to produce biological insights through statistical enrichment of metadata relating to knowledge concerning protein functions , kinase substrates , and sequence motifs . We applied MCAM to a set of dynamic phosphorylation measurements of the ERRB network to explore the relationships between algorithmic parameters and the biological meaning that could be inferred and report on interesting biological predictions . Further , we applied MCAM to multiple phosphoproteomic datasets for the ERBB network , which allowed us to compare independent and incomplete overlapping measurements of phosphorylation sites in the network . We report specific and global differences of the ERBB network stimulated with different ligands and with changes in HER2 expression . Overall , we offer MCAM as a broadly-applicable approach for analysis of proteomic data which may help increase the current understanding of molecular networks in a variety of biological problems .
Large and complex high-throughput proteomic experimental studies are becoming more accessible through the use of powerful , swiftly developing platforms such as mass spectrometry ( MS ) , flow cytometry ( FC ) , and various kinds of protein microarrays ( PMA ) [1]–[3] . As one particular example of increasing attention , there has been an explosion in large-scale datasets for receptor tyrosine kinase ( RTK ) network signaling by the combination of protein post-translational modification enrichment followed by quantitative MS methods [4] . In receptor tyrosine kinase ( RTK ) networks , such as those activated by the ERBB family of receptors , phosphorylation plays a central role in the translation of extracellular cues into phenotypic changes , such as differentiation , proliferation , and migration [5] . Phosphorylation on proteins in the RTK network induce a variety of signaling events including protein-protein interactions , enzymatic activation and inactivation , and cellular localization changes , such as translocation to the nucleus or recruitment to the plasma membrane [6]–[8] . Understanding RTK networks , and the phosphorylation that occurs within them , will be essential for understanding RTK signaling in normal and dysregulated conditions . Mass spectrometry measurement of phosphorylation events in cellular signaling networks is greatly increasing our understanding of the specific modifications occurring in the cell as well as their relative changes in response to network perturbations , such as ligand stimulation or kinase inhibition . However , the pace of identification of phosphorylation sites in cellular networks has outstripped our ability to understand the function and regulation of the measured phosphorylation sites as evidenced by the sharp increase in phosphorylation database repository sizes , such as Phospho . ELM [9] . Unsupervised computational learning methods , applied to quantitative phosphoproteomic data , provides one method by which to infer function and regulation of phosphorylation within cell signaling networks . Previous studies have shown success in application of unsupervised learning to phosphoproteomic data [10]–[12] , however each has presented the results of a single clustering solution , and to date , no extensive study of the relationship between unsupervised learning and phosphoproteomic data has been produced . Phosphoproteomic data represents a new challenge in unsupervised learning and currently , no gold standard exists as a method for judging the success of a clustering solution of phosphoproteomic data . Unsupervised learning approaches , such as various types of clustering algorithms , have been heavily utilized to productive effect in the biological community in application to other kinds of high-throughput biomolecular measurements including gene sequence [13] , gene expression [14] , and metabolomics [15] . Unsupervised learning algorithms , often referred to as clustering , seek to group components of a multidimensional dataset into clusters where intra-cluster differences are minimized and inter-cluster differences are maximized . Successful application of clustering has demonstrated its usefulness in reducing dataset dimensionality and providing biological hypotheses through the use of inference . Application of unsupervised learning to biological datasets is extensive and includes a seemingly endless option of algorithms , such as Kmeans [14] , hierarchical clustering [16] , self organizing maps [17] , and affinity propagation [18] . Since most clustering algorithms base clustering solutions on the similarity , or dissimilarity , between two vectors of measurements based on a distance metric , the metric used will have an impact on the final solution as will any alteration applied to the vectors by data transformations . Several studies have examined the effect different distance metrics [19]–[21] and data transformations [21] , [22] have on clustering solutions of experimentally-derived data . In addition to choosing an appropriate algorithm , distance metric , and data transformation , scientists are also faced with having to determine a suitable number of clusters ( K ) in which to partition their dataset , since few algorithms incorporate concurrent optimization of K . A variety of methods for determining the natural cluster structure of a dataset have been proposed , see [23] for an example of their comparative performances using a set of microarray experiments . Taken as a whole , the historical application of unsupervised learning to microarray data , as well as other large biological datasets , indicates a vast landscape of decisions required to apply unsupervised learning effectively . We view this vastness as offering an opportunity for enhanced capability to gain biochemical and biological insights , via an approach that takes advantage of the diversity by seeking consistencies and contrasts . In this work we create a framework , Multiple Clustering Analysis Methodology ( MCAM ) , in which one can apply a vast array of algorithms , distance metrics , data transformations , and cluster set sizes , in a combinatorial fashion , i . e . by the exhaustive combination of all chosen parameters , to a biological dataset and then subsequently compile and evaluate the outcome of all solutions . In this framework , enrichment for biological terms within clustering partitions , relative to the dataset , gives us a metric for evaluating the success of any particular clustering implementation . The use of this framework allows scientists to apply unsupervised learning in a way that requires no a priori knowledge or assumptions regarding the most useful clustering algorithm , distance metric , data transformation , or set size . We apply MCAM to diverse phosphoproteomic time-course datasets arising from studies of the ERBB network , where interesting biological enrichment includes Gene Ontology terms , protein domains , kinase and phosphopeptide binding domain predictions , and amino acid sequence motifs . We explore results from these analyses for relationships between biological metrics and the parameters of clustering and develop methods for combining results of clustering to produce robust biological hypotheses and inference . Using this information we explore the difference between two independent measurements of ERBB signaling and find interesting modules of signaling that may be responsible for the migratory potential of HER2 amplification in cancer . The implementation of MCAM was done through a mix of Matlab and Perl scripts . The Perl programs for analyzing enrichment have been incorporated directly into the PTMScout interface ( http://ptmscout . mit . edu ) [24] and the Matlab tools to generate MCAM cluster sets and analyze resulting enrichment are available for download from PTMScout as well . PTMScout is an open-access web and data resource that contains experimental information concerning protein phosphorylation and lysine acetylation , tools for analyzing proteomic datasets , and the ability for users to directly load new experiments for analysis and data-sharing purposes .
The MCAM framework is depicted in Figure 1 . Our specific interest was to cluster dynamic phosphorylation measurements of the ERBB network in order to find phosphorylation events with similar temporal dynamics leading to possible hypotheses regarding shared regulation or shared functionality . In the MCAM framework , a biological dataset , here dynamic phosphorylation measurements , is first subjected to a set of data transformations , such as log transformation or mean-centering; this transformed data is then clustered via the exhaustive combination of different algorithms , distance metrics , and solution set sizes ( K ) . In our implementations , this produces on the order of 1 , 500 to 2 , 500 clustering sets , which we refer to as an MCA . Each set in an MCA represents the complete clustering solution of a single combination of clustering parameters . Next , we calculate enrichment in all biological metric terms , as well as markers for time dynamics , Table 1 , within each cluster , compared to the full dataset , using the hypergeometric function and the False Discovery Rate , FDR [25] , procedure to correct for multiple hypothesis testing for every metric across a set , Figure 1B . Enrichment in biological metrics will give us a method by which we can judge the fitness of a clustering solution for known biological information , where those categories include information about structure and function of a protein as well as information about the regulation and function of the particular phosphorylation sites , Table 1 . Finally , we test , for every parameter of clustering , the impact on the biological information if all sets with that parameter are removed . Parameters are removed from the final set when the removal of that parameter results in a significant improvement of overall biological enrichment and does not significantly decrease the enrichment of any single category of biological information . This pruning , or parameter refinement , allows us to remove clustering solutions which , according to the current biological metrics , are ineffective at producing solutions of biological import . We call this post-pruned set of solutions , for the final Multiple Clustering Analysis . It is important to consider that each clustering solution in an MCA is derived by the quantitative measurements alone , in an unsupervised manner . However , biological enrichment is used to shape the final MCA representing a feature selection step in the MCAM method . is typically on the order of 500 clustering sets . With this final MCA we can begin to compile the biological information that is contained , explore how clustering parameters compare to each other using mutual information , how clustering parameters relate to particular biological metrics , and consider biological hypotheses generated throughout all of these steps . In this work we will focus on studies of ERBB network dynamics . In a first example , we evaluate a four time point measurement of the 184A1 human mammary epithelial cell ( HMEC ) line stimulated with a saturating concentration of EGF , where measurements were taken before stimulation ( 0 min ) and then subsequently at 5 , 10 and 30 minutes following EGF addition [26] . Enrichment and fractionation steps focused on capturing tyrosine phosphorylation signaling events in the ERBB network . This dataset represents extensive measurement of the phosphotyrosine ERBB signaling network , with 77 unique phosphopeptide measurements on 68 proteins . Throughout this work we will refer to this dataset as EGF4 for brevity . This dataset , a 77×4 matrix , represents the relative quantitative measurements of 77 phosphopeptides in time following EGF stimulation , where the vectors used in clustering are the dynamic measurements of a single phosphopeptide . The full dataset is plotted in principal component space in Supplementary Figure S1 . The clustering parameters originally applied , and those removed during pruning , are described in Table 2 . The original MCA included 1 , 320 sets and after pruning , included 331 total sets . Pruning indicates that for this dataset and set of parameters , the use of Hierarchical clustering , set sizes of two and four , the differential transformation , and the use of correlation and cosine as a distance metrics are uninformative in producing enrichment in the categories tested . The 331 sets remaining after parameter refinement produces a wide diversity of biological enrichment across all categories of information . In order to ensure that the basis of our analysis for clustering fitness is a product of actual biological power , versus the production of a large degree of false positives due to Type I error , we performed ten random controls , Figure 2 and Supplementary Figure S2 , using two different randomization methods . In the first method , the data matrix is randomly reshuffled and subjected to clustering using the same parameters as in , followed by enrichment analysis . This method should control for the mechanics of clustering random data . The second method randomized the biological labels assigned to the vector dynamics by randomly reshuffling the assignment of phosphopeptides to the measured phosphorylation dynamics . This randomization process should control for the potential limitations in the annotations used for enrichment . Both methods gave similar results , and if not explicitly stated otherwise , the results described in this paper refer to randomization using the first method described . Figures 2A and B depict the extent of biological enrichment found in one representative metric , PFAM domains , for both real data and a representative random control . Figures 2A and B represent the total number of enriched PFAM labels found per cluster set , in order of descending degree of enrichment per set . The ‘top’ and ‘bottom’ quartiles , according to total enrichment per set , are indicated since they will be used extensively throughout this work . As expected , we do see some enrichment in the random control , which represents a small number of false positives , however the contrast between the results of clustering real biological data and random data is stark . These observations indicate that clustering of temporal phosphorylation measurements is capable of producing meaningful biological enrichment and may therefore be useful for inferring function and regulation of poorly understood phosphorylation events in the ERBB network . To ensure that Type I error , or the false positive rate , is controlled , at least empirically , at the target rate of 0 . 05 or better , we plotted the rate of null hypothesis rejection across all metrics for real data and for the 10 random controls , Figure 2C . Figure 2C shows that for most categories the rate of null hypothesis rejection is higher than the random controls , and for all controls Type I error is empirically controlled at 0 . 05 or better , with only one outlier in the case of GO Molecular Function terms . We also wanted to explore the total number of statistically significant labels produced in real versus random MCA sets , shown in Figure 2D and Supplementary Figure S2 for randomization using the second method described above . Bonferroni correction was also used and the results were similar regarding empirical control of the false positive rate , but demonstrated a large loss in statistical power , Supplementary Figure S3 . Given these results , FDR correction is used in all further results presented . Figure 2 shows some categories , like motifs , which have roughly the same rate of null hypothesis rejection as random controls , have a much larger number of total hypothesis tests , resulting in a large degree of enrichment in the case of real biological data . However , some categories perform no better than random controls based on both rate of null hypothesis rejection and total enrichment . These categories include Scansite Kinase predictions , PhosphoELM kinase annotations , and phosphorylation within known PFAM domains ( pfam_site ) . We noticed that despite the fact that overall enrichment for Scansite Kinase terms was poor in , clusters derived using the Fast Fourier Transform , FFT , were correlated with production of Scansite Kinase enrichment in a significant manner . We looked at random controls and real enrichment for Scansite Kinase terms in just FFT derived clusters and found there was an appreciable improvement in enrichment for this relative to random controls Supplementary Figure S4 , indicating that for specific enrichment , subsets of could be considered . This presents a method in which to target clustering results for producing meaningful information in a particular biological category . For example , here it appears that the FFT subset , which performs better than random controls for producing Scansite Kinase predictions , may provide clusters of phosphorylation sites with the power of inferring shared kinase regulation . Taken together , these results demonstrate that the use of any single algorithm , distance metric , and transformation would highlight only a small fraction of the possibly interesting relationships between the data . We originally hypothesized that the choice of optimal unsupervised learning parameters would be dependent on the biological information desired in the resulting clustering solution . Therefore , we explored the relationship between clustering parameters and the enriched biological information . We rank ordered the sets in an MCA in nine ways , according to the degree of total enrichment the sets contained within each of the nine biological metrics , depicted in Figure 2A for PFAM domain ranking . We found that there was no single set that performed in the top 25% of all nine metric rankings . When we took the union of the top 25% in the nine orders , we found that 273 sets of the 331 sets in the MCA , or 83% of all sets , were required to capture the top quartile of all biological information . These results indicate that , indeed , the choice of ‘optimal’ parameters will be based on the desired type of categorical relationships , since almost all sets are required to capture the full array of possible information . To understand how parameters directly relate to specific types of biological information , we tested for the overrepresentation of clustering parameters in the top and bottom quartiles of all nine biological metrics , and four dynamic metrics , Table 3 . We found that the parameter performance is based on the particular metric tested . For example , the cityblock distance metric consistently performs poorly for producing clustering sets with clusters that have enrichment in GO Molecular Function terms , but the same distance metric consistently performs well in producing sets with clusters containing PFAM enrichment . Interestingly , Kmeans clustering is in the bottom quartile of all biological metrics . However , since Kmeans was not cut in the reduction stage of creating , it indicates that is still useful at producing biological enrichment , unlike those parameters shown in Table 2 that were pruned based on their inability to produce significant biological meaning . The lack of enrichment for a parameter in a particular category indicates that the top and bottom quartiles have diversity of parameters in this category . For example , there are transforms , log10 and normMax , that consistently produce clusters with motif labels , but there is no particular distance metric or algorithm that outperforms others in this capacity . This observations indicate that there is indeed an important relationship between the parameters chosen in clustering and the resulting biological information produced . We discerned additional results in Table 3 to be worth exploring further . For example , it seems that the FFT transformation performs well in producing multiple protein-level metric types . To elucidate this , we compared the set rankings in all 13 metrics ( biological and dynamic ) in a pair-wise fashion . Using a bootstrapping method , we compared the overlap in the top and bottom quartile rankings with that expected by random and highlighted when there was either a significant increase or a significant decrease between any two metric rankings compared to the expected overlap , Figure 3A . The most striking results from this analysis are that there tends to be relatively good agreement in the top and bottom quartiles of sets between protein-level information and site-level information , with the exception of motifs . However , there is very little overlap between different site-level categories in the top quartile , except for Scansite Binding predictions and motifs . There are roughly twice as many positively enriched overlaps in the bottom quartile than there are the top quartile indicating there tends to be more agreement in the worst performing sets across different categories than in the best performing sets . Enrichment for dynamic terms is meant to act as a marker for those features in the dynamics that may have led to a particular clustering outcome . There is a good degree of agreement between dynamic enrichment , but very little agreement with protein-level information and opposing agreement with site-level information . In order to compare how the phosphopeptide cluster mappings between different clustering solutions compare , we calculated the mutual information between sets in a pairwise fashion . The pairwise MI calculations have been hierarchically clustered in Figure 3B , high values of MI indicate a closer agreement in two clustering solutions than low values of MI . The clustered heat map of MI terms indicates there are pockets of sets with high agreement . We bisected the MI cluster tree at various levels and searched for indications that a set of parameters drives similarity between set architectures . At the coarsest level , and depicted in Figure 3C , similarity appears to be driven by the data transformation . At the next level , it appears the algorithm defines similarity and finally the distance metric . There was no clear evidence that cluster set size , K , was important in determining similarity , but this is a poor method for determining the exact effect of K , since the maximum possible MI between any two sets , or self-MI , will be dependent on their cluster set sizes K . We took subsets of , based on having a single shared parameter , such as all those sets derived with a log10 transform , and found the average MI in that subset . The distributions , Supplementary Table S1 , indicate that there is far more similarity in the clustering set solutions when single transform subsets are explored , than when single algorithm or distance metric subsets are explored , which is in agreement with the observations from Figure 3C . Using this same subset MI averaging method , we can look at direct comparison of the similarity of any two parameters of clustering by creating a subset made up of both parameters , Supplementary Table S1 . Rank-ordered lists for data transformations also agree with observations from Figure 3B , for example , that log10 transform clustering solutions are most related to the power transform solutions and least related to rangeScale transform solutions . Additionally , we find that Kmeans and AP clustering algorithms perform the most similarly to each other . However , both Ncut and SOM algorithm subsets show a higher similarity with other algorithms than they do with themselves . These observations could be very useful in determining a subset of clustering parameters that could be chosen to generate the maximum difference in clustering solutions . In the previous section , we developed methods for comparisons of parameter and metric information in order to understand the impact parameters have on biological enrichment , the similarities between biological metric categories , and the impact parameters have on clustering solution architectures . In addition to this , we wanted to develop a way to use the resulting MCAM information in a more traditional manner , regarding biological inference and hypothesis generation . There are a variety of ways in which one could select a manageable number of clustering solutions from for manual evaluation based on selecting the sets with the largest differences in architecture or resulting biological information . However , here we decided to look at robust relationships that result despite the high degree of variability that occurs through the implementation of a combinatorial set of clustering parameters . We did this in two ways , first by combining the enrichment results of and second by looking at the frequency of which any two phosphopeptides co-cluster in . Supplementary Table S2 lists all biological and dynamic enrichment that occurs in and the number of times that label occurs . Across the biological categories of interest , there are 539 unique labels enriched at least once in , and many of these labels appear in more than one clusterset . A histogram of the number of unique biological labels enriched in versus the number of clustersets they occur in is given in Supplementary Figure S5 . Biological labels enriched several times are considered to be ‘robust’ labels , for example GO Biological Process terms “DNA binding” and “transcription factor binding” both appear in 23% of the cluster sets . Since these robust labels may contain specific information worth exploring , we seeded the generation of a “robust cluster” based on the number of times phosphopeptides participate in a cluster with a particular enrichment term . Figure 4A and B show the results of two such terms , GO Biological Process term “MAPKKK cascade” and Cellular Compartment term “lamellipodium” and all phosphopeptides that were in a cluster giving rise to these terms at least 50% of the time . In addition to being enriched for the seed term , Figure 4A , this ‘robust’ cluster is also enriched for GO:BP term “positive regulation of DNA” proliferation . This group is composed of multiple phosphopeptides from SHC1 ( Swissprot P29353 ) , an adaptor protein recruited to the EGFR in response to EGF and upstream of the MAPK cascade . Additionally , it includes the activation sites of MAPK1 ( ERK2 , Swissprot Q1HBJ4 ) and MAPK3 ( ERK1 , Swissprot Q7Z3H5 ) and a phosphorylation on a relatively poorly characterized protein , FAM59A ( Swissprot Q9H706 ) . A recent study showed FAM59A acts in the MAPK pathway in response to EGF stimulation by binding to Grb2 in a manner that is dependent on the phosphorylation of Y453 [27] . The group associated with the cellular compartment “lamellipodium” is composed of only three phosphopeptides from the proteins cortactin ( CTTN , Swissprot Q14247 ) , paxillin ( PXN , Swissprot P49023 ) , and ENO1 ( Swissprot Q96GV1 ) , an enolase . Paxillin and cortactin are both labeled as being localized in lamellipodium , however , ENO1 , at least in full length is currently thought to be only cytoplasmic . This robust “lamellipodium” cluster also has enrichment for two sequence motifs , E . E . VyS , which is shared in both PXN and CTTN and G . . . . Oy ( O indicates a degenerate search for hydrophobic amino acids and ‘ . ’ for any amino acid ) , common to ENO1 and CTTN . These motifs may be indications of shared enzyme or binding domain recognition . These findings indicate that productive biological inference regarding function , localization and regulation is possible using the MCAM framework . An alternate method for understanding relationships between data measurements is to consider the frequency that any two phosphopeptides appear in the same cluster . We calculate this frequency , or co-occurrence , for every pairwise combination . The hierarchically clustered co-occurrence matrix is shown in Figure 4C . It appears in Figure 4C that there are essentially two groups of peptides that co-cluster a great deal with themselves and very little with the other group , listed explicitly in Supplementary Table S3 . We highlighted these two groups , and one we consider a ‘transition’ group and plot the average dynamics of those groups +/− two standard deviations in Figure 4D . The transition group clearly has the largest variability in their dynamics , which may explain why they co-cluster to some extent with both outlying groups . The “blue group” , which is also the largest , appears to have the least variance in the early time points following EGF stimulation . There is also a marked difference between the downregulation of the two outlying groups . Analysis of the co-occurrence matrix indicated there are roughly only two main dynamics in tyrosine phosphorylation dynamics within this dataset . However , given that enrichment analysis was very fruitful for finer gradations , it is clear that several levels of dataset separation are useful . With 77 phosphopeptides , there are 2 , 926 pairs of phosphopeptides to consider in the EGF4 dataset , listed in Supplementary Table S3 . In , only 63 , or 2 . 15% , of the possible relationships never occur . Fifty-percent of the possible relationships occur less than 34 times . There are 272 pairwise clusterings that occur more than half the time in , which accounts for 9 . 3% of possible relationships . We found no particular relationship between the parameters of clustering and the production of infrequently , or highly frequently , occurring phosphopeptide clustering . The most robust relationship seen is STAT3 Y705 ( isoform 1 , Swissprot P40763 ) phosphorylation with STAT3 Y704 ( isoform 2 , Refseq NP_003141 ) phosphorylation , co-clustering all but two times . This variability highlights the importance of considering more than a single clustering solution when deriving hypotheses for further testing . We have illustrated only a few examples of a large number of possibilities to demonstrate the power of MCAM in deriving biologically meaningful hypotheses . We encourage others to make full use of the Supplementary Information and Matlab scripts to continue to explore the results in this dataset . In particular , the methods that are the most promising , which are described in this section , are to explore robust relationships found either: 1 ) Through the exploration of a particular enriched label of interest , or 2 ) Through the exploration of a particular phosphopeptide of interest and those phosphopeptides that co-cluster the most robustly with it . We wished to explore the utility of MCAM in comparing independent measurements of the same network , and so we turned to a different study of ERBB network dynamics by Wolf-Yadlin et . al . [12] . In this study , which will be referred to here as the HER2 dataset , the authors were interested in the signaling response downstream of EGF and HRG and in response to HER2 amplification , which is common to several breast cancers . They used two cell lines , the wild type , or parental HMEC cell line , which is the same cell line used in the EGF4 study described previously in this paper , which has 20 , 000 HER2 receptors per cell , and 24H , a HER2 overexpressing cell line , which has 600 , 000 HER2 receptors per cell . Both cell lines express roughly the same number of EGFR and HER3 receptors . EGF ligand binds EGFR , which will drive EGFR homodimers and EGFR:HER2 heterodimers , whereas heregulin , HRG , which only binds the HER3 and HER4 family members will instead drive HER3 and HER4 containing heterodimers . When HER2 is overexpressed , the majority of the dimers in HRG treatment will be composed of HER2:HER3 and a larger proportion of EGFR:HER2 dimers will occur in response to EGF treatment . The authors found HER2 overexpressing cells were more migratory in response to stimulation by either ligand when compared to wild type HMECs . One quarter of this dataset represents the same measurement conditions as the EGF4 dataset ( 0 , 5 , 10 , and 30 minutes following a saturating dose of EGF ) of wild type HMECs ( termed P for parental ) . The remainder of the dataset is the measurement , at the same time points , of parental with HRG treatment and with HER2 overexpressing ( 24H ) cells treated with EGF and HRG . We were interested in applying MCAM to the HER2 dataset in order to see if two independent measurements of a network would agree and if signaling differences between EGF and HRG stimulation , and wild-type versus HER2 amplification , could be distinguished using MCAM and could therefore highlight potential signaling mechanisms related to the increased migratory behavior of HER2 overexpressing cells . We applied MCAM to the HER2 dataset and to five subsets of the dataset in order to tease apart: 1 ) differences in signaling between the various HER2 states and ligand treatments , and 2 ) how measurements of a conditional nature ( for example measurements at only one time point across the four conditions ) would differ from the dynamic measurements and how these differ from the full dataset . The datasets analyzed by MCAM include: the full 16-point measurement set ( Full ) , a “conditional” dataset made up of the 5-minute measurements in all conditions ( 5 minTimePoints ) , and four subsets representing the four dynamic measurements ( 0 , 5 , 10 , and 30 minutes ) of each condition named P_EGF for EGF stimulation of Parental cells , P_HRG for HRG stimulation of parental cells , 24H_EGF for EGF stimulation of HER2 overexpressing cells , and 24H_HRG for HRG stimulation of HER2 overexpressing cells . To understand more globally how the MCAM results compare across these different datasets , we found the correlation between the co-occurrence matrixes . The MCAM co-occurrence matrix of the EGF4 dataset has the highest correlation with that of the Parental EGF subset of the HER2 dataset , followed closely by the EGF treatment of the 24H cell line , there are 36 phosphopeptides common to both the EGF4 and HER2 datasets , Table 4 . There is poorer correlation with the HRG treated subsets and the worst correlation with 5 minuteTimePoint subset . In general , we see that EGF and HRG treatments correlate much better with themselves than they do with each other . For example , all EGF treatment comparisons appear in the top of the rankings of correlations , with an average correlation of 0 . 46 . However , all cross-treatment comparisons have correlation of 0 . 35 or worse , with an average correlation of only 0 . 25 . Using the co-occurrence matrix that results from MCAM application to different measurements of the same system is a useful way to derive a single , global metric for the agreement between different datasets . In order to explore the differences in clustering results of any two datasets , we are faced with analyzing an overwhelming number of comparisons . To simplify the search range , and highlight those differences that are the most extreme , and potentially the most biologically interesting , we looked for co-occurrences that move from one extreme of co-clustering to another . Specifically , we define an extreme difference as the case when two phosphopeptides change from co-clustering at least 75% of the time in one dataset to co-clustering less than 25% of the time in another . Full results for all dataset comparisons are provided in Supplementary Table S4 . There are several notable changes in cross-comparing the EGF4 dataset and all sets formed from the HER2 dataset , but the ones we found the most interesting occur in the difference between EGF treatment of parental cells versus 24H cells . In this comparison , we noticed two sites on SHC1 experience very different phenomena . In the case of parental EGF treatment , both sites co-cluster robustly with each other and with phosphorylation on EGFR pY1172 and pY1197 ( Swissprot P00533 ) , both known to bind SHC1 . However , EGF treatment of HER2 overexpressing cells indicates that although pY427 continues to robustly co-cluster with those sites on EGFR , SHC1 pY349 does not , and instead most robustly co-clusters with catenin delta-1 , CTTND1 ( Swissprot O60716 ) , phosphorylation of Y228 . Dynamics of these sites , and their most robustly co-clustered partners , under both conditions are shown in Figure 5 . We also observe large differences in the co-clustering of multiple sites on p130Cas , also known as BCAR1 ( Swissprot P56945 ) , between the two cell lines in response to EGF treatment and an extreme change in the association of ENO1 pY44 with annexin A2 ( ANXA2 , Swissprot P07355 ) phosphorylation on Y238 , where ENO1 Y44 phosphorylation associates with ANXA2 phosphorylation in parental but not HER2 overexpressing cells . The largest number of differences is observed when comparing the EGF treatment of parental cells to the 24H cell lines treated with HRG , which also has the lowest correlation of any dynamic subset comparison . Both pieces of evidence point to maximum differences in signaling dynamics when both HER2 expression levels and the stimulating ligands are altered . Further hypothesis generation can be accomplished by exploring the remaining pairwise dataset comparisons for other meaningful signaling changes that occur , which are highlighted by extreme differences in phosphopeptide co-clustering , provided in Supplementary Table S4 . Although the best agreement the EGF4 MCAM results have is with the parental EGF treatment of the HER2 MCAM results , there is less than ideal correlation . We looked to see if there were any serious disagreements by using the same ‘co-cluster swap’ method as mentioned above and found there were only two such instances , which highlight very different measurements made on GIT1 ( Swissprot Q9Y2X7 ) pY545 and EFNB2 ( Swissprot P52799 ) pY304 in HMEC cells in response to EGF treatment , Supplementary Table S4 . Interestingly , though these measurements are quite different , the extreme differences of co-occurrences with other phosphopeptides are only highlighted in two cases , both with regards to Ephrin family phosphorylation sites that are relatively similar in the two datasets , GIT1 pY545 with EPHB1 ( Swissprot P54762 ) pY600 and EFNB2 pY304 with EPHA2 pY772 ( Swissprot Q96HF4 ) . This indicates that the MCAM co-occurrence matrix is a precise way to identify similarities and discrepancies between independent measurements of a system . Since this methodology relies on comparing the relationships between measurements within a dataset and then comparing this abstracted metric across datasets , one could imagine that the measurement scheme would not need to be the same between the two datasets of interest .
MCAM was developed to capitalize on the success unsupervised learning has had on biological inference in the past and apply it to a new challenge in the field , that of understanding the function and regulation of phosphorylation in the ERBB network . Since the dynamic trajectory of a phosphorylation site will be shaped by the combination of the kinase and phosphatase activities , as well as any protective influence accumulated through binding to other proteins , and the exposure to all of these proteins by localization with or away from their regulatory proteins , then co-regulation of multiple phosphorylation sites may yield testable hypotheses regarding one or all of these possible shared traits . Encouragingly , biological information is enriched in dataset partitions derived through separation by clustering , whereas multiple random controls do not demonstrate a significant relationship with biological category enrichment . In addition to using this information to gain understanding of the phosphorylation events in the ERBB network , we also undertook to account for the fact that a large number of mathematically correct solutions can be derived through the use of various data transformations , distance metrics , algorithms , and target set sizes . We found that given a variety of clustering solutions , one could see almost any pair of phosphopeptides co-cluster , which may not necessarily be a product of co-regulation that can be linked directly to spatial localization , shared functionality , or shared enzymatic regulation . Therefore , we feel any single clustering implementation will have hypotheses likely to yield testable and supportable information as well as a good mix of those that may not . Each clustering set within an MCA represents the dimensionality reduction of a large , multidimensional dataset . However , we found there is unique value in the majority of all clustering sets in , according to the variety of biological enrichment , and so no single set could be considered as an optimal solution and evaluated in a traditional way . Therefore , the most significant development that had to be made in MCAM was to reduce the complexity of a large number of clustering sets . To this end we focused on two aspects of information: 1 ) Understanding how the parameters affected the final clustering solution , and how this related the power of any particular type of biological inference that could be made , based on statistically enriched information and 2 ) How to derive meaningful and testable biological hypotheses , through inference , concerning the function and regulation of protein phosphorylation . We found the resulting methods provided important insights when comparing similar measurements , indirectly , across multiple datasets . Through the use of this technique we found that indeed , optimal clustering parameters one would choose for clustering a dataset would vary greatly dependent on the type of information that was desired at the outcome . However , for those biological categories chosen in this study , there were a few parameters that performed badly across the board , including Hierarchical clustering and the differential transform . These are somewhat unsurprising since the differential transform will reduce the vector size by one dimension , which in a dataset with few measurements may be detrimental to the ability to separate the measurements . However , it is an important observation to mention that lessons found in one dataset should not be extrapolated to all datasets , even of the same type . Although the cosine and correlation distance metrics were uninformative in producing biological enrichment in the EGF4 dataset , they were useful in clustering various portions of the HER2 dataset . This observation highlights the importance of a broad-spectrum application of multiple parameters of clustering . It is important to consider the impact the feature selection process has on the final result in the MCAM method , which is applied during the parameter refinement , or pruning , step of the MCAM method . Given the varied relationships we observed between particular types of information and the parameters of clustering that best gave rise to them , it is likely that the final result will change with the addition of new biological features , or the alteration to existing features , such as the improvement of GO annotations . One can imagine the parameters pruned during the feature selection process would decrease with the addition of new features , which would redefine the value of a successful clustering solution . Although this suggests the results may need to be reconsidered as annotations improve , we feel this parameter refinement process helps to avoid the consideration of a large portion of currently uninformative sets . The open source nature of the MCAM software project allows for flexibility in altering the specific categories and the thresholds used during parameter refinement . We found that there is no single ‘optimum’ clustering solution , or one that performed in the top quartile of all biological metrics of interest . Depending on the application , a small number of solutions could be chosen and analyzed in a way that is more traditionally performed in the field . However , we decided to focus on allowing the agreement of many solutions to highlight a potential area of robust biological inference through the agreement of biological enrichment and alternately the agreement of co-clustering of phosphorylation dynamics . These methods produced a large list of possibly interesting biological inferences , of which we highlighted just a few possibilities to demonstrate MCAM's power . In Figure 4 we highlight two ‘robust’ clusters based on repeated enrichment of categorical terms , which creates a hybrid cluster from a combination of multiple clustering sets within , based on a particular enrichment label of interest . The hybrid cluster , like any single cluster produced from a single clustering method , represents a cluster of phosphopeptides that are strongly co-regulated . The first cluster , Figure 4A , was produced based on enrichment for co-regulation of phosphorylation sites on proteins involved in the MAPK cascade . For those proteins not currently annotated in the MAPK cascade , there is individual evidence that they are involved in regulating MAPK activity . FAM59A was recently named GAREM , which stands for Grb2-associated and regulator of Erk/MAPK activity [27] . Specifically , phosphorylation of Y453 on GAREM was required for association with Grb2 and subsequent activation of Erk by EGF stimulation . PTPN18 , a protein tyrosine phosphatase also known as BDP1 , has been implicated in regulation of HER2 directed MAPK signaling activation [28] . The study specifically found that PTPN18 was capable of inhibiting activation of mitogenic signaling . The robust co-regulation of PTPN18 Y389 phosphorylation with other components of the MAPK cascade , shown here , further implicates PTPN18 in MAPK signaling downstream of EGF stimulation , and highlights a particular mechanism for PTPN18 activity , that of Y389 phosphorylation . This modification may possibly act as a negative regulator of BDP1 activity , thereby relieving its function as a negative regulator of MAPK activity . Alternatively , if Y389 phosphorylation on PTPN18 potentiates its ability to shut down MAPK signaling , then these dynamics suggest it occurs subsequently with MAPK activation . ARHGEF5 , also known as TIM , is a RhoGEF , which has been shown to activate Rac , which is upstream of another MAPK family member , JNK . This suggests that the JNK cascade may be concurrently activated or that this particular RhoGEF has a role in the ERK1/ERK2 cascade directly . These results indicate that MCAM has been useful not only in highlighting a known co-regulation event , that of EGFR phosphorylation on sites that recruit SHC1 to the receptor with that of SHC1 phosphorylation sites that are phosphorylated following recruitment to the receptor , but also in highlighting proteins not yet generally recognized as playing a role in the EGFR/MAPK signaling pathway . This result also supports a role for these proteins in the MAPK pathway in human mammary epithelial cells as previous cell lines explored in previous studies . Additionally , the example of highlighting GAREM ( FAM59A ) phosphorylation on Y453 as playing an important biochemical role in the indicated pathway strengthens the hypothesis that PTPN18 Y389 phosphorylation is also an important biochemical mechanism in the MAPK pathway downstream of EGFR activation . Since we can find no specific study on the effect of PTPN18 Y389 phosphorylation , this hypothesis could not have come from literature mining . Additionally , we observe that traditional application of a single clustering implementation would likely not have highlighted this group of proteins in a way that would have linked PTPN18 Y389 and GAREM Y453 phosphorylation with that of EGFR Y1172 and MAPK Y187 , since this event occurs in less than 15% of clustering set solutions in . In Figure 4B we highlight another ‘robust’ cluster , which indicates the dynamics of phosphorylation of Y44 on ENO1 are very similar to that of cortactin and paxillin phosphorylation , two proteins and phosphorylation sites that play a role in cell motility [29] , [30] and are annotated as being localized to ‘lamellipodium’ . However , since there is relatively little evidence that paxillin is localized to lamellipodia , but instead has a strong association with mature focal adhesions [29] , it may be that ‘lamellipodium’ has been used as a blanket term for leading edge formations in the Gene Ontology since GO is lacking finer gradations of leading edge compartments . What is common to both paxillin and cortactin is localization in invadopodia [31] , [32] , a term not currently included in GO . Such similarity of dynamics indicates potential co-regulation , which is dependent on a variety of factors . There is evidence that all three sites may be targets of Src [30] , [33] , [34] . In addition to shared enzymatic control , similarity of dynamics might also be dependent on shared localization , especially when enzyme activation is localized to an area such as focal adhesions , invadopodia , or lamellipodium . To better conjecture where this co-regulation is occurring we looked at additional robust associations with ENO1 phosphorylation and found in addition to paxillin and cortactin this site is most closely associated with phosphorylation on a protein called AFAP1L2 , which stands for actin filament associated protein like-2 . Despite several other phosphorylation sites on focal adhesion proteins , including integrin ( ITGB4 ) , p130Cas/BCAR1 , and focal adhesion kinase ( FAK ) , none of these sites is similarly regulated . Additional evidence supporting this is that the parental EGF subset of the HER2 dataset also indicates that ENO1 and paxillin phosphorylation are tightly coupled with a third protein , Annexin A2 , which has been implicated in cell spreading and migration [35] as well as formation of invadopodia [36] . Finally , a recent proteomic study found enolase is enriched in invadopodia [37] further supporting the hypothesis that co-regulation of enolase , paxillin , and cortactin phosphorylation is through shared cellular localization . The temporal association of ENO1 phosphorylation with that of cortactin and paxillin is intriguing . ENO1 , in full and dimeric form , is a metabolic enzyme . However , a short isoform of the same gene product has been shown to block Myc transcription factor activity by binding the Myc promoter [38] . This begs the question of whether ENO1 is playing a role in lamellipodium , invadopodia , or focal adhesions , related to its metabolic activity or some other gene product of yet to be discovered functionality . ENO1 phosphorylation on Y44 has long been known and when originally studied did not show a noticeable affect directly on its catalytic activity [34] . However , in these studies it was observed that only a small percentage , roughly 5% , of total enolase was phosphorylated and so catalytic differences due to phosphorylation may have been imperceptible . Perhaps compartmentalization of a fraction of altered enzymatic activity could play an important functional role; this activity may be confined to lamellipodia , invadapodia , or focal adhesions . Despite the excess of ATP in the cell , the induction of an ATP gradient within invadopodia could conceivably act as a method of invadopodia formation , since many components , such as F-actin , would be sensitive to a shift in the ATP to ADP ratio . There are other indications that phosphorylation of metabolic enzymes may indeed be playing a functional role and driving tumor progression [39] and our results may indicate a role for enolase phosphorylation specifically in the metastatic potential of tumor progression . Finally , we found that using the concepts developed in MCAM , we could compare independent measures of the ERBB network and dissect signaling alterations occurring between different perturbations of the network , including ligand and receptor level differences . Using the co-occurrence matrix we were able to look “globally” at the differences between the networks . This study indicated that despite HER2 amplification , EGF stimulation drove signaling that was more similar between the two cell lines than it was for the same cell line under two different ligand stimulations . The greatest difference in signaling occurred when EGF stimulation of wild type cells was compared with HRG stimulation of HER2 overexpressing cells . We can dissect the signaling differences further by looking for those associations that have the most extreme differences between conditions . When we did this , we were fascinated to find that in the presence of HER2 overexpression , EGF drives very different dynamics on two phosphorylation sites of the adaptor protein SHC1 and sites on focal adhesion protein p130Cas/BCAR1 . SHC1 is known to be recruited to EGFR by two different phosphopeptide binding events and it subsequently recruits , and activates , members of the MAPK cascade . Specifically , both Y349 and Y427 phosphorylation on SHC1 has been shown to recruit Grb2 [40] . Through its SH2 domain , SHC1 is recruited to EGFR Y1197 phosphorylation and through its PTB domain , it is recruited to EGFR Y1172 phosphorylation [41] . Therefore , it is no surprise that subsequent phosphorylation of SHC1 on Y427 and Y349 would be tightly co-regulated with that of EGFR Y1172 and Y1197 phosphorylation when the network is stimulated with EGF . However , what is surprising is this co-regulation is broken for only Y349 in the presence of HER2 overexpression . Instead of being most closely co-regulated with the receptor phosphorylations it is instead most closely co-regulated with Y228 phosphorylation on catenin delta-1 , CTTND1 , a protein known to be interact with E-cadherin at cell-cell junctions [42] . The authors of the original study found that HER2 amplification drives a higher migratory potential and posited that breakup of E-cadherin junctions would be essential to this process . Our finding may therefore indicate that SHC1 plays an important role in this process and that the differential regulation of Y349 and Y427 is perhaps driven by two populations of SHC1 , one of which is localized at cell-cell junctions and which is differentially regulated , indicated by the sustained phosphorylation of SHC1 Y349 relative to Y427 phosphorylation , and a second population that is recruited to the receptor , which is probably the dominant population in EGF stimulation of wild type HMECs . The sequence surrounding catenin delta-1 Y228 matches known preferences for SHC1 SH2 recognition . In addition , multiple sites on the focal adhesion protein p130Cas/BCAR1 experience differential regulation in the presence of HER2 overexpession . Increased migration would come as a result of the disruption of both cell-cell contacts and cell-substrate contacts , so these sites might indicate a particular role in how cell-substrate contacts are disrupted in HER2 overexpressing cells . All of these data may help us in understanding the aggressiveness of tumor cells with HER2 amplification . The methodology developed here has wide applicability to data mining of all varieties . Permutations of clustering parameters and their judgement of success by pertinent categorical data has the capability of producing a wide array of solutions that together span a meaningful range of data separation . MCAM , as developed , can be applied directly to any dataset currently in the PTMScout database , and any proteomic dataset that measures phosphorylation or lysine acetylation can be added by the public to PTMScout for analysis by MCAM . Extension to any type of multidimensional biological measurements simply requires the alteration of the target categorical data . For example , in addition to using gene and protein annotation information , one could look for known transcription factors when mining gene expression data . Another benefit of MCAM is that it provides a method for comparing the relationship between independent measurements of a system , even if the overlap of measurements is incomplete . This methodology provides a new method for understanding the relationship of quantitative measurements with each other , and importantly provides a means in which to judge the outcome of a parameter of clustering with regards to resulting power of inference . This is a much-needed tool when one lacks a satisfactory ‘gold standard’ by which to evaluate the impact of various parameters of unsupervised learning .
The EGF4 and HER2 datasets were loaded into PTMScout [24] and then subsequently reassigned to default assignments through the PTMScout ‘ambiguity’ interface to ensure all isoform selections represent the best overlap with current annotations . The default dataset was then reloaded into PTMScout . This modified dataset was then exported from PTMScout's ‘subset evaluation’ page as a tab-separated file , which was then loaded into Matlab for clustering . The flat text file of the dataset was imported into Matlab based on DataRail object structures [43] . Transforms , distance metrics , and algorithms are from the Matlab environment and its toolboxes , downloaded from other resources , or developed for our purposes . Ncut code for Matlab was obtained from http://www . cis . upenn . edu/~jshi/software/ based on the algorithm description in [44] , affinity propagation ( AP ) clustering code was downloaded from http://www . psi . toronto . edu/affinitypropagation/software/apcluster . m based on the algorithm described in [18] . A self organizing map ( SOM ) Matlab toolbox was downloaded from http://www . cis . hut . fi/somtoolbox/ and is based on the algorithm described in [45] . Affinity propagation clustering was modified to accept an arbitrary distance metric , but does not accept an argument for K . SOMs only utilize the Euclidean distance metric . Average linkage distance is used in hierarchical clustering . Kmeans uses the squared value of the Euclidean distance and does not accept the Chebychev distance metric . The largest value of K is bounded by a number that would produce roughly 5 phosphopeptides per cluster , assuming a solution were to equally distribute all phosphopeptides , which in this case is Kmax = 14 . For non-deterministic algorithms , such as SOMs and Kmeans , we store the random seed so that results can be exactly reproduced , but allow the random seed to vary between individual implementations so as to ensure we do not force all implementations of the algorithm into a poorly performing local minima . Clustering assignments are written from Matlab into a tab-separated file , which is then loaded into PTMScout's ‘subset evaluation’ tool and enrichment is calculated on the PTMscout server using the ‘MCAM’ feature . Enrichment is calculated as in PTMScout's subset selection enrichment analysis , using a hypergeometric distribution calculation [24] to test for the overrepresentation of a label in a cluster compared to the full dataset . The MCAM feature has variable arguments , which were set as following: the motif algorithm branch cutoff is set at 1e-2 [46]; Scansite prediction levels of three and better are considered based on an empirical analysis of the tradeoff between the false positive rate and total hypothesis rejection ( data not shown ) ; Domain predictions of 1e-5 and more stringent are allowed; the Benjamini and Hochberg FDR procedure was used with an alpha value of 5e-2 [25] . False discovery rate correction ( FDR ) was performed at the metric and set levels as following: p-value calculations were accumulated for all tests within a category across a cluster set and the p-value satisfying an FDR alpha value of 0 . 05 was used to determine final enrichment for that metric . Three main data structures are translated from the PTMScout MCAM calculations into Matlab structures can then be used in Matlab for parameter refinement and analysis into results . Random controls were performed using two methods described in the results section . The first method randomized the data matrix and the second method randomized the metadata labels by reshuffling the vector to phosphopeptide mappings . Once enrichment structures are completed on the PTMScout server , they are loaded locally into Matlab . In order to determine parameters that produce clustersets with a relatively small degree of biological enrichment , we test the impact on the total enrichment when all clustersets generated with a particular parameter are removed from the full MCA . We perform this test systematically for every parameter used . Parameters whose removal improves the total enrichment across all categories by at least 2% without decreasing the power of enrichment in any single category by more than 10% are pruned to create . Where designated , in the HER2 dataset a total improvement of 3% was required to ensure the final sizes were roughly the same . In order to compare the clustering mappings of two clustering set solutions , we calculated the pairwise mutual information , I ( X;Y ) , between clustering sets X and Y according to the equation given below , where X is composed of clusters and Y is composed of clusters , p ( x , y ) indicates the joint probability distribution function of X and Y , and and indicate the marginal probability distributions of X and Y , respectively .
|
Proteomic measurements , especially modification measurements , are greatly expanding the current knowledge of the state of proteins under various conditions . Harnessing these measurements to understand how these modifications are enzymatically regulated and their subsequent function in cellular signaling and physiology is a challenging new problem . Clustering has been very useful in reducing the dimensionality of many types of high-throughput biological data , as well inferring function of poorly understood molecular species . However , its implementation requires a great deal of technical expertise since there are a large number of parameters one must decide on in clustering , including data transforms , distance metrics , and algorithms . Previous knowledge of useful parameters does not exist for measurements of a new type . In this work we address two issues . First , we develop a framework that incorporates any number of possible parameters of clustering to produce a suite of clustering solutions . These solutions are then judged on their ability to infer biological information through statistical enrichment of existing biological annotations . Second , we apply this framework to dynamic phosphorylation measurements of the ERBB network , constructing the first extensive analysis of clustering of phosphoproteomic data and generating insight into novel components and novel functions of known components of the ERBB network .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biotechnology",
"bioengineering",
"systems",
"biology",
"biochemistry",
"biological",
"systems",
"engineering",
"biology",
"computational",
"biology",
"proteomics",
"genetics",
"and",
"genomics",
"engineering"
] |
2011
|
MCAM: Multiple Clustering Analysis Methodology for Deriving Hypotheses and Insights from High-Throughput Proteomic Datasets
|
Dysregulation of sleep and feeding has widespread health consequences . Despite extensive epidemiological evidence for interactions between sleep and metabolic function , little is known about the neural or molecular basis underlying the integration of these processes . D . melanogaster potently suppress sleep in response to starvation , and powerful genetic tools allow for mechanistic investigation of sleep–metabolism interactions . We have previously identified neurons expressing the neuropeptide leucokinin ( Lk ) as being required for starvation-mediated changes in sleep . Here , we demonstrate an essential role for Lk neuropeptide in metabolic regulation of sleep . The activity of Lk neurons is modulated by feeding , with reduced activity in response to glucose and increased activity under starvation conditions . Both genetic silencing and laser-mediated microablation localize Lk-dependent sleep regulation to a single pair of Lk neurons within the Lateral Horn ( LHLK neurons ) . A targeted screen identified a role for 5′ adenosine monophosphate-activated protein kinase ( AMPK ) in starvation-modulated changes in sleep . Knockdown of AMPK in Lk neurons suppresses sleep and increases LHLK neuron activity in fed flies , phenocopying the starvation state . Further , we find a requirement for the Lk receptor in the insulin-producing cells ( IPCs ) , suggesting LHLK–IPC connectivity is critical for sleep regulation under starved conditions . Taken together , these findings localize feeding-state–dependent regulation of sleep to a single pair of neurons within the fruit fly brain and provide a system for investigating the cellular basis of sleep–metabolism interactions .
Dysregulation of sleep and feeding has widespread health consequences , and reciprocal interactions between these processes underlie a number of pathologies [1–4] . Sleep loss correlates with increased appetite and insulin insensitivity , while short-sleeping individuals are more likely to develop obesity , metabolic syndrome , type 2 diabetes , and cardiovascular disease [1 , 3 , 4] . Although the neural basis for sleep regulation has been studied in detail , little is known about how feeding state and changes in metabolic function modulate sleep [5 , 6] . Understanding how sleep and feeding states are integrated may provide novel insights into the comorbidity of disorders linked to sleep and metabolic regulation . Animals balance nutritional state and energy expenditure in order to achieve metabolic homeostasis [6 , 7] . In both flies and mammals , diet potently affects sleep regulation , supporting the notion that sleep and metabolic state interact [5 , 6 , 8] . Starvation leads to sleep loss or disrupted sleep architecture , presumably to induce foraging behavior , while high-calorie diets have complex effects on sleep depending on macronutrient content [9–12] . Behavioral and physiological responses to changes in feeding state are modulated both by cell-autonomous nutrient centers in the brain that detect changes in circulating nutrients and through communication between brain and peripheral tissues [13] , yet the neural basis for the integration of sleep and feeding state remain poorly understood . The fruit fly , D . melanogaster , provides a powerful model for investigating sleep regulation . Flies display all the behavioral hallmarks of sleep , including extended periods of behavioral quiescence , rebound following deprivation , increased arousal threshold , and species-specific changes in posture [14 , 15] . Many genetic mechanisms regulating sleep are conserved from flies to mammals . In addition , high-throughput systems for sleep analysis in Drosophila have led to the identification of both novel and highly conserved sleep genes [16 , 17] . Further , stimulants including caffeine , amphetamine , and cocaine have been shown to suppress sleep in flies [15 , 18 , 19] . Thus , at the molecular , pharmacological , and behavioral levels , flies provide a model for studying genetic regulation of mammalian sleep . A number of genes and neurons that are required for the integration of sleep and feeding states have been identified , including core-circadian clock genes , metabolic hormones , and sensory neurons [9 , 20–22] . While many identified regulators of sleep–metabolism interactions broadly impact these processes [6] , a mutation of the DNA/RNA binding protein translin ( trsn ) disrupts starvation-induced sleep suppression without affecting sleep or metabolic regulation under fed conditions . Targeted knockdown in approximately 30 leucokinin ( Lk ) neurons phenocopies trsn mutants , raising the possibility that these neurons are required for the integration of sleep and metabolic state [23] . Here , we identify a single pair of Lk neurons in the lateral horn of the fly brain that are required for the integration of sleep and metabolic state . These neurons project near the insulin-producing cells ( IPCs ) , which are critical modulators of sleep and metabolic regulation [24–26] . Lateral Horn leucokinin ( LHLK ) neurons are dispensable for sleep under fed conditions but are required for starvation-induced sleep suppression . Functional imaging reveals that LHLK neurons have reduced activity in response to glucose application and increased activity under starved conditions . The identification of single neurons that integrate sleep and metabolic state provide a model for investigating the cellular mechanisms regulating the integration of sleep and metabolic state .
Leucokinin ( Lk ) neuropeptide has been implicated in regulation of feeding , sleep , and circadian activity [27 , 28] . To determine whether Lk is required for metabolic regulation of sleep , we measured sleep under fed and starved conditions in mated female flies with disrupted Lk expression . In agreement with previous findings , driving Lk-RNAi ( RNA interference ) under control of Lk-galactose-responsive transcription factor ( GAL4 ) ( yeast transcription factor , Lk-GAL4>upstream activation sequence [UAS]-dicer-2 [dcr2] , Lk-RNAi ) significantly reduced Lk expression ( Fig 1A and 1B and S1B Fig ) . Control flies harboring Lk-RNAi or Lk-GAL4 transgenes alone significantly suppressed sleep during starvation , while no significant differences were detected between the fed and starved states in Lk>Lk-RNAi knockdown flies ( Fig 1C and 1D ) . To confirm that these phenotypes are not due to off-target effects , we measured starvation-induced sleep suppression in two Lk mutants . Lkc275 is a hypomorphic allele containing a piggyBac element upstream of the Lk gene transcription start site ( S1A Fig ) , with approximately 30% reduction in Lk levels ( Fig 1E and 1F and S1B Fig ) [28] . Flies homozygous for Lkc275 failed to suppress sleep when starved ( Fig 1G and 1H ) . We also used Clustered Regularly Interspaced Short Palindromic Repeats ( CRISPR/Cas9 ) gene-editing to generate a recombinant transgenic line ( Lk−/− ( GAL4 ) ) by replacing bases 1 to 7 downstream of the Lk ATG start codon with a GAL4-element containing cassette ( S1A Fig ) . Lk protein was not detected in the brains of Lk−/− ( GAL4 ) mutants , confirming that this genetic modification resulted in a robust reduction of Lk function ( Fig 1I and 1J and S1B Fig ) . Sleep while fed did not differ between Lk−/− ( GAL4 ) lines and w1118 controls , suggesting that Lk is not required for sleep regulation under fed conditions ( Fig 1K and 1L ) . Conversely , control w1118 flies , but not Lk−/− ( GAL4 ) flies , robustly suppressed sleep under starved conditions during the day and night , indicating that Lk is required for starvation-induced sleep suppression ( Fig 1K and 1L ) . In agreement with previous findings , starvation induces hyperactivity in control flies ( S1C and S1D Fig ) [20 , 29] . Starvation did not alter waking activity in Lkc275 and Lk−/− ( GAL4 ) flies , indicating that Lk is required for both starvation-induced changes in sleep regulation and hyperactivity ( S1C and S1D Fig ) . To determine whether the sleep phenotype was caused by loss of Lk , we restored Lk in the background of each mutant and measured sleep . Pan-neuronal rescue ( Lkc275; embryonic lethal abnormal vision [elav]-GAL4/UAS-Lk ) restored starvation-induced sleep suppression ( S1E Fig ) and starvation-induced hyperactivity ( S1F Fig ) . Rescue flies did not differ from heterozygous controls ( Lkc275; elav-GAL4/+ or UAS-Lk;Lkc275/+ ) . The GAL4 insertion in Lk−/− ( GAL4 ) drives expression to a pattern similar to Lk antibody , providing the opportunity to restore Lk to its endogenous expression pattern ( S1G Fig ) . Similar to Lkc275 , rescue in Lk neurons ( Lk−/− ( GAL4 ) ; UAS-Lk ) restored starvation-induced sleep suppression ( S1H Fig ) and starvation-induced hyperactivity ( S1I Fig ) , confirming that Lk is required for the metabolic regulation of sleep . Flies heterozygous for Lk−/− ( GAL4 ) failed to suppress sleep in response to starvation ( Fig 1K ) and displayed reduced starvation-induced hyperactivity compared to control ( S1D Fig ) , raising the possibility that haploinsufficiency impacts sleep regulation . This result is consistent with the notion that the moderate reduction in Lk levels in Lkc275 mutants can affect diverse behavioral phenotypes [27] . Expression of a rescue transgene in Lk neurons of flies heterozygous for Lk ( Lk+/− ( GAL4 ) ;UAS-Lk ) also restored starvation-induced sleep suppression ( S1J Fig ) , confirming the specificity of phenotype in Lk−/− ( GAL4 ) flies . Taken together , three independent genetic manipulations that perturb Lk expression inhibit starvation-induced changes in sleep and activity . Leucokinin antibody labels the LHLK neurons and the subesophoageal ganglion Lk ( SELK ) neurons , as well as a number of abdominal Lk ( ABLK ) neurons in the ventral nerve cord ( Fig 2A ) . To localize the population of neurons that regulate starvation-induced sleep suppression , we restricted GAL4 expression primarily to the brain through the expression of GAL80 , a GAL4 repressor , in the ventral nerve cord using teashirt-GAL80 ( tsh-GAL80 ) [30] . Expression of CD8::GFP ( Lk-GAL4>CD8:green fluorescent protein [GFP];tsh-GAL80 ) revealed tsh-GAL80 blocks expression in all but two ventral nerve cord neurons without affecting expression in the brain ( Fig 2B ) . Silencing the remaining Lk neurons with light-chain tetanus toxin ( TNT ) ( tsh-GAL80;Lk-GAL4>UAS-TNT ) abolished starvation-induced sleep suppression , phenocopying the effects of silencing all Lk neurons ( Lk-GAL4>UAS-TNT ) ( Fig 2C and 2D ) [31] . Further , no differences in sleep were detected between groups in fed flies , and there was no effect of expressing an inactive variant of TNT light chain ( impTNT ) in Lk neurons ( Fig 2C and 2D ) . These findings suggest Lk neurons within the brain are required for sleep–metabolism interaction . It has previously been reported that Apterous-GAL4 ( Apt-GAL4 ) drives expression in the LHLK neurons , as well as neurons in the optic lobe and antennal mechanosensory and motor centers ( AMMCs ) , and a small population of mushroom-body neurons ( S2A Fig ) [27 , 32] . Immunostaining with Lk antibody in Apt-GAL4>UAS-mCD8::GFP flies confirmed colocalization exclusively within the LHLK neurons ( S2A and S2B Fig ) . To functionally assess the role of LHLK neurons , we genetically silenced LHLK neurons , as well as other non-Lk cells labeled by Apt-GAL4 . Silencing neurons labeled by Apt-GAL4 ( Apt-GAL4>UAS-TNT ) inhibited starvation-induced sleep suppression and promoted sleep while fed , while no effects were observed in flies expressing impTNT ( Apt-GAL4>UAS-impTNT ) ( Fig 2E ) . While these findings suggest a role for Apt-GAL4–labeled neurons in sleep regulation , it is possible that the phenotype is independent of Lk function . To verify that the sleep phenotype was due to blocking Lk release from LHLK neurons , we sought to disrupt Lk function selectively in Apt-GAL4–labeled neurons . Expression of Lk-RNAi in Apt-Gal4 neurons ( Apt-GAL4>UAS-dcr2 , Lk-RNAi ) disrupted starvation-induced sleep suppression ( Fig 2F ) . As a complementary approach , we restored Lk to Apt-GAL4 neurons in the Lkc275 mutant background and measured sleep . Rescue in Apt-GAL4 neurons ( Lkc275; Apt-GAL4>Lkc275; UAS-Lk ) restored starvation-induced sleep suppression to heterozygote control levels ( Apt-GAL4; Lkc275/+ and UAS-Lk; Lkc275/+ ) , whereas flies harboring either GAL4 or UAS in the Lkc275 mutant background failed to suppress sleep ( Apt-GAL4; Lkc275 and UAS-Lk; Lkc275; Fig 2G ) . These data support a role for the LHLK neurons in starvation-induced sleep suppression but do not rule out the possibility that other neurons labeled by Apt-GAL4 also contribute to this phenotype . To complement genetic silencing experiments , we sought to precisely ablate the LHLK neurons and measure their role in starvation-induced sleep suppression ( Fig 3A ) . Multiphoton microscopy has been used in diverse genetic models for targeted ablation of neuronal cell types [33–35] . All adult Lk neurons are present in third-instar larvae and labeled by the Lk-GAL4 . The SELK and anterior Lk ( ALK ) neurons are easily visualized , providing the opportunity to independently ablate these subtypes and measure the effect on adult behavior in an intact animal ( Fig 3B ) . We selectively induced bilateral ablations of LHLK neurons or unilateral ablation of two control ALK neurons in immobilized third-instar larvae with a titanium sapphire multiphoton laser . Ablation of individual neurons could be visualized in larvae as a disruption of the GFP-labeled neuronal cell body ( Fig 3C ) . Following ablation , larvae were transferred back into food vials , and 5- to 7-day–old adult flies were tested for sleep under fed and starved conditions . After behavioral testing , brains were dissected and imaged to verify selective bilateral ablation of the LHLK neurons or unilateral ablation of two ALK neurons ( Fig 3F ) . Flies with ablated ALK neurons suppressed sleep during starvation similarly to controls ( Fig 3D and 3E ) . Conversely , bilateral ablation of the LHLK neurons abolished starvation-induced sleep suppression without affecting sleep while fed , revealing an essential role for the LHLK neurons in the integration of sleep and metabolic state ( Fig 3D , 3E and 3F ) . The finding that LHLK neurons are required for starvation-induced sleep suppression raises the possibility that the activity of Lk neurons is modulated by nutritional state . We selectively expressed a GFP-calmodulin and MP13 peptide sequence ( GCaMP6m ) -mCherry ( UAS-Gerry ) fusion protein that allows for ratiometric detection of Ca2+ activity [36 , 37] in Lk neurons and measured the response to nutrients . The brains of fed or 24-hr–starved flies were imaged for GCaMP and mCherry signal ex vivo ( Fig 4A ) . Flies expressing Gerry in Lk neurons suppressed sleep similarly to control flies harboring Lk-GAL4 alone , indicating that expression of the Ca2+ sensor does not affect starvation-induced regulation of sleep ( S3A Fig ) . The GCaMP/mCherry ratio was elevated in the LHLK neurons of starved flies compared to fed controls , suggesting these neurons are more active during starvation ( Fig 4B ) . Conversely , no difference in the GCaMP/mCherry ratio between the fed and starved states was detected in the SELK neurons ( Fig 4C ) . In mammals , the activity of some sleep- and wake-promoting neurons are directly modulated by glucose and other circulating nutrients [38 , 39] . It is possible that the activity of Lk neurons is modulated in accordance with feeding state by sensory detection of tastants or indirectly result upon detection of changes in circulating nutrients . To differentiate between these possibilities , the brains of fed flies were removed and treated with either glucose or the competitive inhibitor of glycolysis , 2 deoxy-glucose ( 2DG ) [40–42] . Application of glucose reduced Ca2+ activity in LHLK neurons compared to controls treated with Drosophila artificial hemolymph alone , suggesting these neurons are sensitive to circulating glucose ( Fig 4D ) . To verify that application of glucose was specific to LHLK neurons , we applied glucose or artificial hemolymph while measuring Ca2+ response in SELK neurons . No significant differences in Ca2+ activity were found between controls and glucose application ( S3B Fig ) . Further , the combined application of 2DG and glucose increased Ca2+ activity to levels greater than hemolymph alone , mimicking the starved state ( Fig 4D ) . Taken together , these findings indicate that the activity of Lk neurons are modulated in accordance with nutrient availability and support the notion that the LHLK neurons are more active during starvation , thereby suppressing sleep . To examine whether the activity in Lk neurons is modulated by feeding state in an intact animal , we performed in vivo recordings in tethered flies ( Fig 4E ) . Briefly , a portion of the head capsule was removed so that the LHLK neurons were accessible . The activity of LHLK neurons was then recorded in flies that had been previously fed or starved for 24 hr . In agreement with ex vivo findings , the GCaMP/mCherry ratio was elevated in the LHLK neurons of starved flies , fortifying the notion that Lk neurons are more active during starvation ( Fig 4F ) . Refeeding flies with standard food reduced the GCaMP/mCherry ratio 3 hr after refeeding ( Fig 4F ) . Taken together , these findings suggest the activity of LHLK neurons are modulated by circulating nutrient levels . The identification of nutrient-dependent changes in activity of LHLK neurons raises the possibility that cell-autonomous nutrient sensors or signaling pathways function within Lk neurons to modulate sleep . To identify regulators of sleep that modulate the activity of Lk neurons , we expressed RNAi targeted to 28 RNAi lines encoding putative nutrient sensors or signaling pathways using Lk-GAL4 and measured starvation-induced changes in sleep ( S4A Fig ) . RNAi knockdown of AMP-activated protein kinase alpha ( AMPKα ) in Lk neurons ( Lk-GAL4>UAS- AMPKα ) alone abolished starvation-induced sleep suppression compared to GAL4 controls crossed to the isogenic host strain for the RNAi library ( Fig 5A and 5B and S4A Fig ) [43] . Feeding did not differ in flies expressing AMPKα-RNAi in Lk neurons , suggesting the sleep phenotype is not due to generalized changes in hunger ( S4B Fig ) . Targeting AMPKα-RNAi with a second , independently derived RNAi line also abolished starvation-induced sleep suppression ( Lk-GAL4>UAS-dcr2 , UAS-AMPKα-RNAi #2; S4C Fig ) . Genetically restricting AMPK knockdown in flies harboring tsh-GAL80 ( tsh-GAL80;Lk-GAL4>AMPKα-RNAi ) also impaired starvation-induced sleep suppression ( Fig 5C and 5D ) , suggesting AMPKα functions in the Lk neurons within the brain to regulate sleep . To determine whether inhibition of AMPK signaling changes the activity of Lk neurons to resemble a starved-like state , we genetically expressed AMPKα-RNAi under control of Lk-GAL4 and measured neuronal activity in vivo using UAS-Gerry ( Fig 5E and 5F; Lk-GAL4>UAS-AMPKα-RNAi; UAS-Gerry ) . Genetic disruption of AMPKα increased Ca2+ activity in LHLK neurons of fed flies compared to flies expressing UAS-Gerry alone ( Fig 5E ) . The increase Ca2+ activity phenocopies changes found in starved control flies , suggesting that the loss of AMPK increases the activity of Lk neurons , thereby suppressing sleep ( Fig 5F ) . Together , these findings suggest AMPK is active within LHLK neurons during the fed state , and reduced AMPK signaling during starvation increases LHLK activity . Lk signals through a single Lk receptor ( Lkr ) that is highly expressed in the IPCs and the dorsal fan-shaped body ( dFSB ) , both of which have been implicated in sleep regulation [24 , 44–46] . To determine the role of Lkr in sleep regulation , we used CRISPR/Cas9 gene-editing to generate a recombinant transgenic line ( Lkr−/− ( GAL4 ) ) with a GAL4 element inserted 106 to 111 base pairs preceding the ATG translational start site , disrupting its function . Consistent with previous reports of Lkr expression , transgene expression in Lkr−/− ( GAL4 ) >UAS-mCD8::GFP flies labeled the IPCs , dFSB , and a number of other brain regions ( Fig 6A ) . To determine the role of Lkr in sleep , we tested flies for sleep under fed and starved conditions . Lkr−/− ( GAL4 ) flies failed to suppress sleep when starved , phenocopying loss of Lk function ( Fig 6B ) . Restoring Lkr to Lkr mutant flies ( Lkr−/− ( GAL4 ) >UAS-Lkr ) rescued starvation-induced sleep suppression ( Fig 6B ) . The promoter-fusion R65C07-GAL4 predominantly labels the dFSB , while R67D01-GAL4 primarily labels the IPCs ( [47] , Fig 6C and 6F ) . Silencing R65C07-GAL4–labeled neurons with TNT reduced sleep in fed flies , consistent with a sleep-promoting role for the dFSB ( [45]; Fig 6D ) . However , there was no effect of Lkr knockdown in these cells ( R65C07>UAS-dcr2 , Lkr-RNAi ) , suggesting that Lk does not signal through the dFSB to modulate sleep . Conversely , genetic silencing or expression of Lkr-RNAi in R67D01-GAL4 neurons that label the IPCs abolished starvation-induced sleep suppression ( Fig 6G and 6H ) . Together , these findings suggest Lkr function in the IPCs is required for starvation-induced sleep suppression . To validate a role for the IPCs , we selectively targeted Lkr in neurons labeled by Drosophila insulin-like peptide 2 ( Dilp2 ) ( Fig 6I ) . Selectively knocking down Lkr in Dilp2 neurons ( Dilp2-GAL4>UAS-dcr2 , Lkr-RNAi ) prevented starvation-induced sleep loss , indicating that Lkr is required in Dilp2 neurons for starvation-induced sleep suppression ( Fig 6J and 6K ) .
Our findings reveal that Lk signals through a single Lkr to integrate sleep and metabolic state . Previous studies have identified a number of genes required for starvation-induced changes in sleep or locomotor activity , yet many of these genes have pleiotropic functions on behavior or metabolic function [20 , 29 , 48 , 49] . For example , the glucagon-like adipokinetic hormone ( AKH ) is responsible for energy mobilization , and genetic disruption of AKH induces obesity and abolishes starvation-induced hyperactivity [29 , 50 , 51] . Similarly , the circadian transcription factors clock and cycle are required for starvation-dependent regulation of behavior , and loss of function affects sleep both in fed and starved conditions [20] . Conversely , neuropeptide F functions within a subpopulation of circadian neurons and is selectively required for metabolic regulation of sleep [22] . Our findings that genetic manipulations that inhibit Lk signaling selectively disrupt starvation-induced modulation of sleep suggest that different neural mechanisms regulate sleep under basal conditions and in response to environmental perturbation . The failure of Lk mutants to suppress sleep under starved conditions phenocopies mutation of the RNA/DNA binding protein trsn . Loss of trsn does not impact feeding behavior but impairs starvation-induced sleep suppression , suggesting that trsn is not generally required for hunger-induced behavior [23] . While trsn is broadly expressed in the fly nervous system [52 , 53] , we previously found that selective knockdown of trsn in Lk neurons disrupted starvation-induced sleep suppression [54] . These findings raise the possibility that trsn functions to regulate changes in the physiology of Lk neurons to modulate sleep under starved conditions . A central question is how Lk neurons modulate numerous complex behaviors and physiological processes . In adults , Lk is expressed in four pairs of neurons in the brain and 11 pairs in the ventral nerve cord , which regulate diverse behaviors and physiological processes [55–57] . Recent work suggests the ABLK neurons in the thoracic ganglion are critical regulators of water consumption and contribute to the altered stress resistance and water content in Lk mutant flies [46] . The SELK neurons connect the gustatory receptors to the subesophageal ganglia and ventral nerve cord . Although a specific function has not been identified to SELK neurons , silencing of all Lk neurons disrupts gustatory behavior , and a mutation in the Lk locus affects meal size [28 , 58] , raising the possibility that these behaviors are regulated by SELK neurons . Lastly , the LHLK neurons project to the superior lateral protocerebrum , medial protocerebrum , and peduncle and axonal stalk of the mushroom bodies [56] . The LHLK neurons we identify here as regulating sleep–metabolism interactions receive inputs from clock neurons and are thought to modulate locomotor activity and sleep [27] . Supporting the notion that LHLK neurons are outputs of the clock , previous work has found that silencing Lk within these neurons attenuates circadian rhythms [27] . Conversely , we find that silencing of these neurons has little effect on sleep yet abolishes starvation-induced sleep suppression . Together , these findings support a role for LHLK neurons in sleep regulation , yet discrepancies remain about the conditions under which these neurons regulate sleep . Lk neurons have also been implicated in the modulation of sleep following the ingestion of a meal . Postprandial sleep is enhanced in Lk-deficient flies and reduced in flies with thermogenetically activated Lk neurons sleep , whereas Lk knockdown increases the probability of falling asleep after a meal [59] . Similarly , we found that knockdown of Lk in Lk-expressing neurons results in increased sleep during starvation . In contrast , it was previously reported that Lk signals the fan-shaped body neurons to regulate postprandial sleep [59] . These results raise the possibility that distinct circuitry regulating starvation-induced sleep suppression and postprandial sleep , which could be influenced by distinct time-scales at which both behaviors are being executed . In addition , it is possible that a distinct subpopulation of Lk neurons is responsive to individual nutrients in comparison to starvation itself , which is sensed by LHLK neurons . Taken together , these findings suggest different neural mechanisms underlie Lk-dependent regulation of postprandial feeding and circadian rhythms . The Drosophila genome encodes for a single Lk target , the Lkr , that is expressed in the lateral horn , the ventral nerve cord , the IPCs , and the sleep-promoting fan-shaped body [27 , 28 , 59] . The fan-shaped body , a subregion of the Drosophila central complex , is a primary sleep-promoting region [45 , 60] , while the IPCs are proposed integrators of sleep and feeding state [61] . Previous studies suggest that Lkr function within the fan-shaped body is required for proper regulation of circadian rhythms and postprandial sleep [27 , 59] . Here , we find that targeted knockdown of Lkr within the IPCs phenocopies LHLK ablation , suggesting the LHLK neurons signal to the IPCs to modulate starvation-induced regulation of sleep . Supporting these findings , recent work has shown that Lk and Lkr mutants display increased levels of Dilp2 and Dilp3 immunoreactivity in the IPCs [46] . However , anterograde trans-synaptic labeling revealed no direct synaptic inputs between Lk neurons and the IPCs , raising the possibility that that Lk inputs to the IPCs could occur via paracrine signaling [46] . Taken together with previous studies , these findings suggest different Lk targets regulate postprandial feeding , circadian rhythms , and sleep or that Lk functions through paracrine signaling to modulate different targets . In Drosophila , a number of circulating nutrients including fructose , trehalose , and glucose have been found to affect central brain physiology and behavior [42 , 62 , 63] . While nutrients may be detected by gustatory receptors expressed in the periphery to regulate sleep [9 , 64] , sugar receptors and transporters are also expressed in the brain [65] . The identification of LHLK neurons as being active under starvation conditions and suppressed by glucose provide a system to investigate feeding-state–dependent changes in neural activity . A number of neurons in the fly brain are acutely regulated by feeding state , including the starvation-active Taotie neurons that inhibit IPCs of the pars intercerebralis to regulate insulin-like peptide release under nutrient deprivation conditions [66–68] . Conversely , the IPCs function as cell-autonomous nutrient sensors that are activated by glucose through the inhibition of KATP channels [69] . Further , the LHLK nutrient phenotype is similar to the neurons within the ellipsoid body labeled by the sodium/glucose co-transporter SLC5A11 that are active during starvation and promote feeding [41 , 70] . SLC5A11 and its cognate neurons are required for a variety of hunger-induced feeding behaviors , but the effect on sleep has not been identified [70] . Our screen found that knockdown of SLC5A11 in Lk neurons did not affect starvation-induced sleep suppression , suggesting alternative regulators of sleep . The identification of LHLK neurons as starvation-active neurons provides a system for identification of additional nutrient sensors that regulate sleep . AMPK functions as a cell-autonomous regulator of energy allocation and induces physiological changes associated with starvation [71 , 72] . AMPK consists of a heterotrimeric complex that is activated by AMP and modulates diverse intercellular signaling pathways , including mammalian target of rapamycin ( mTOR ) , forkhead box ( FoxO ) , and sirtuin-1 ( SIRT1 ) [73] . Canonically , AMPK is activated during starvation and increases neuronal activity , though this effect varies by neuronal subtype [74 , 75] . For example , in Caenorhabditis elegans , starvation-induced AMPK activation leads to inhibition of neurons that modulate local search behavior in response to food deprivation while promoting activity in neurons that trigger dispersal behavior [76] . Here , we find that knockdown in LHLK neurons using multiple independently derived RNAi lines results in flies that reduce sleep under the fed state and increases the activity of LHLK neurons , similar to neural activity seen during the starved state . Ubiquitous disruption of AMPK in Drosophila induces hypersensitivity of the locomotor response to starvation and reduces starvation resistance [74] . Conversely , we find that selectively disrupting AMPK function in Lk neurons promotes starvation-induced hyperactivity and sleep loss during the fed state , suggesting a neural-circuit–specific function for AMPK . While our findings suggest that AMPK functions as an important modulator of LHLK neuronal activity and state-dependent changes of activity within LHLK neurons , it is also possible that AMPK generally modulates the activity of Lk neurons , resulting in sleep loss . The findings that the activity of SELK neurons is not elevated during starvation raises the possibility of neuron-specific AMPK function . Taken together , the identification of LHLK neurons as critical modulators of sleep–metabolism interactions provides a system for identifying novel nutrient-sensing and signaling mechanisms that modulate sleep . These findings illustrate the need to determine how Lk neurons modulate different aspects of sleep regulation , including their reported role as circadian output neurons , in regulation of sleep–metabolism interactions , and in postprandial sleep regulation [23 , 27 , 59] . Further investigation of feeding-state–dependent changes in Lk signaling and the identification of neuronal inputs and targets of LHLK neurons will provide mechanistic insight into how animals integrate sleep with changes in their internal and external environments .
Flies were grown and maintained on standard food ( Bloomington Recipe , Genesee Scientific , San Diego , CA , USA ) . Flies were kept in incubators ( Dros52; Powers Scientific , Warminster , PA , USA ) at 25 °C on a 12:12 LD cycle with humidity set to 55%–65% . The background control line used in this study is the w1118 fly strain , and all experimental flies were outcrossed 6–8 generations into this background . All the experiments performed in this manuscript used mated female flies . The following fly strains were ordered from Bloomington Stock Center: w1118 ( 5905; [77] ) , Lkc275 ( 16324; [28] ) , elav-GAL4 ( 8765; [78] ) , Apt-GAL4 ( 3041; [79] ) , UAS-TNT ( 28996; [31] ) , UAS-impTNT ( 28840; [31] ) , UAS-mCD8::GFP ( 32186; [80] ) , UAS-dcr2 ( Chr II;24650; [43] ) , UAS-dcr2 ( Chr III;24651; [43] ) , AMPKα-RNAi#2 ( 35137; [81] ) , UAS-Lkr-RNAi ( 65934; [81] ) , UAS-luciferase ( 31603; [81] ) , Lkr-GAL4 ( 39344; [47] ) , and Lkr-GAL4 ( 39412; [47] ) . The following lines were generated in this study: Lk−/− ( GAL4 ) , Lkr−/−GAL4 ) , and UAS-Lk . UAS-Gerry was a kind gift from Greg Macleod , Lk-GAL4 and Dilp2-GAL4 from Young-Joon Kim , and UAS-Lkr from Bader Al Anzi [28] . tsh-GAL80 [30] was provided by Julie Simpson . Drosophila lines used in the RNAi screen and UAS-Lk-RNAi ( 14091 ) originate from the VDRC library [43] and are described in Table 1 . Lk−/− ( GAL4 ) and Lkr−/− ( GAL4 ) were generated by Wellgenetics ( Taipei City , Taiwan ) using the CRISPR/Cas9 system to induced homology-dependent repair ( HDR ) using one guide RNA ( gRNA ) ( Lk−/− ( GAL4 ) : GATCTTTGCCATCTTCTCCAG and Lkr−/− ( GAL4 ) : GTAGTGCAATACATCTTCAG ) . At the gRNA target site , a donor plasmid was inserted containing a GAL4::VP16 and floxed 3xP3-RFP cassette [82] . For Lk−/− ( GAL4 ) , following the translational ATG start site , bases 1 to 7 were replaced by the knock-in cassette . For Lkr−/− ( GAL4 ) , preceding the ATG start site , bases 111 to 106 were replaced by the knock-in cassette . All lines were generated in the w1118 background [77] . Proper insertion loci for both mutations were validated by genomic PCR . The full-length open reading frame of Lk was amplified from the Lk-pOT2 plasmid ( Drosophila Genomics Resource Center [DGRC] , #1378621 ) using specific primers ( forward primer: GCCTTTGGCCGTCAAGTCTA and reverse primer: CTCCAAGTACCGCAGGTTCA ) generated by Integrated DNA Technologies ( Coralville , IA , USA ) . Amplified sequence was inserted into the pENTER vector ( Invitrogen ) via TOPO cloning and subsequently recombined into pTW destination vector ( DGRC , #1129 ) using standard gateway cloning protocol as per manufacturer’s instructions ( Invitrogen , Carlsbad , CA , USA ) . The plasmids were verified by sequencing ( Genewiz , Morrisville , NC , USA ) . Transgenic lines were established via phiC31-mediated integration at the attp40 landing site [83] on the second chromosome ( BestGene , Chino Hills , CA , USA ) . The DAMS detects activity by monitoring infrared beam crossings for each animal [84] . These data were used to calculate sleep information by extracting immobility bouts of 5 minutes using the Drosophila Counting Macro [85 , 86] . For experiments examining the effects of starvation on sleep , flies were kept on a 12:12 LD cycle . Mated female flies were briefly anesthetized with CO2 and placed into plastic tubes containing standard food . All flies were given 24 hr to recover after being anesthetized . Activity was recorded for 24 hr in food tubes prior to transferring flies into tubes containing 1% agar diluted in dH2O ( Fisher Scientific ) at Zeitgeber time ( ZT ) 0 . Activity was monitored for an additional 24 hr on agar . For the screen , percent change in sleep during starvation was calculated as the sleep duration on agar minus the sleep duration in food tubes , divided by the sleep duration in food tubes for each fly assayed multiplied by a hundred [11 , 54] . The brains of 5- to 7-day–old female flies were dissected between ZT 4–ZT 9 in ice-cold phosphate-buffered saline ( PBS ) and fixed in 4% paraformaldehyde , PBS , 0 . 5% Triton-X for 30 minutes as previously described [87] . Brains were then rinsed 3× with PBS , 0 . 5% Triton-X ( PBST ) for 10 minutes and overnight . In the following day , brains were incubated for 24 hr in primary antibody ( 1:1 , 000 rabbit anti-Lk [88] and mouse 1:20 nc82; Iowa Hybridoma Bank , University of Iowa , Iowa City , IA , USA ) diluted in PBST at 4 °C . Brains were rinsed in PBST 3× for 10 minutes and placed in secondary antibody ( 1:400 donkey anti-rabbit Alexa 555 and 1:200 donkey anti-mouse Alexa 647; Thermo Fisher Scientific , Waltham , MA , USA ) , diluted in PBST for 90 minutes at room temperature . Finally , all samples were washed in PBST for a total of 120 minutes and mounted in Vectashield ( VectorLabs , Burlingame , CA , USA ) . Samples were imaged in 2-μm sections with a Nikon A1R confocal microscope ( Nikon , Tokyo , Japan ) using a 20× or 60× oil immersion objective . Images were then processed with NIS Elements 4 . 40 ( Nikon ) . Briefly , flies were maintained on standard fly food . At ZT 0 , flies were transferred to vials containing 1% agar , 5% sucrose , and 2 . 5% blue dye ( FD&C Blue Dye No . 1 ) . Following 30 minutes of feeding , flies were flash frozen on dry ice , and four flies were homogenized in 400 μL PBS ( pH 7 . 4 , Thermo Fisher Scientific ) per sample . Color spectrophotometry was then used to measure absorbance at 655 nm in a 96-well plate reader ( iMark; Millipore Sigma , Burlington , MA , USA ) . Baseline absorbance was determined by subtracting the absorbance measured in non-dye–fed flies from each experimental sample . Five- to seven-day–old female flies were collected and placed in vials containing fresh food ( fed ) or a wet KimWipe paper ( starved ) for 24 hr . All experiments were done between ZT 4–ZT 7 to account for rhythmic excitability of Lk neurons [27] . For imaging brain explants , previously established methods for calcium imaging were used with modifications [42 , 65] . Brains of fed or 24-hr–starved flies were dissected and placed in glass wells ( Pyrex , Corning , Corning , NY , USA ) containing artificial hemolymph ( 140 mM NaCL , 2 mM KCl , 4 . 5 mM MgCl2 , 1 . 5 mM CaC2 , and 5 mM HEPES-NaOH with pH 7 ) and allowed a 5-minute recovery period before being recorded . For 2DG experiments , fed brains were dissected and placed in 400 mM 2DG ( Sigma Aldrich ) in artificial hemolymph , 200 mM glucose ( Sigma Aldrich ) in artificial hemolymph , or artificial hemolymph alone for a total of 70 minutes . Every 20 minutes , solutions were exchanged . Coverslips were treated with poly-L-lysine ( Sigma Aldrich ) to ensure that brains were in the same position during imaging and placed onto chamber ( RC-21BBDW; Warner Instruments , Hamden , CT , USA ) . Fly brains were bathed in artificial hemolymph solution and imaged using a 20× air objective lens on an inverted confocal microscope ( Nikon A1R on a Ti-E inverted microscope ) . The pinhole was opened to 244 . 43 μm to allow a thicker optical section to be monitored . UAS-GCaMP-R ( GCaMP and mCherry ) was expressed in Lk neurons and simultaneously excited with wavelengths of 488 nm ( FITC ) and 561 nm ( TRITC ) . Recording were taken for 120 seconds , capturing 1 frame/5 seconds with 512 × 512 resolution . For analysis , regions of interest ( ROIs ) were drawn manually , capturing the same area between experimental and control . The mean fluorescence intensity was subtracted from background mean fluorescence intensity for FITC and TRITC per frame . Then , the ratio of GCaMP6 . 0 to mCherry was calculated and plotted as an average of the total time recorded per brain imaged . In vivo imaging was performed using a previously described protocol with some modifications [89 , 90] . Briefly , fed , 24-hr–starved , or 3-hr re-fed ( standard Bloomington Recipe ) flies were anesthetized on ice and secured in a 200-μL pipette tip with head and proboscis accessible . The pipette tip was placed in a small chamber at an angle of 140° , allowing the dorsal and posterior surface of the brain to be imaged . A small hole was cut in the tin foil and fixed to the stage and fly head , leaving a window of cuticle exposed , then sealed using dental glue ( Tetric EvoFlow; Ivoclar Vivadent , Schaan , Lichtenstein ) . The proboscis was extended , and dental glue was used to secure it in place , ensuring the same position throughout the experiment . A 21-gauge 1 1/4 needle ( PrecisionGlide; Becton Dickinson , Franklin Lakes , NJ , USA ) was used to cut a window in the fly cuticle . A drop of artificial hemolymph was placed on the cuticle , and the connective tissue surrounding the brain was dissected . Flies were allowed to recover from the procedure for 30–45 minutes in a humidified box . Mounted flies were placed under a confocal microscope ( Nikon A1R on an upright microscope ) and imaged using a 20× water-dipping objective lens . The pinhole was opened to 244 μm to allow a thicker optical section to be monitored . The settings and data analysis were performed as described above . Female third-instar larvae expressing GFP in Lk neurons were selected and anesthetized in ethyl ether ( Thermo Fisher Scientific , E134-1 ) for 2–5 minutes . Larvae were placed dorsally on a microscope slide , and a coverslip was placed on the larvae . Ringer’s solution was applied onto the larvae below the coverslip . Larvae were imaged using a 25× water-dipping objective lens on a multiphoton microscope ( Nikon A1R ) containing a Chameleon Vision II Ti:Sapphire tunable laser . Excitation laser light of 870 nm was used . Images were acquired at 1 frame per second with a resolution of 512 × 512 pixels . For each neural ablation , a total of four frames were acquired . Two frames were captured prior to ablation for a duration of approximately 2 seconds , followed by ROI stimulation of 2–4 seconds and two frames after ablation . Following ablations , larvae were placed in vials containing food and allowed to grow . Sleep in food tubes and on agar was measured 5–7 days posteclosion in the DAMS . In order to verify which neurons were ablated after behavioral assay , flies were anesthetized on ice , and the central nervous system ( CNS ) was dissected . Fly CNS was fixed in 4% paraformaldehyde , 0 . 5% PBST for 30 minutes . Following fixation , samples were imaged in 2-μm sections with a Nikon A1R confocal microscope ( Nikon ) using a 20× oil immersion objective . Ablations that resulted in the formation of supernumerary neurons or deletions of two different subpopulations of Lk neurons were removed from analysis . The experimental data are presented as means ± SEM . Unless otherwise noted , a one-way or two-way analysis of variance ( ANOVA ) followed by Tukey’s post hoc test was used for comparisons between two or more genotypes and one treatment and two or more genotypes and two treatments . Unpaired t test was used for comparisons between two genotypes . All statistical analyses were performed using InStat software ( GraphPad Software 6 . 0 ) with a 95% confidence limit ( p < 0 . 05 ) .
|
Neural regulation of sleep and feeding are interconnected and are critical for survival . Many animals reduce their sleep in response to starvation , presumably to forage for food . Here , we find that in the fruit fly Drosophila melanogaster , the neuropeptide leucokinin is required for the modulation of starvation-dependent changes in sleep . Leucokinin is expressed in numerous populations of neurons within the two compartments of the central nervous system: the brain and the ventral nerve cord . Both genetic manipulation and laser-mediated microablation experiments identify a single pair of neurons expressing this neuropeptide in the brain as being required for metabolic regulation of sleep . These neurons become active during periods of starvation and modulate the function of insulin-producing cells that are critical modulators of both sleep and feeding . Supporting this notion , knockdown of the leucokinin receptor within the insulin-producing cells also disrupts metabolic regulation of sleep . Taken together , these findings identify a critical role for leucokinin signaling in the integration of sleep and feeding states .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"rna",
"interference",
"chemical",
"compounds",
"statistics",
"sleep",
"green",
"fluorescent",
"protein",
"carbohydrates",
"neuroscience",
"animals",
"organic",
"compounds",
"glucose",
"animal",
"models",
"physiological",
"processes",
"analysis",
"of",
"variance",
"drosophila",
"melanogaster",
"model",
"organisms",
"luminescent",
"proteins",
"mathematics",
"experimental",
"organism",
"systems",
"epigenetics",
"drosophila",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"rna",
"sequence",
"analysis",
"animal",
"cells",
"genetic",
"interference",
"animal",
"studies",
"proteins",
"mathematical",
"and",
"statistical",
"techniques",
"gene",
"expression",
"bioinformatics",
"chemistry",
"insects",
"arthropoda",
"biochemistry",
"rna",
"cellular",
"neuroscience",
"eukaryota",
"cell",
"biology",
"nucleic",
"acids",
"organic",
"chemistry",
"physiology",
"neurons",
"genetics",
"database",
"and",
"informatics",
"methods",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"monosaccharides",
"statistical",
"methods",
"organisms"
] |
2019
|
A single pair of leucokinin neurons are modulated by feeding state and regulate sleep–metabolism interactions
|
After obtaining certification of the absence of transmission of the Trypanosoma cruzi by Triatoma infestans in 2006 , other native species of protozoan vectors have been found in human dwellings within municipalities of the State of Paraná , Southern Brazil . However , the spatial distribution of T . cruzi vectors and how climatic and landscape combined variables explain the distribution are still poorly understood . The goal of this study was to predict the potential distribution of T . cruzi vectors as a proxy for Chagas disease transmission risk using Ecological Niche Models ( ENMs ) based on climatic and landscape variables . We hypothesize that ENM based on both climate and landscape variables are more powerful than climate-only or landscape-only models , and that this will be true independent of vector species . A total of 2 , 662 records of triatomines of five species were obtained by community-based entomological surveillance from 2007 to 2013 . The species with the highest number of specimens was Panstrongylus megistus ( 73%; n = 1 , 943 ) , followed by Panstrongylus geniculatus ( 15 . 4%; 411 ) , Rhodnius neglectus ( 6 . 0%; 159 ) , Triatoma sordida ( 4 . 5%; 119 ) and Rhodnius prolixus ( 1 . 1%; 30 ) . Of the total , 71 . 9% were captured at the intradomicile . T . cruzi infection was observed in 19 . 7% of the 2 , 472 examined insects . ENMs were generated based on selected climate and landscape variables with 1 km2 spatial resolution . Zonal statistics were used for classifying the municipalities as to the risk of occurrence of synanthropic triatomines . The integrated analysis of the climate and landscape suitability on triatomines geographical distribution was powerful on generating good predictive models . Moreover , this showed that some municipalities in the northwest , north and northeast of the Paraná state have a higher risk of T . cruzi vector transmission . This occurs because those regions present high climatic and landscape suitability values for occurrence of their vectors . The frequent invasion of houses by infected triatomines clearly indicates a greater risk of transmission of T . cruzi to the inhabitants . More public health attention should be given in the northern areas of the State of Paraná , which presents high climate and landscape suitabilities for the disease vectors . In conclusion , our results–through spatial analysis and predictive maps–showed to be effective in identifying areas of potential distribution and , consequently , in the definition of strategic areas and actions to prevent new cases of Chagas' disease , reinforcing the need for continuous and robust surveillance in these areas .
Chagas disease ( CD ) is an important zoonosis caused by the protozoan hemoflagellate Trypanosoma cruzi [1] and represents an important public health problem in Latin American countries where approximately 6–7 million people are currently infected by the parasite [2] . Between 2000 and 2013 , Brazil recorded 68 , 206 deaths per CD , an average of 4 , 872 deaths per year [3] . Human infection occurs mainly through vector transmission that is , contact with excreta of infected triatomines ( Hemiptera: Reduviidae ) with T . cruzi [4] . In 2006 , after the implementation of a vector control program , Brazil was certified as free from T . cruzi transmission by Triatoma infestans ( Klug 1834 ) . This species was considered the main vector of CD during the last century [5] . After the control of T . infestans , Panstrongylus megistus ( Burmeister 1835 ) is considered the main vector of T . cruzi in Brazil . However , within the Brazilian Amazon , ecological niche models ( ENMs ) projected as main vector species Rhodnius robustus ( Larrousse 1927 ) and R . pictipes ( Stal 1872 ) [6] . In the State of Paraná , located in the Southern Brazil , the first record of P . megistus naturally infected by T . cruzi occurred in 1917 , in the municipality of Jataí , located in the valley of the Tibagi river in the north of the state [7] . This species presents a wide geographical distribution , has high rates of infection by T . cruzi and , in addition , present a high ability to colonize artificial ecotopes [8] . In the State of São Paulo , near the State of Paraná , the increasing proximity of this species to human dwellings has been verified , which characterizes it as an important vector for T . cruzi transmission [9] . Rhodnius prolixus ( Stal 1859 ) is among the five main vectors of T . cruzi and is native to Colombia and Venezuela , where it is considered the main vector of the parasite [10] . Triatoma sordida ( Stal 1859 ) has been considered as emerging vector prevalent in peridomestic habitats , mainly associated with bird nests and frequently found to be infected by T . cruzi . Nevertheless , there is still no evidence of vector transmission of T . cruzi to humans by this species . Another factor that may favor the emergence of cases of human infection by T . cruzi comes from the increase of natural and semi-natural areas inhabited by human population . These dwellings can overlap areas where natural cycles of T . cruzi occur , causing the emergence of artificial "niches" in homes and the maintenance of synanthropic reservoirs , such as mammals that have adapted to live inside or in close proximity to human dwellings [11] . ENMs use knowledge of the current distribution of the species , based on the known occurrences , together with environmental variables and mathematical algorithms , to infer suitable localities for species occurrence . Therefore , ENMs have become a widely used approach for predicting the potential distribution of species [12 , 13] . Once ENMs infer potential species distribution , this technique can be used to predict geographic patterns of disease transmission risk [14] . Several studies have demonstrated that the spatial distribution of triatomines populations is heavily influenced by climate conditions and changes in landscape level [6 , 12 , 15 , 16] . Higher temperature and humidity are one of the main determinants of the occurrence of triatomines . Additionally , the drier climate favors the dispersal of these insects , while the destruction of natural vegetation , such as forests , can modify species dispersal and species survival and local population persistence . However , no study has jointly assessed the landscape or climactic effects on the spatial distribution of triatomines in Southern Brazil . Currently , ENMs approaches have been applied to understand aspects of T . cruzi transmission , since these techniques are able to predict species geographic potential distribution [14] . Other species of triatomines , considered native , have invaded areas previously occupied by T . infestans , demonstrating evidence of ecological niche overlap [17] . This is the case of the State of Paraná , an ancient endemic area for CD . From 1995 , T . sordida , P . megistus and Rhodnius neglectus ( Lent 1954 ) were found in the intra and peridomicile , i . e . inside and outside the human dwellings respectively , in the rural area of municipalities of the Northwest and Central regions of the state , with infection rates by T . cruzi of up to 43% [7 , 18] . Moreover , residual foci of T . infestans are still detected in others Brazilian states and their elimination requires continuous and intensive entomological surveillance [5 , 19] . A higher number of synanthropic triatomines has also been observed in areas with higher seasonal temperature variation . The expansion of human-inhabited areas , including cities , and the modification of natural areas for agricultural activities and construction of human habitations , may disturb sites where natural cycles of T . cruzi population growth occur . This can lead triatomines to invade artificial ecotypes ( like domiciles ) , exposing the population to these vectors , and also maintaining synanthropic reservoirs close to dwellings . This fact reinforces the effects of landscape modification on species dispersal and population persistence in no-native sites highlighting the need for an updated survey in order to direct vector prevention and control measures and , subsequently , reduce the risk of infestation of triatomines and human infection . The main goal of the present study was to analyze and spatially predict areas with environmental potential suitability for the presence of triatomines and , hence , the risk for the vector transmission of T . cruzi in the State of Paraná . For this , we used data from the period 2007 to 2013 in combination with ENM techniques based in a new approach that include climate and landscape variables . We hypothesize that both landscape and climate conditions influence the potential risk for vector transmission of T . cruzi . However , we expect that the combination of both group of variables are more powerful than climate-only or landscape-only models , and that this will be true independently of vector species .
We developed a cross-sectional , observational , ecological study using spatial analysis techniques , based on occurrence locations ( i . e . geographic coordinates ) of triatomines species in the State of Paraná , Brazil . It is important to point out that the transmission of CD depends on multiple factors for it to occur , in addition to the simple presence of triatomines infected with T . cruzi . Only in a few cases all the necessary conditions to transform a potential vector into an effective vector of CD are fulfilled . In addition to the population density of the vector and its susceptibility to T . cruzi , or to one of the parasite genetic lineages , the effective vector must attend to other conditions such as adaptation to human dwelling , high degree of anthropophilia , and short time between hematophagy and defecation . These factors are important in the contaminative vector transmission , but not for oral transmission , which is independent of some of these factors [6] . The State of Paraná is located in the Southern Region of Brazil in the Southern Plateau region ( Fig 1 ) , between 22°29'33'' and 26°42'59'' of latitude south , and 48°02'24'' and 54°37'38'' of longitude west , covering an area of ~200 thousand km2 . Currently the state is subdivided into 399 municipalities with an estimated population of more than 11 million people in the year 2016 [20] . Although it covers only 2 . 5% of the country area , Paraná has in its territory the majority of the main phytogeographic units that occur in the country . It presents 97 . 8% of its territorial area inserted in the Atlantic Forest biome [21] , where five large phytogeographic units stand out . ( 1 ) The Atlantic Ombrophylous Forest region is located in the eastern portion of the state that is bordered by the natural geographic barrier of the Serra do Mar . The west of this barrier , occupying the plateau portions of the state are the ( 2 ) Mixed Ombrophilous Forest with Araucaria [22] . In the northern and western regions of the state there is ( 3 ) Semideciduous Seasonal Forest ( seasonal forest ) , characterized by two periods of climatic influence ( rain and drought ) , partially losing its leaves , and harboring a rich biodiversity . Its original distribution occupied 37 . 3% of the state area and , nowadays , there are only remnants in 3 . 4% of it ( Fig 1 ) . ( 4 ) Areas of Estepe ( Campos ) , cover about 14% of the state's surface , located generally in the higher portions of the three plateaus of the State of Paraná , and ( 5 ) Cerrado , located in the north and northeast , occupying about 1% of its area [22] . The state presents favorable climatic conditions for the development of several types of forest vegetation , which is mainly determined by the uniformity in the rainfall distribution during the year and absence of a clearly defined dry season . Following the Köeppen classification [23] , the predominant climate of Paraná state is the Cfa , subtropical moist mesothermic , with average temperatures in the coldest month below 18°C and average temperature in the warmest months above 22°C , with hot summers and frosts infrequent , extending between the Paranapanema and Paraná rivers until reaching regions with altitudes between 600 and 800 m a . s . l . The highland and plateau regions present a humid subtropical climate ( Cfb ) with average temperature in the coldest month below 18°C , with fresh summers , and average temperature in the warmer month below 22°C . Already in the northwest region of the state , the climate is tropical modified by altitude ( Cfah ) . The occurrence data of the synanthropic triatomine species of the State of Paraná were obtained during the activities of entomologic surveillance of Chagas disease and provided by the State Department of Health of Paraná / Division of Vector-borne Diseases based on captures in domiciliary environments between 2007–2013 . The use of the term synanthropic refers to the species of triatomines that frequently invade human habitations or home annexes and eventually colonize these environments [24] . The triatomines were captured through active research conducted by health agents and with the participation of community members who reported the presence of a suspected insect in their houses to health workers . The technical staff was trained to visit the house , to capture and performed a full entomological evaluation . The insects were confirmed as triatomines at the species level and submitted to examination of intestinal contents to identify positivity for T . cruzi using optical microscopy , according to the Southern Cone Initiative protocol [25] . The entomological indicators recommended by Organización Panamericana de La Salud and considered in this study were: number of insects captured per municipality , catch site ( intra or peridomiciliary ) and rate of natural infection . As the occurrence data are derived of municipality health departments , all triatomine species occurrence are associated with a central geographic coordinate that refers to the municipality centroid . The ENM approach was used to achieve the proposed objectives . The data of the bioclimatic variables were obtained from WorldClim dataset v . 1 . 4 ( www . worldclim . org ) . The landscape variables were the same used by Jorge et al . ( 2013 ) [26 , 27]: vegetation cover ( % ) , structural connectivity ( in log ( ha ) *100 ) , functional connectivity to 200 m ( in log ( ha ) /100 ) , Euclidean Distance to the nearest road and Human density ( hereafter Anthropogenic Distance ) , and homogeneity from Earthenv ( http://www . earthenv . org/texture ) [28] . We choose functional connectivity to 200 m because we agree that it is an interesting value considering the movement on triatomines in the landscape All variables were used with spatial resolution of 1 km2 for Atlantic Forest delimitation and cropped to our study area , State of Paraná , entitled Political-Administrative Division of the State of Paraná in the year of 2010 [20] . These variables were used in models building because is known the effects of landscape metrics in species dispersal , species survival and population persistence [29] , factors that increase the needs of use of landscape metrics to predicting transmission risk of Chagas disease . Initially , we used all 19 bioclimatic variables and two relief variables ( S1 Table ) and five landscape variables ( S2 Table ) . A factorial analysis was performed , similar to the methods applied by Sobral-Souza et al . [28] , to select the climatic variables that were less correlated with each other and which explained the greater environmental variation of the studied area . The selected climatic variables for the construction of climate-based models were isothermality , temperature annual range , mean temperature of warmest quarter , precipitation of wettest quarter and precipitation of coldest quarter ( S1 Table ) . The landscape variables used are related of the effects of landscape fragmentation in the triatomine species dispersion , since triatomines are associated with human occupations . The type of vegetation cover of an area , the plant species , its status of preservation and diversity of species , may influence the occurrence and abundance of certain species of triatomine vectors . Thus , we used vegetation cover and functional connectivity as variables for the construction of landscape models ( S2 Table ) . To build ENMs , the values of each selected variables were extracted cell by cell using 1 km2 as cell-size resolution . Knowing that distinct triatomine species are listed as potential vectors , even differing as to its importance in the epidemiology of Chagas disease , all occurrence points gathered regardless of species or sampling date . Thus , the occurrence locations used in the ENMs are related to the five most frequent species in the State of Paraná during the study period ( Figs 2 and 3 ) . Different analytical methods are available to infer the geographical distribution of species . Based on idea that the combined use of different mathematical algorithms increase the accuracy of the prediction results by considering different tolerances in the potential distribution of the species [30 , 31] , four mathematical algorithms were used that belong to two classes of models: two algorithms that consider only presence , represented by the climatic envelope—Bioclim [32] and by the distance method—Domain Gower Distance [33]; and two algorithms of presence/background: Support Vector Machines ( SVM ) [34] and Maximum Entropy ( MaxEnt v . 3 . 3 . 3k ) [35] . To evaluate the generated models , the occurrence points were divided into two subsets , training and test , which contained 75% and 25% of the occurrence locations , respectively . Since these subsets are part of the same set of data ( points of occurrence ) , this process was repeated 10 times , using the k-folding ( k = 2 ) technique , as a way of decreasing the autocorrelation between the data [13] . Thus , 40 different predictions were generated ( 10 randomizations x 4 algorithms ) for climate-based models , and another 40 predictions for landscape models , separately . The threshold values of each model were calculated as a way of transforming them in binary maps , from the maximum specificity and sensitivity . This threshold was used because it maximizes the correctness of presences and absences and has shown to be more efficient in predicting occurrences based on presence-only models previously used [36] . After defining the thresholds , the ensemble forecasting technique [30 , 31] was used to obtain the final prediction map of the triatomine distribution . The maps were generated based on climate and landscape separately . We binarized the 10 maps belonging each algorithm ( replicates ) by their respective threshold values previously calculated and than summed the maps of the same algorithm and between the algorithms . Thus , the suitability values of each cell in the final maps varied from 0 to 40 , demonstrating the frequency with which each cell was predicted as occurrence for the triatomines . In order to evaluate each of the generated models , the values of TSS ( True Skill Statistic ) were estimated . The values of TSS vary from -1 to 1 , where negative values or close to 0 indicate that the models do not differ statistically from randomly generated models; values close to 1 indicate excellent models , and values above 0 . 5 are considered as adequate models [37] . The final distribution maps of triatomines , both climate-based and landscape-based , were of 1 km2 cell-size resolution . However , as the objective of this study is to understand the risk of transmission per municipalities , we used the ENMs in 1 km2 as input and suitability average per municipality using zonal statistics . Thus , we built two suitability maps per municipality: climate-based and landscape-based map . First , niche models based on climatic variables were constructed as a way of inferring the influence of the climate on distribution of triatomine species in the municipalities of the State of Paraná . Likewise , models based on landscape variables were also generated , as a way to predict the effect of the landscape on the distribution of species . We built ENMs separately because landscape and climate condition affects species occurrence in different scales ( narrow and broad-scale ecological processes ) . Finally , we generated a scatterplot with climate municipality suitability values ( axis X ) and landscape municipality suitability values ( axis Y ) to infer which municipalities in the State of Paraná have high climatic and landscape suitability for the occurrence of triatomines , municipalities with high climatic adequacy and low landscape suitability , municipalities with high landscape adequacy , but low climatic suitability , and municipalities with low climatic and landscape suitability ( Fig 4 ) . Here we proposed this new method to combine the effects of landscape and climate on species distribution , hereafter EcoLand Analysis . We used a 0 . 75 suitability threshold value for considering that a municipality is suitable or not for species occurrence ( see Fig 5 ) . Thus , the final maps of the distribution of triatomines weighs the effect sizes of climatic and landscape conditions separately . The State Environmental Surveillance Center ( CEVA ) of the State Department of Health of Paraná ( SESA ) ( Protocol 13 . 669 . 712–9 ) authorized data collection of notifications and localities of triatomine species .
During the study period , a total of 2 , 662 specimens of triatomines belonging to five species were captured in the State of Paraná . No distinction between developmental stages was available . The species with the highest number of occurrences was P . megistus ( 73 . 0%; n = 1 , 943 ) of the specimens , followed by P . geniculatus ( 15 . 4%; 411 ) , R . neglectus ( 6 . 0%; 159 ) , T . sordida ( 4 . 5%; 119 ) and R . prolixus ( 1 . 1%; 30 ) ( Table 1 ) . Of the total number of insects captured , 71 . 9% ( 1 , 914 ) were found within houses , with the species P . megistus corresponding to 54 . 8% ( 1 , 461 ) of the specimens . The rate of infection by T . cruzi was 19 . 7% in 2 , 472 triatomines examined and P . megistus had the highest infection rate ( 24 . 7% ) . The municipalities of Paraná with the highest number of captured triatomines were Guamiranga ( 6 . 3%; 196 specimens ) , Rosário do Ivaí ( 5 . 6%; 148 ) , Querência do Norte ( 4 . 4%; 118 ) , Santana do Itararé ( 4 . 3%; 115 ) and Nova Londrina ( 3 . 5%; 94 ) ( S3 Table ) . The triatomines were captured in 39 . 9% ( 159/399 ) of the municipalities and in 47 . 8% ( 76 ) of them , there were positive insects for T . cruzi , and in 13 . 2% ( 21 ) of the municipalities , the insect infection rate was 100% . Between 2007 and 2013 , the number of insects caught varied between 492 and 223 , with a peak of 664 specimens in the year 2009 . P . megistus prevailed in 6/7 years ( except for 2012 when P . geniculatus was the most captured ) , showing a decline , so much that in 2007 that species corresponded to 97 . 4% of the insects caught and in 2013 44 . 1% ( Fig 3 ) . All generated models had TSS values above than 0 . 5 ( S4 Table ) . The results of the predictions show that in the analysis of 1 km2 cell-size and in the zonal model ( which demonstrates the mean values for each municipality ) , the northern and central regions of the state , identified in red on the map , present the most suitable areas for the presence of triatomines species ( Fig 4A and 4B ) . The southern and eastern regions , highlighted in blue on the map , have areas with low potential for the presence of triatomines species . The most suitable climate for the occurrence of triatomines is the hot and humid and the optimum levels for most species are around 26–29°C and 70% relative humidity [38] . On the other hand , the main physiographic regions where occur four of the five triatomines species found in this study are tropical and subtropical , dry and xerophytic forests ( i . e . develop in areas with low humidity and is the main representatives of the Subtropical Forests of southern Brazil ) and , to a lesser extent , savannas [39] . The landscape-based predictions show a rather different distribution pattern , with all the northern extension and part of the western part of the state being considered suitable for triatomine occurrence ( Fig 4C and 4D ) . When comparing climate and landscape prediction maps , it can be observed that most municipalities appear in red , indicating areas with high suitability for the occurrence of triatomines species . However , some blue areas on the climate map were orange in the landscape map ( medium suitability ) . The EcoLand results , combining climate and landscape effects on triatomine occurrence indicated that the northern and northwestern regions are more suitable for the occurrence of triatomine species ( Fig 4E and 4F ) . The scatterplot ( climate suitability x landscape suitability ) results indicates that there are municipalities with a high suitability for the occurrence of triatomines , both for landscape and by climate conditions ( Fig 5 ) . However , it has been identified that many municipalities have either adequate climate or landscape and other locations have low suitability for both climate and landscape . It is noticeable that some municipalities in the northwestern , northern and northeastern of the state ( marked in red on the map ) have a greater risk of T . cruzi vector transmission due to the high climatic and landscape suitability for the occurrence of their vectors ( S1 Table ) . Besides these , a subset of others conditions under which the vector is able to survive and reproduce in must also be met . Of the 399 municipalities analyzed , 26% ( n = 104 ) were classified as having high climate and landscape suitability , and only 4% ( 16 ) had low climate and landscape suitability for the occurrence of the vectors .
During the seven years of study , five species of synanthropic triatomines were captured in close to 40% ( 159/399 ) of the municipalities of the State of Paraná . Most of the entomological data came from passive surveillance and 72% of captures occurred in the intradomicile . P . megistus , strongly associated with humid regions , was the most prevalent species , it had the highest rate of infection by T . cruzi ( 24 . 7% ) , about five times higher than the other species , and most of the specimens were found in the intradomiciliary ( 54 . 8% ) . This species therefore should receive more attention from the health authorities , adopting stricter entomological surveillance . Previous studies showed T . sordida as the species most frequently captured in Paraná , usually with high infection rates for T . cruzi in the peridomicile [7 , 18] . However , that was not the case in the present study , since P . megistus was the most frequently captured species . Moreover , T . infestans were not found , confirming the efficacy of the Southern Cone Initiative to Control Chagas Disease program in the state [40] . P . geniculatus was the second most captured species in the study period , overcoming P . megistus in 2012 in number of captured specimens—but presenting a low infection rate ( 4 . 7% ) . This is considered a wild species associated with armadillo holes [41] , however , it has also been observed to be associated with domestic pigs in the Amazon basin in Northern Brazil [42] . Attracted by light , they are frequently found inside human dwellings and , together with P . megistus , they occur in more than 20 Brazilian states and in at least three Brazilian biomes , showing large potential for adaptation to different ecological conditions [43] . The low frequency observed of the species R . neglectus and T . sordida may have been influenced by the type of vegetation cover of the studied area . Gurgel-Gonçalves et al . [6] used ENMs for generate suitability maps of these species in the Cerrado biome , unlike the State of Paraná that , as shown previously , has only 1% of Cerrado in its territorial area . All Rhodnius species have been primarily associated with palm trees even though some species were found in other sylvatic habitats . Although R . neglectus has a wider distribution across the Cerrado it has also been found in the adjacent regions of Central Brazil . In the decades after the implementation of the control program that led to the elimination of T . infestans , T . sordida was the most common species in Paraná , corresponding to 87 . 4% and 98 . 7% of the captured specimens , with rates of infection by T . cruzi ranging from 13 . 4% to 43 . 0% [7 , 18] . The areas of higher occurrence of T . sordida are the ones related to the agricultural activities in the past , what could explain its presence in areas of the state of Paraná that suffered ecologic impact due to significant loss of vegetation . However , the results of the current study show that their abundance in this state did not hold , occupying the 4th place in terms of frequency ( 4 . 7% ) with low infection rate ( 1 . 8% ) . Among the five triatomine species we analyzed , R . prolixus was the least frequent ( 1 . 1% of the captured specimens ) , showing a slight increase in the last two years . A low frequency of this species was also observed in a study of the spatial distribution of triatomines in the State of Minas Gerais , southeastern Brazil [44] . For being considered one of the main vectors of T . cruzi , it deserves attention from the sanitary authorities to prevent the expansion in its geographical range . Several studies on the distribution of triatomines in Brazil have considered the capacity of domiciliary or home invasion associated with climatic conditions [15] and biogeographic aspects related mainly to vegetation cover [16] . In this study , an ENM approach was used , which allows the exploration of geographic and ecological phenomena based on known occurrences of species [32] . From these techniques , risk areas can be defined by analyzing the geographical distribution of disease cases , vectors or reservoirs . In a study by Peterson et al . [45] it was possible to develop distribution hypotheses for 15 species of mammals and insects that interact in the potential transmission of T . cruzi in Mexico . The variables isothermality , temperature annual range , mean temperature of warmest quarter , precipitation of wettest quarter and precipitation of coldest quarter , were selected for ENMs because they are less correlated with each other and explain the greater climatic variation of Paraná state . However , the subtropical humid mesothermic climate ( Cfa ) , which predominates in the north and central regions of the state , is more suited to the triatomine encounter , differing from the climate with fresh summers and temperatures below 22 oC occurring in the mountains and plateau regions . Our climate-based models predicted that all selected climate variables affect triatomine species occurrence and indicate that northwest , north and northeast regions has suit climate and landscape condition to maintain triatomine population . The southern region of Paraná state has severe climate condition ( low temperature ) for triatomine occurrence . The influence of temperature on triatomine distribution has been reported in literature [15 , 46 , 47 , 48] . Higher temperature values favor greater geographical dispersion of wild vectors [38] . Moreover , the type of vegetation cover and the structural connectivity ( that is , the landscape's ability to facilitate biological flow ) also influenced the prediction of species occurrence suitability . The areas of seasonal forest and , to a lesser extent , Mixed Ombrophylous Forest , were more suitable for the occurrence of triatomines . These types of vegetation contain several species of our flora considered triatomines ecotypes , such as hollow trees in arboreal habitats , palms and epiphytes [6] . Besides the variables structural and functional connectivity may favor the flow not only of triatomine species , but also of mammalian and avian species that can act as a food source of these insects , as a source of infection ( wild mammalian of different orders , known reservoirs of T . cruzi ) . However , as we hypothesized , models using climate and landscape effects on triatomine distribution allowed us to clearly identify the municipalities that present greater suitability for the occurrence of synanthropic triatomines and , consequently , the risk for vector transmission of T . cruzi . These municipalities are located mainly in areas of mesothermal humid subtropical climate and in the phytogeographic unit of the Atlantic Forest biome of the seasonal forest . Main efforts should concentrate where both suitability values converge , in order to apply the best control measures . On places where there is medium of both suitability values , initiatives of prevention should be applied , in order to keep the population informed and prevented against CD . These results allow to predict the profile of municipalities on which there is a high potential risk of T . cruzi vector transmission , without the need for sampling the focal insects . These predictions generate ecological and biogeographic knowledge regarding triatomines providing information to subsidize health services , guiding the actions of control and prevention of CD in the State of Paraná . Although our results predict a reliable spatial potential risk of T . cruzi , there are some limitations . The use of pooling data from multiple vector species , and our assumption that gathering the vector species occurrence data is plausible ( whether infected or not , and whether found indoors or not ) should be reviewed to improve prediction of risk of T . cruzi . It is also important to differentiate the developmental stage of triatomines since the encounter of immature forms inside houses means colonization of these insects . However , to do it , we need others field expeditions to search new triatomine species in unknown and unstudied sites . With accurate spatial and temporal information about triatomine occurrence directing disease control and preventive actions will be more effective .
|
Trypanosoma cruzi , the etiologic agent of Chagas' disease ( CD ) , circulates among humans , vectors and domestic and wild reservoirs . Frequent invasion of housing by infected triatomines may present a risk of vector transmission from T . cruzi to humans . The objective of this study was to predict the potential distribution of triatomines in the State of Paraná and to infer potential high risk areas for T . cruzi vector transmission . Five distinct species of three different genera of triatomines were collected , the majority of which were Panstrongylus megistus ( 73% of occurrences ) . More than half of collected specimens ( 54 . 8% ) were captured intradomicile . The infection rate was 24 . 7% , which may represent a risk of occurrence of new cases of CD by vector transmission . Ecological Niche Models indicated suitable areas to triatomine occurrence in municipalities of Northwest , North and Northeast of Paraná state . Thus , more effective management actions are needed to ensure epidemiological surveillance of CD . P . megistus is the main species that colonizes the intradomicile , reinforcing the importance of the maintenance of surveillance for directing vector control actions of the CD in the local health systems in this and similar regions , mainly in the northern region of the State of Paraná , with the intensive participation of the community and also of municipal agents .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"ecology",
"and",
"environmental",
"sciences",
"medicine",
"and",
"health",
"sciences",
"atmospheric",
"science",
"population",
"dynamics",
"geographical",
"locations",
"tropical",
"diseases",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"animals",
"simulation",
"and",
"modeling",
"protozoans",
"neglected",
"tropical",
"diseases",
"population",
"biology",
"triatoma",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"south",
"america",
"ecosystems",
"protozoan",
"infections",
"brazil",
"insects",
"disease",
"vectors",
"arthropoda",
"people",
"and",
"places",
"trypanosoma",
"cruzi",
"trypanosoma",
"eukaryota",
"chagas",
"disease",
"climatology",
"ecology",
"forests",
"earth",
"sciences",
"climate",
"modeling",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"organisms",
"terrestrial",
"environments",
"geographic",
"distribution"
] |
2018
|
Spatial prediction of risk areas for vector transmission of Trypanosoma cruzi in the State of Paraná, southern Brazil
|
The leishmanioses , vector-borne diseases caused by the trypanosomatid protozoan Leishmania , are transmitted to susceptible mammals by infected phlebotomine sand flies that inoculate promastigotes into hemorrhagic pools created in host skin . We assumed that promastigotes are delivered to a blood pool , and analyzed early promastigote interactions ( 0–5 min ) with host components , which lead to parasite endocytosis by blood leukocytes , and to host infection . Promastigotes were incubated with NHS or with heparinized blood in near-physiological conditions , and we used cell radioimmunoassay and flow cytometry to measure the on-rate constants ( k+1 ) of promastigote interactions with natural opsonins and erythrocytes . We obtained quantitative data for parasitized cells to determine the time-course of promastigote binding and internalization by blood leukocytes . In these reactions , promastigotes bind natural opsonins , immune adhere to erythrocytes and activate complement cytolysis , which kills ∼95% of promastigotes by 2 min post-infection . C3-promastigote binding is a key step in opsonization; nascent C3-promastigotes are the substrate for two simultaneous reactions , C3-promastigote immune adherence ( IA ) to erythrocytes and complement-mediated promastigote killing . The k+1 for IA was 75-fold greater than that for promastigote killing , showing that IA facilitates promastigote endocytosis and circumvents lysis . At 5 min post-infection , when reaction velocity is still linear and promastigote concentration is not limiting , 17 . 4% of granulocytes and 10 . 7% of monocytes had bound promastigotes , of which ∼50% and ∼25% , respectively , carried surface-bound ( live ) or internalized ( live and dead ) leishmanias . Of other leukocyte types , 8 . 5% of B cells bound but did not internalize promastigotes , and T cells , NK cells and CD209+ dendritic cells did not bind parasites . These data show that , once in contact with blood , promastigote invasion of human leukocytes is an extremely rapid and efficient reaction , and suggest that the IA reaction constitutes a central strategy for this parasite in subverting host innate immune defenses .
The leishmanioses are a group of vector-borne zoonotic diseases caused by trypanosomatid parasites of the genus Leishmania . Leishmanias are heteroxenous protozoa with a life cycle in two different hosts , in the Psychodidae diptera of the genera Phlebotomus and Lutzomyia and in mammals . In the sand fly , Leishmania lives in the digestive tract as an extracellular motile flagellated promastigote; in mammals , it dwells as a sessile aflagellated amastigote inside macrophages [1] . Female phlebotomine sand flies are hematophagous arthropods that require blood proteins for oogenesis . Sand flies feed from hemorrhagic spots created in the host dermis . When feeding on a Leishmania-infected host , flies can ingest amastigotes or amastigote-laden macrophages and become infected . In the vector gut , amastigotes differentiate first into procyclic promastigotes and subsequently into more mature promastigote morphotypes . Promastigote differentiation generates mature , non-dividing parasites termed metacyclic promastigotes , considered the Leishmania forms that infect mammals [2] . In heavily infected flies , the lumen of the food canal appears choked by a promastigote-derived mucin-like gel ( PSG ) containing large numbers of promastigotes embedded in a filamentous proteophosphoglycan ( fPPG ) matrix [3] . During sand fly engorgement , PSG limits the food flow and jams the vector feeding system , hampering intake of an adequate blood meal; this prompts sand fly regurgitation and delivery of promastigotes together with saliva and fPPG to the intradermal pool , thus causing infection [4] . Leishmaniosis is transmitted to mammalian hosts when infected sand flies take a second blood meal [5] , [6] . Cases of Leishmania infection have been reported in humans with no apparent blood uptake by the vector; in this case , promastigotes are presumably deposited into the extracellular matrix in the dermis [7]–[9] . Parasite transmission without blood involvement is also described in experimental rodent infection , in which a considerable fraction of transmitting flies apparently did not ingest blood while feeding [10] , [11]; nevertheless , one of these studies shows that promastigote transmission was 2 . 6-fold higher among flies that had taken a second blood meal [10] . Studies of natural sand fly feeding habits showed that a majority ( 63–68% ) of females trapped around animal shelters were blood-fed , and that 58 . 7% of blood-fed flies were PCR-positive for Leishmania DNA , double the number of positives found in non-blood-fed flies [12] . These data indicate that blood uptake by the vector is frequent in leishmaniosis transmission . In natural Leishmania infection , promastigotes can therefore be delivered into hematomas or into a bloodless context in the skin . This is not an irrelevant issue , as promastigotes interact in blood with leukocyte populations , whereas in the dermal matrix they interact with fibroblasts , dermal dendritic cells ( DDC ) , mast cells , and macrophages [13] , [14] . The environment and cell target repertoire in which infection occurs can influence the course of disease development , as well as the type and intensity of immune response induced [15] , [16] . Recent work in mice explored parasite fate after intradermal promastigote inoculation [13] , [17] , but the physiological and functional differences between the innate immune systems of mouse and man preclude direct extrapolation of results . In humans , comprehensive studies of the initial stages of leishmaniosis transmission in the blood pool are lacking , and most infection studies have been carried out using isolated leukocyte populations [18]–[26] . For a previous study , we designed an ex vivo model of infection in human blood to analyze the early stages of promastigote-host interaction [27] . Opsonisation , binding and internalization of promastigotes by target leukocytes occur within minutes ( early infection ) ; ensuing reactions triggered by the cells that endocytose parasites take hours or days to develop or to reach full intensity . Using two Leishmania species with different tropism , L . amazonensis and L . donovani , we studied the kinetics of early Leishmania infection of human blood , and measured the rate constants ( k+1 ) of promastigote opsonization reactions and the kinetics of promastigote binding and internalization by blood leukocytes during the very early infection period ( 0–5 min ) . Based on these data , we propose a kinetic model of ex vivo human blood infection by Leishmania promastigotes .
This study was approved by the Ethics Committee ( Comité de Ética de la Investigación y Bienestar Animal ) of the Instituto de Salud Carlos III ( Ref CEI PI 12_2009 ) . All human participants were volunteers and gave written consent . Leishmania donovani Khartoum 1246 ( MHOM/SD/43/124 ) and Leishmania amazonensis Maria ( MHOM/Br/79/Maria ) isolates were cultured in RPMI 1640 complete medium as described [27] . Stationary phase parasites were harvested by centrifugation ( 1 , 500×g , 15 min , 20°C ) , washed twice in PBS pH 7 . 2 , and adjusted to the desired concentration . Promastigotes were labeled with [111In]-oxine as described [27] . For labeling with 5-chloromethylfluorescein diacetate ( CMFDA; Invitrogen , Carlsbad , CA ) , early stationary phase promastigotes in 1 ml culture medium ( ∼2×107 cells ) were incubated ( 15 min , 37°C ) with 3 µM CMFDA; after incubation , cells were washed in PBS by centrifugation ( 11 , 000×g , 1 min ) and adjusted to 107 CMFDA-labeled promastigotes/ml . Blood was drawn from healthy donors into preservative-free heparin ( 10 IU/ml ) , kept at room temperature ( 20°C ) , and used in the IA reaction within hours of extraction . Normal human serum ( NHS ) was collected from clotted blood ( 20°C , 30 min ) and serum aliquots stored in liquid nitrogen . Monoclonal antibodies ( mAb ) used were anti-CD15-PE ( clone HI98 ) , -CD14-allophycocyanin ( APC ) ( clone M5E2 ) , -CD3-PE-Cy5 ( clone 5HIT3a ) and -CD56-APC ( clone B159 ) ( all from BD Pharmingen , San Jose , CA ) , -CD19-APC ( clone HIB19; BioLegend , San Diego , CA ) and CD209-PE ( clone eB-h209; eBioscience , San Diego , CA ) . FITC-labeled rabbit anti-human μ chain was from Dako ( Glostrup , Denmark ) . mAb SIM 27–49 ( anti-human C3 α chain; IgG2b ) was Cy5-labeled ( GE Healthcare ) following manufacturer's instructions . Goat anti-μ ( 50 µg ) and SIM27–49 ( 25 µg ) were labeled with 5 µl sodium 125I ( carrier-free , 105 . 36 mCi/ml; DuPont/NEN Life Science ) in Iodogen ( Pierce , Rockford , IL ) -coated tubes and [125I]-SIM27–49 activity was measured [28] . The Leishmania opsonization pathway can be depicted as a sequence of four reactions: 1 ) promastigote+natural antibodies→promastigote-IgM , 2 ) +complement→promastigote-C3 ( nascent promastigote-C3 are the substrate of two subsequent competing reactions ) , 3 ) promastigote-C3+E→promastigote-E IA and 4 ) promastigote-C3+C5b-C9→promastigote propidium iodide ( PI ) uptake . In 25% NHS , the concentrations of anti-Leishmania IgM antibodies ( [IgM5] ) , complement components and receptors for complement C3b fragments on erythrocytes ( [E-CR1] ) are in moderate to great excess over the concentration of promastigote IgM binding sites ( [Pmbs] ) and bound C3b with which these molecules interact in reactions ( 1 ) to ( 4 ) . To simplify measurement of on-rate constants , reactions ( 1 ) to ( 4 ) were considered one-way bimolecular interactions of the type A reactant+B reactant→P ( product ) , and analyzed as described [29] , [30] . The rate of P formation is described asin which k+1 is the second-order on-rate constant for the reaction . When the concentration of reactant [B] is >> [A] , [B] does not change substantially and the reaction is said to be pseudo-first-order . The velocity equation is writtenwhere kapp , the apparent pseudo-first-order rate constant , is kapp = k+1 [B] . The rate expression simplifies to L ( Bmax/Bmax−Bt ) = kapp t , in which L is the natural logarithm , Bmax is the percent of maximum binding , and Bti , the percent of binding at times ti . Kinetic data plotted as L ( Bmax/Bmax−Bti ) against incubation time ( ti ) render a straight line of slope kapp . The second-order rate constant of the reaction is obtained from k+1 = kapp/[B] . Sigmoid kinetic data were fitted by non-linear regression using a four-parameter Hill equation ( SigmaPlot 9 . 0 ) . Aliquots ( 50 µl ) of L . donovani or L . amazonensis promastigotes ( 107/ml ) were mixed with 50 µl 50% PBS-diluted pooled NHS and incubated for varying time periods between 0 and 60 sec . The reaction was terminated by dilution with 1 ml cold ( 4°C ) PBS containing 2 . 5% FCS and 0 . 05% NaN3 ( PFS buffer ) , followed by centrifugation ( 11 , 000×g , 1 min ) . To avoid cell loss , untreated promastigotes ( 5×106 ) were then added to the pellet and the samples washed twice in cold PFS ( 11 , 000×g , 1 min ) to remove traces of serum IgM that could block 125I-goat anti-μ binding . Parasites were resuspended in 0 . 2 ml PFS containing 2×105 cpm 125I-goat anti-μ ( 107 cpm/µg ) , incubated 1 h on ice , washed twice as above , and promastigote-bound antibody was determined . Assay conditions were identical to those for promastigote-IgM binding , except that the reaction was terminated by dilution with 1 ml cold PFS containing 5×106 untreated promastigotes , and tube contents were washed twice by centrifugation ( 11 , 000×g , 1 min ) . Promastigote pellets were resuspended in 0 . 2 ml PFS containing 5×105 cpm [125I]-SIM27–49 ( 6×106 cpm/µg ) and incubated 1 h on ice . After incubation , samples were processed as above and promastigote-bound [125I]-SIM27–49 cpm determined . Aliquots ( 50 µl ) of [111In]-labeled promastigotes ( 107 cells/ml ) were mixed with 50 µl heparinized blood and incubated for varying time periods ( 0 to 60 sec ) . EDTA ( final concentration 5 mM ) was added to terminate the reaction , and samples were immediately fractionated by centrifugation ( 500×g , 3 min ) through 1 . 5 ml 72% Percoll . [111In]-labeled E-bound ( E pellet ) and free parasites ( Percoll solution ) were then filtered through glass fiber discs ( GF/C; Whatman ) , washed three times with cold PBS , and retained [111In] cpm determined . Single aliquots ( 50 µl ) of L . donovani or L . amazonensis promastigotes ( 107/ml ) were mixed with 50 µl 50% PBS-diluted pooled NHS and incubated for varying times ( 0 to 140 sec ) . The reaction was terminated by diluting the sample with 1 ml FACSFlow sheath fluid ( BD Biosciences , San José , CA ) containing 5 µg/ml propidium iodide ( PI; Sigma-Aldrich , St . Louis , MO ) and PI uptake by killed promastigotes was measured in a FACSCalibur flow cytometer ( BD Biosciences ) [28] . To estimate the number of live parasites after opsonization in different NHS concentrations , single aliquots ( 0 . 3 ml ) containing L . amazonensis promastigotes ( 2×105 ) , 10 µg/ml PI , and serially diluted ( 50% to 0 . 78% ) NHS were incubated ( 37°C ) for 1 to 9 min . Leishmania killing was determined by measuring promastigote PI uptake in real-time by flow cytometry ( FACSCalibur ) . The precise percentage of complement-killed promastigotes cannot be determined by flow cytometry due to background signal from small particles and debris in stationary cultures; to reduce background , promastigotes were gated ( SSC vs . FSC ) and PI uptake emission measured in a dot plot of FL-2 ( 585/42 nm ) vs . time ( sec ) ; data were analyzed with CELLQuest software ( Becton Dickinson ) . To measure promastigote complement-resistance in 50% NHS , aliquots ( 50 µl ) of promastigotes ( 2×106/ml ) of six L . amazonensis stationary cultures ( 96 . 6% viable ) were incubated in 50% pooled NHS ( 37°C , 5 min ) , after which promastigote PI uptake was measured by flow cytometry . Additional experiments ( n = 3 ) compared PI uptake by stationary promastigotes and promastigotes enriched in metacyclic forms by centrifugation in a 10% to 30% Ficoll density gradient ( “top 10” promastigotes ) [31] , [32] . Briefly , 1 ml of stationary-phase promastigotes ( 2×108/ml ) was layered onto a discontinuous gradient of 2 ml 10% and 2 ml 30% Ficoll solutions , and centrifuged ( 1300×g or 365×g , no brake; 22°C , 10 min ) . Promastigotes at the 10% Ficoll interface and the upper part of the 10% Ficoll cushion were pooled , diluted with RPMI 1640 complete medium and washed by centrifugation ( 1500×g , 15 min ) . The pellet , resuspended in complete RPMI 1640 , was washed again and adjusted to 2×107cells/ml . Aliquots ( 50 µl ) of stationary-phase and “top 10” promastigotes were incubated ( 37°C , 5 min ) in 50% NHS; for controls , NHS was replaced by PBS . The reaction was terminated by dilution with complete RPMI 1640 and centrifugation . Promastigote PI incorporation was determined by flow cytometry after incubating a 10 µl aliquot of each sample in 0 . 2 ml of PBS containing 2 µl PI ( 1 mg/ml ) . Promastigotes were gated ( SSC vs . FSC ) to eliminate small particles and debris , and PI uptake emission was collected in the FL2 detector through a 585/42 nm band pass filter . After opsonization in different NHS concentrations , the percentage of live promastigotes was quantitated by light microscopy . Aliquots ( 50 µl ) of stationary-phase L . amazonensis promastigotes were incubated ( 37°C , 5 min ) in serially diluted NHS ( 1∶2; 50% to 0 . 78% concentration ) . The reaction was terminated by dilution with PBS , and live parasites counted under a microscope . It should be noted that after parasite incubation in ≥3% serum , motile promastigotes with a slender shape are observed very infrequently . Most parasites registered as viable undergo marked changes in cell geometry , lose the flagellum and become ellipsoidal or small round refractile bodies with limited but noticeable drifting; such changes complicate identification . To measure initial blood leukocyte binding of opsonized leishmanias , we incubated CMFDA-labeled promastigotes with heparin-treated blood and used flow cytometry to determine the percentage of each leukocyte subpopulation that bound parasites after 5 min . We mixed 1 ml CMFDA-labeled L . amazonensis or L . donovani promastigotes ( 107/ml ) with 1 ml heparinized blood , followed by incubation ( 37°C , 5 min , waterbath ) . The mixture was divided into 200 µl aliquots and E were lysed by incubation with 2 ml E-lysing reagent ( EasyLyse , DakoCytomation; 10 min ) . Cells were washed in PBS by centrifugation ( 500×g , 10 min , 20°C ) . Each pellet was incubated ( 30 min , 20°C ) with one of the following mAb: anti-CD14-APC , -CD15-PE , -CD3-PE-Cy5 , -CD56-APC , -CD19-APC or -CD209-PE . Cells were washed in PBS by centrifugation ( 500×g , 10 min ) . Leukocyte subpopulations were identified by flow cytometry and represented in a dot plot as side scatter ( SSC ) vs . the specific fluorescent label of each population . The percentage of each subpopulation that bound promastigotes was analyzed in a secondary plot by independently representing the fluorescence intensity of each gated mAb-labeled population vs . that of cell-bound CMFDA-labeled promastigotes ( FL1 , green 530 nm ) following excitation with a 488 nm argon ion laser . To account for nonspecific binding , we subtracted promastigote binding at time zero from all values registered . In heat-inactivated serum ( HIS ) , promastigote binding by leukocytes was negligible at the times tested ( Fig . S1 ) . Use of HIS as control would require cell separation from plasma and its replacement with HIS , a less physiological approximation of infection conditions . The number of events acquired by the cytometer varied from 5×104 to 2×105 . A total of 5×103 events were counted for each leukocyte population , except in the case of CD209+ cells , for which 2×105 leukocytes were acquired and 150 events counted . To measure the kinetics of promastigote binding and internalization by granulocytes and monocytes , 50 µl aliquots of heparinized human blood were mixed with 50 µl CMFDA-labeled Leishmania promastigotes ( 107/ml ) and incubated ( 37°C ) for various times ( 0 , 0 . 5 , 1 , 1 . 5 , 2 , 2 . 5 , 3 , 4 , 5 , 10 , 30 and 60 min ) . The reaction was terminated by addition of 2 ml E-lysing reagent . After 10 min incubation , 3 ml of sheath fluid were added to each tube and centrifuged ( 500×g , 5 min ) ; the pellet was washed with 5 ml sheath fluid as above and resuspended in 200 µl sheath fluid . Samples were divided in two 100 µl aliquots . To quench extracellular fluorescence , 10 µl trypan blue ( TB ) solution ( 10 mg/ml; final concentration 1 mg/ml ) were added to one sample . To register the percentage of leukocytes that bound and internalized promastigotes , granulocytes and monocytes were gated by SSC vs . forward scatter ( FSC ) and plotted independently in a secondary plot of SSC vs . FL1 ( green , 530 nm ) following excitation with a 488 nm argon ion laser . Quenched samples were acquired for 10 min after TB addition . A total of 2×104 events were acquired for each measurement and analyzed with CELLQuest software ( BD Biosciences ) .
Non-immune serum from most vertebrates contains natural anti-trypanosomatid antibodies that act as an innate recognition system in the host [28] , [39] . To determine the velocity of IgM binding to Leishmania parasites , we measured the association rate constant ( k+1 ) by incubating promastigotes in NHS for varying times . The reaction was terminated by sample dilution with PFS , and promastigote-bound IgM was measured using 125I-goat anti-μ . Pentameric IgM ( IgM5 ) binding reached maximum after 20 sec incubation for both L . donovani ( Fig . 1A ) and L . amazonensis ( Fig . 1B ) . [IgM5] in adult NHS is ∼1 . 3 g/L; if the Mr of IgM5 is considered to be 950 , 000 , then serum [IgM5] is ∼1 . 4×10−6 M . Exhaustive adsorption of 25% NHS with L . amazonensis and L . donovani promastigotes removed ∼15% and ∼30% of IgM5 , respectively [28] , indicating that [IgM5] anti-L . amazonensis in unadsorbed serum is ∼0 . 5×10−7 M and that of anti-L . donovani , ∼1×10−7 M . At equilibrium , L . amazonensis and L . donovani promastigotes bind ∼5 , 000 and ∼2 , 500 IgM molecules/cell , respectively . Early in the promastigote-IgM5 interaction , antibody binding is assumed to be monovalent [29] , which would render [Pmbs] on L . amazonensis promastigotes of ∼4 . 1×10−11 M and on L . donovani promastigotes of ∼2 . 1×10−11 M . In a monovalent promastigote-IgM5 interaction , the [IgM5]/[Pmbs] ratio is ∼1 , 300∶1 in L . amazonensis and ∼5 , 000∶1 in L . donovani; in the case of a decavalent promastigote-IgM5 interaction , this ratio would be ∼130∶1 ( L . amazonensis ) and ∼500∶1 ( L . donovani ) . In both cases , the promastigote-IgM5 interaction obeys pseudo-first-order kinetics . The percentage of IgM5 bound , plotted as L ( IgM5max/IgM5max−IgM5ti ) against reaction time ( ti ) , gives straight lines whose slopes are the kapp values of the reactions , ∼0 . 23 sec−1 for L . amazonensis and ∼0 . 18 sec−1 for L . donovani . Second-order rate constants were obtained from k+1 = kapp/[IgM5] , with values of ∼4 . 4×106 M−1 sec−1±0 . 05×106 M−1 sec−1 for L . amazonensis and ∼1 . 8×106 M−1 sec−1±0 . 05×106 M−1 sec−1 for L . donovani . The mean k+1 for natural IgM5 binding to promastigotes of both Leishmania species was ∼3×106 M−1 sec−1 . Early in Leishmania infection , human complement has effects both advantageous for and harmful to the promastigote . Complement activation triggers binding of C3 fragments to the parasite , which mediates IA and Leishmania internalization by host leukocytes; however , promastigote-C3 binding also nucleates assembly of the C5 convertases , activating the complement lytic cascade that kills the parasites . To establish how Leishmania circumvents complement effector activity , the kinetics and velocity of complement activation must be understood . We incubated promastigotes with NHS and measured the reaction time course and k+1 of promastigote-C3 binding . Promastigote-C3 binding follows a sigmoidal course , with an initial ∼30-sec lag during which 10 to 20% of total C3 ligands are fixed; C3 molecules that bind during this lag period probably attach to IgM . This amount of C3 is insufficient to activate the alternative pathway and the lytic cascade , but is sufficient to establish multipoint contacts with clustered E-CR1 , triggering the IA reaction [40]–[42] . Thereafter , C3 binds at an exponential rate for ∼1 min , during which C3 is bound at an average rate of ∼1 , 800 molecules/sec for L . donovani ( Fig . 2A ) and ∼1 , 200 molecules/sec for L . amazonensis ( Fig . 2B ) . During the C3 lag period , IgM5 binding to Leishmania reaches maximum and the [IgM5] bound is ∼4 . 1×10−11 M ( L . amazonensis ) and ∼2 . 1×10−11 M ( L . donovani ) . In 25% NHS , the concentration of complement classical pathway components ( [C1–C3] ) ranges from 4 . 5×10−8 M for C1 to 1 . 75×10−6 M for C3 . If the C1 interaction with promastigote-bound IgM5 is considered the rate-limiting step in classical complement pathway activation , the serum C1 concentration is ∼2000-fold greater than that of promastigote-bound IgM5 for L . amazonensis and ∼1000-fold for L . donovani . [C1–C3] is >> [promastigote-IgM5] , and C3 binding to promastigotes would proceed under pseudo-first order conditions . We analyzed the late phase of the exponential course , fitting the data to a pseudo-first order rate equation . Plots of the percentage of C3 bound as the L ( C3max/C3max−C3ti ) against time ( ti ) yielded kapp of ∼0 . 063 sec−1 for L . amazonesis and ∼0 . 061 sec−1 for L . donovani . The second-order rate constants were ∼3 . 6×104 M−1 sec−1±0 . 01×104 M−1 sec−1 for L . amazonesis and ∼3 . 5×104 M−1 sec−1±0 . 005×104 M−1 sec−1 for L . donovani; the mean k+1 value for Leishmania promastigote-C3 opsonization was ∼3 . 5×104 M−1 sec−1 . In human blood , C3-opsonized promastigotes immune adhere to erythrocytes [27] . IA is a C3-mediated innate immune mechanism , almost ubiquitous in mammals , that enhances phagocytosis and clearance of opsonized microorganisms from blood [43] . To study the velocity of IA in the reaction sequence of Leishmania infection , we followed the kinetics of IA complex formation and determined the on-rate constant of the interaction between promastigote-C3 and erythrocytes . Within seconds of serum contact , nascent C3-opsonized promastigotes bind to CR1 on E and form promastigote-E IA complexes . This interaction has an initial lag time of ∼15 sec followed by a period in which IA complexes are formed at an exponential rate until the reaction reaches completion at 30 to 40 sec of incubation . The IA reaction is so rapid that it gives the impression that promastigote-E binding proceeds before C3 opsonization ( Fig . 3A , B ) . Considering the average number of CR1 molecules per erythrocyte to be 500 [44] , the [E-CR1] in this assay is ∼2 . 1×10−9 M . At equilibrium , L . donovani promastigotes bind ∼180 , 000 C3 molecules/cell [28]; at the onset of the IA reaction in L . donovani ( ∼12 sec ) , there are ∼27 , 000 C3 molecules ( ∼15% of maximum binding ) bound to promastigotes , indicating a promastigote-bound C3 concentration ( [Pm-C3] ) of ∼2 . 2×10−10 M . At the onset of the IA reaction , the [E-CR1]:[Pm-C3] ratio is thus ∼10 . In the case of L . amazonensis , at equilibrium promastigotes bind ∼120 , 000 C3 molecules/cell , and at the onset of the IA reaction ( ∼15 sec ) there are ∼22 , 000 C3 molecules bound/cell ( ∼18% of maximum binding ) ; this yields a [Pm-C3] of ∼1 . 8×10−10 M , and a [E-CR1]:[Pm-C3] ratio of ∼12 . In these assays , the [E-CR1] exceeds the [Pm-C3] by ∼10-fold and the reaction proceeds under pseudo-first-order conditions . Regression analysis of the percentage of IA plotted as L ( IAmax/IAmax−IAti ) against incubation time ( ti ) yielded kapp values of ∼0 . 15 sec−1 ( L . amazonensis ) and ∼0 . 21 sec−1 ( L . donovani ) , and k+1 constants of ∼0 . 8×108 M−1 sec−1±0 . 02×108 M−1 sec−1 ( L . amazonensis ) and ∼1×108 M−1 sec−1±0 . 04×108 M−1 sec−1 ( L . donovani ) . The average k+1 for the Leishmania IA reaction was thus ∼9×107 M−1 sec−1±0 . 4×107 M−1 s−1 . Complement activation leads to assembly of the C5 convertases ( C4b3b2a , C3b2Bb ) , triggering the cytolytic complex ( C5b–C9 ) that causes promastigote death . We measured the time course of parasite membrane damage by complement as PI uptake and determined the on-rate constant of the reaction . The kinetics of promastigote PI uptake is very rapid; it begins at ∼50 sec after complement activation and by ∼100 sec after serum contact , most leishmanias have incorporated PI ( Fig . 4A , B ) . During that period , the [Pm-C3] varies from ∼6×10−10 to ∼1×10−9 M in L . amazonensis and from ∼5×10−10 to ∼1 . 5×10−9 M in L . donovani . In 25% human serum , the concentration of C5b–C9 components ranges from 9 . 2×10−8 M for C5 to 2 . 1×10−7 M for C9 . C5b deposition is considered the rate-limiting step; at the onset of the lytic reaction , [C5b–C9] is greater than [promastigote-C3] by ∼150-fold ( L . amazonensis ) and ∼180-fold ( L . donovani ) . The rate of complement-mediated PI uptake by promastigotes was calculated by plotting the percentage of PI incorporation as L ( PImax/PImax−PIti ) against incubation time ( ti ) . kapp values for L . amazonensis and L . donovani PI uptake were 0 . 13 sec−1 and 0 . 083 sec−1 , respectively , and the k+1 constants were ∼1 . 4×106 M−1 sec−1±0 . 008×106 M−1 sec−1 ( L . amazonensis ) and ∼0 . 9×106 M−1 sec−1±0 . 005×106 M−1 sec−1 ( L . donovani ) . The average k+1 value for promastigote PI uptake was ∼1 . 2×106 M−1 sec−1 . The on-rate constant values of opsonization , immune adherence , and PI uptake reactions of L . donovani and L . amazonensis promastigotes in human blood are summarized in Table 1 . To measure initial blood leukocyte binding of opsonized leishmanias , we incubated CMFDA-labeled promastigotes with heparin-treated blood and used flow cytometry to determine the percentage of each leukocyte subpopulation that bound parasites after 5 min . Leukocyte subpopulations were identified with fluorochrome-labeled anti-CD15 ( for granulocytes ) , -CD14 ( monocytes ) , -CD3 ( T cells ) , -CD19 ( B cells ) , -CD56+ ( NK cells ) and -CD209 ( monocyte/dendritic cells ) mAb . Data from a representative experiment are shown ( Fig . 5 ) in which fluorescence intensity of each gated leukocyte subpopulation is represented in a secondary plot against that of cell-bound CMFDA-labeled promastigotes ( FL-1 ) as the percentage of each subpopulation that bound L . amazonensis ( Fig . 5A ) or L . donovani ( Fig . 5B ) parasites . Cells of each subpopulation that bound promastigotes are expressed as a percentage of total leukocytes in the sample ( Fig . 5C; mean for eight experiments ) . After 5 min incubation , 13% of leukocytes bound promastigotes , of which 10 . 7%±0 . 15% were CD14+ , 76 . 3%±0 . 8% CD15+ , 2 . 7%±0 . 1% CD3+ , 8 . 5%±0 . 2% CD19+ , 1 . 3%±0 . 02% CD3− CD56+ and 0 . 49%±0 . 03% CD209+ cells . A substantial fraction of B cells ( 8 . 5% ) bound promastigotes in this early period . Granulocytes are the main subpopulation that bound promastigotes; nevertheless , the percentage of promastigote-binding cells in each subpopulation was nearly identical for CD15+ ( 19 . 1±2 . 4% ) , CD14+ ( 17 . 3%±2 . 3% ) and CD19+ ( 17 . 4%±2 . 5% ) ( Fig . S6 ) . Other leukocyte subpopulations did not bind promastigotes appreciably . We infected heparin-treated blood with CMFDA-labeled L . amazonensis or L . donovani promastigotes . At various times post-infection , granulocytes and monocytes were gated by SSC vs . FSC , and the percentage of each subpopulation that bound parasites was calculated in a secondary dot plot by representing SSC-gated cells against fluorescence intensity of granulocyte- and monocyte-bound CMFDA-labeled promastigotes ( FL1 ) . We simultaneously determined the percentage of granulocytes and monocytes that internalized promastigotes by quenching fluorescence emitted by cell-bound complement-killed CMFDA-labeled promastigotes , using trypan blue [45] . The time course of granulocyte and monocyte binding of promastigotes was linear in the first 5 to 10 min of infection . Promastigote concentration subsequently became limiting and the kinetics followed a hyperbolic course; the binding reaction was complete by 60 min ( Fig . 6 ) . After 30 min incubation , 39 . 2% of granulocytes ( ∼94% of total binding ) had already bound L . amazonensis and 45 . 1% ( ∼90% of total binding ) had bound L . donovani promastigotes . At this time , 23 . 5% of monocytes ( ∼76% of total binding ) had bound L . amazonensis and 20 . 2% ( ∼75% of total binding ) , L . donovani promastigotes . We analyzed leukocyte binding and promastigote internalization in the period between 0 and 5 min , when the velocity of promastigote binding is proportional with time and the cytolytic activity of complement on Leishmania is probably not yet complete ( Fig . 6A , B; insets ) . After 3 min incubation , the ratio of cells that bound∶internalized promastigotes was 12 . 9∶5 . 4% for granulocytes and 8 . 9∶2% for monocytes ( mean value of L . amazonensis and L . donovani ) ; after 5 min , these values were 17 . 4∶8 . 4% for granulocytes and 10 . 7∶2 . 6% for monocytes .
Phlebotomine sand flies transmit promastigotes to mammalian hosts in two ways , by direct inoculation of parasites into a blood pool in the skin or by their delivery into the interstitial space of the dermis [4] , [14] . Here we used human blood as a surrogate model of skin hematoma for a comprehensive kinetic analysis of promastigote interactions with host blood components in the first minutes of infection . As the time course of L . amazonensis and L . donovani promastigote reactions were similar , we compiled only the data for L . amazonensis ( Fig . 7 ) . Immediately after promastigote interaction with human blood , natural anti-trypanosomatid IgM/IgG antibodies bound the parasite with a hyperbolic course characteristic of first-order or pseudo-first order reactions; in contrast , the kinetics of promastigote-C3 deposition , promastigote IA , and promastigote PI uptake showed S-shaped curves , indicating cooperative ( IA ) or multimolecular ( C3 deposition , PI uptake ) mechanisms . Once opsonization was triggered , all reactions proceeded simultaneously , and by ∼2 min most parasites had been killed by complement activity . In human blood , Leishmania survival probably depends on the relative rate by which newly C3-opsonized promastigotes are incorporated into IA and PI uptake reactions . The 75-fold greater velocity constant of the promastigote IA ( k+1 , Pm IA∼9×107 M−1 sec−1 ) than that of promastigote PI uptake ( k+1 , Pm PI∼1 . 2×106 M−1 sec−1 ) permitted the IA reaction to be 50% complete by ∼30 sec before parasite PI uptake began . PI incorporation into promastigotes started ∼50 sec after serum contact , and required an additional 20 sec to reach 90% maximum; due to these differences in reaction times , complement killing begins at ∼70 sec after serum contact . In this interval , Leishmania IA can ferry promastigotes to blood leukocytes . Leukocyte populations that express receptors for opsonic ligands , principally C3 complement fragments , bound promastigotes in proportion to the concentration of each subpopulation in the blood pool . In human blood , granulocyte concentration exceeds that of monocytes , B cells and NK cells by 6- to 8-fold , and that of natural killer ( NK ) and CD209+ dendritic cells ( DC-SIGN ) by two to three orders of magnitude; the latter constitute a very small cell population ( 0 . 01–0 . 05% ) that binds diverse microorganisms in vitro , including promastigotes and axenic amastigotes of Leishmania [46]–[48] . At 5 min post-infection , 13% of blood leukocytes bound promastigotes in the proportion granulocytes ( 76 . 3% ) >monocytes ( 10 . 7% ) >B lymphocytes ( 8 . 5 ) >CD3+ cells ( 2 . 7% ) >CD3−CD56+ cells ( 1 . 2% ) cells>CD-209+ cells ( 0 . 5% ) ( %Fig . 5C ) . This shows that CD3+ , NK and DC-SIGN+ cells have no relevant role in early blood infection . In addition to granulocytes and monocytes , ∼8 . 5% of B lymphocytes bound promastigotes; to our knowledge , this interaction has not been previously reported . B cell binding of opsonized promastigotes was confirmed with Raji lymphoblastoid B cells , which bound C3-opsonized promastigotes to the same extent as U937 monocytes ( unpublished data ) . Promastigote-C3 binding by B cells is probably mediated by the IA reaction . In autologous serum and in PBS/RPMI 1640 medium , human erythrocytes transfer CR1-bound C3-opsonized promastigotes or antigen-antibody immune complexes ( IC ) to granulocytes , monocytes , B lymphocytes , and U937 cells [27] , [49]–[51] . Binding of C3-opsonized promastigotes allows B lymphocytes to present antigens to monocytes and macrophages through a CR2-mediated reaction similar to the transfer of E-bound C3-IC to B cells [52] . B lymphocytes have a still-undefined role in host immune response to Leishmania . Early data showed that BALB/c mice depleted of B cells by anti-IgM treatment had enhanced resistance to L . major infection [53] and subsequent studies indicated that B lymphocytes are involved in mouse susceptibility to Leishmania infection and disease pathogenesis [54]–[57] . Our results confirm that B lymphocytes have an early role in Leishmania immunity . In the ex vivo Leishmania blood infection experiments , promastigote inoculum was 5×103 parasites/0 . 5 µl blood , a 1 . 8∶1 promastigote∶leukocyte ratio . Assuming that the number of live opsonized promastigotes in the inoculum is 5% , 0 . 5 µl of host blood would contain ∼250 live parasites . In the first minutes after infection , granulocytes and monocytes bound parasites at a constant rate , and promastigote binding and ingestion was detectable from 3 min incubation ( Fig . 6 , insets ) . At 5 min post-infection , 17 . 4% of granulocytes bound promastigotes and 8 . 4% carried TB-unquenched parasites; for monocytes these figures were 10 . 7% and 2 . 6% , respectively . At that time , 8 . 5% of B cells bound promastigotes , but did not internalize them ( unpublished data ) . To illustrate the number of Leishmania-infected leukocytes , after 5 min incubation a 0 . 5 µl blood pool with 5 . 6×106 leukocytes/ml ( 66 . 9% granulocytes , 6 . 9% monocytes ) would have ∼330 granulocytes with bound promastigotes , ∼160 of which would have surface-bound ( live ) and internalized ( live and dead ) parasites . In the case of monocytes , there would be ∼20 cells with surface-bound promastigotes and ∼5 cells with surface-bound and internalized parasites . The early period of Leishmania infection was recently addressed in the mouse using intravital two-photon microscopy . Ng et al . highlight the role of DDC in invasion , and showed that between 55 and 70% of DDC had ingested parasites by 2 to 3 h post-infection [13] . Host neutrophil depletion before infection did not affect the number of DDC that internalized promastigotes , suggesting that these cells act independently of neutrophils . This apparent lack of neutrophil involvement in infection control could be due to the method of parasite inoculation or to the high intradermal promastigote∶DDC ratio used in these experiments . In any case , the intradermal cell compartment is very different from the hematoma environment; this experiment probably mimics promastigote transmission in a bloodless context , and it is not appropriate to compare it with the blood pool . In another study , Peters et al . transmitted infection by intradermal injection of large numbers of promastigotes or with infected flies fed ad libitum on restrained mice [17] . Both inoculation methods are likely to have caused dermic hematomas and to have induced substantial neutrophil infiltrate at the injection site . One day post-infection , ∼90% of infected neutrophils harbored live promastigotes; after a week , all leishmanias were inside macrophages and DDC . This change in host cells could be explained by the neutrophil Trojan horse model , which proposes that macrophages are infected when they dispose of apoptotic leukocytes [58] . Peters et al . [17] and others [59] nonetheless consider that when infected neutrophils become apoptotic , they release live promastigotes or amastigotes that are phagocytosed by macrophages and DDC ( “Trojan rabbit strategy” ) . To compare early Leishmania infection reactions in mouse and man , we must consider that the earliest host-protective mechanisms , serum complement and professional blood phagocytes , differ in their anti-promastigote activity and in cell proportion in these species . In human blood , the granulocyte∶monocyte ratio is 6–7∶1 , whereas in mouse it is from 2–3∶1 [60] . This is important , as these cells are the main parasite targets during infection and thereafter , when they are recruited to the inoculation site . Persistence of neutrophils harboring promastigotes is considered of paramount importance for subsequent disease development [24] , [61]–[63] . We anticipate that the size of such a neutrophil reservoir would differ between mouse and man . Human complement is highly cytotoxic to promastigotes , whereas mouse complement is not [28] , [64] . After inoculation , we estimate the number of live promastigotes in human blood to be one twentieth of that in mouse , which would considerably reduce the parasite load of infected neutrophils . Based on these data , we outline a kinetic model of Leishmania infection in human blood that incorporates rate constants for promastigote interactions , which measure the speed of these reactions ( Fig . 8 ) . At 50 sec post-Leishmania inoculation , promastigote opsonization reactions are terminated and complement-dependent parasite killing has not begun , indicating that rapid leukocyte uptake of promastigotes promotes Leishmania survival in the host [23] , [27] . We suggest that a crucial step in the infection pathway is determined by the high on-rate constant of the Leishmania IA reaction . This mechanism competes with complement-mediated parasite lysis by promoting promastigote binding and internalization by granulocytes and monocytes , and binding by B cells ( Fig . 8 ) . The substantial number of promastigote-infected leukocytes in the blood pool is probably sufficient to establish host infection . Nevertheless , subsequent macrophage and DDC phagocytosis of promastigotes released from neutrophils , or of apoptotic neutrophils , would boost the parasite load during the early silent phase post-infection . During natural transmission , sand flies deliver salivary bioactive components to the host , two of which could potentially interfere with early infection mechanisms: saliva and fPPG ( PSG ) . Sand fly saliva inhibits hemostasis and facilitates feeding [65] . Mice inoculated with sand fly salivary gland extracts show exacerbated lesion development , which is associated with early ( 6 h ) increase in type-2 cytokine production and with early ( 2 to 4 h ) expression of macrophage-recruiting chemokines that promote inflammatory cell influx to the injection site [66]–[68] . fPPG inoculated during sand fly bite causes substantial disease exacerbation in mice [6] . fPSG facilitates early parasite establishment by two mechanisms , macrophage recruitment to the infection site and enhancement of alternative macrophage activation , which upregulates arginase activity and promotes amastigote growth [69] . Early leukocyte recruitment to the infection site is detected in vivo from 40 min to 6 h [17] and in vitro after 4 h; the effect of enhanced alternative macrophage activation takes 24–48 h to develop [69] . In mouse , induction of arginase mRNA peaks at day 3 post-stimulation [70] and in man , leukocyte arginase is constitutively expressed only in granulocytes , independently of proinflammatory or anti-inflammatory stimuli [71] . In addition , neither PSG nor sand fly saliva appear to affect macrophage phagocytosis of L . mexicana metacyclic promastigotes [69] . We did not have access to PSG material and thus did not examine PSG activity in our system . We therefore cannot assert that PSG does not affect Leishmania early infection; nevertheless , the data cited above strongly suggest that PSG and bioactive salivary components exert their effects at a later stage of early infection reactions . Future experiments will help clarify the PSG effect on Leishmania ex vivo blood infection . The most effective innate mechanism against pathogens is said to be denial of access [72] . Leishmania has developed an extremely rapid and effective infection strategy , and the prospect of blocking initial promastigote access to a host seems highly improbable . Research efforts should focus on development of therapeutic approaches to prevent Leishmania establishment of permanent infection through enhancement of macrophage leishmanicidal mechanisms .
|
Leishmania infection is transmitted to mammalian hosts by phlebotomine sand flies . During the vector's bloodmeal , promastigotes are inoculated into hemorrhagic spots in the skin or are delivered into the extracellular matrix of the dermis . In the first case , blood is involved in transmission; in the second , it apparently is not . This is important , as the cellular milieu of infection can be critical for induction of the anti-parasite immune response and the subsequent course of disease . In humans , there are few comprehensive studies of the initial stages of Leishmania transmission in blood . Using blood to mimic the skin hematoma , we carried out kinetic and quantitative analyses of the reaction with serum and blood components of promastigotes from two Leishmania species with different tropism . We describe the kinetics of the promastigote reaction pathway that leads to blood infection and provide quantitative data for the cell types infected in the first five minutes of leishmaniosis transmission .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"infectious",
"diseases/protozoal",
"infections",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"immunology/innate",
"immunity"
] |
2010
|
Kinetic Analysis of Ex Vivo Human Blood Infection by Leishmania
|
The t-tubules of mammalian ventricular myocytes are invaginations of the cell membrane that occur at each Z-line . These invaginations branch within the cell to form a complex network that allows rapid propagation of the electrical signal , and hence synchronous rise of intracellular calcium ( Ca2+ ) . To investigate how the t-tubule microanatomy and the distribution of membrane Ca2+ flux affect cardiac excitation-contraction coupling we developed a 3-D continuum model of Ca2+ signaling , buffering and diffusion in rat ventricular myocytes . The transverse-axial t-tubule geometry was derived from light microscopy structural data . To solve the nonlinear reaction-diffusion system we extended SMOL software tool ( http://mccammon . ucsd . edu/smol/ ) . The analysis suggests that the quantitative understanding of the Ca2+ signaling requires more accurate knowledge of the t-tubule ultra-structure and Ca2+ flux distribution along the sarcolemma . The results reveal the important role for mobile and stationary Ca2+ buffers , including the Ca2+ indicator dye . In agreement with experiment , in the presence of fluorescence dye and inhibited sarcoplasmic reticulum , the lack of detectible differences in the depolarization-evoked Ca2+ transients was found when the Ca2+ flux was heterogeneously distributed along the sarcolemma . In the absence of fluorescence dye , strongly non-uniform Ca2+ signals are predicted . Even at modest elevation of Ca2+ , reached during Ca2+ influx , large and steep Ca2+ gradients are found in the narrow sub-sarcolemmal space . The model predicts that the branched t-tubule structure and changes in the normal Ca2+ flux density along the cell membrane support initiation and propagation of Ca2+ waves in rat myocytes .
Ventricular cardiac muscle cells have deep invaginations of the extracellular space known as t-tubules [1]–[14] . In rodents , these invaginations branch within the cell to form a complex network that allows rapid propagation of the electrical signal ( i . e . the action potential , AP ) to the subcellular location ( i . e . the sarcoplasmic reticulum , SR ) where the intracellular Ca2+ required for the cell contraction is stored [14] . The release of Ca2+ from the SR depends on “trigger Ca2+” entering the cytosol from the extracellular space by activating sarcolemmal Ca2+ channels ( L-type Ca2+ channels , LCC ) and by Ca2+ entry via Na+/Ca2+ exchanger ( NCX ) , [3] , [9] , [14] , [15] . The trigger Ca2+ activates SR Ca2+ release channels ( ryanodine receptors , RyRs ) by the process of “Ca2+-induces Ca2+-release” ( CICR ) which amplifies the modest increase in intracellular Ca2+ concentration ( [Ca2+]i ) caused by the LCC and NCX influxes to provide sufficient Ca2+ for the proteins regulating muscle force ( i . e . troponin C , TN ) ) , [14] . Thus , by working together , the microanatomy of t-tubules and SR permits spatially homogeneous and synchronized SR Ca2+ release and spatially uniform Ca2+ transients throughout the cell [5] , [14] , [16] . It has been also observed that the spatially uniform Ca2+ transients might be achieved if the SR Ca2+ release and uptake are abolished [5] . Yet , despite a wealth of information on ventricular cell function and structure , the mechanisms causing the synchrony of activation and the similarity of levels of [Ca2+]i across the myocyte still remain unclear . In cardiac muscle cells , several computational models have been introduced to investigate the Ca2+ signaling , buffering and diffusion [17]–[19] and Ca2+ wave initialization and propagation [12] , [20]–[23] . All these studies , however , are conducted on simplified geometries ( such as cylindrical or rectangular shapes ) and it has been pointed out that a small geometric change ( even in the case the change is uniformly applied ) could greatly influence the suggested homogeneous Ca2+ distribution by initiating wave propagation in the computer simulation [20] , [22] . Several laboratories , using common pool modeling approaches , have investigated also the effects of LCC and NCX distributions on global [Ca2+]i transients in dyadic , sub-sarcolemmal and cytosol compartments [10] , [24] , [25] . Recently , to examine how the distribution of Ca2+ flux along the sarcolemma affects Ca2+-entry and Ca2+ diffusion and buffering , we developed a 3-D continuum model in rat ventricular cells [19] . An important limitation of this model is that a cylindrical t-tubule geometry was assumed while several studies have provided evidence that in rodent ventricular myocytes the realistic t-tubule geometry is quite complex ( with large local variations in the diameter and transverse-axial anatomies ) , [9]–[12] . These experimental findings suggest that replacing our idealistic t-tubule model with a realistic geometry is needed . The use of idealistic shapes will change the diffusion distances and realistic Ca2+-transporting protein localizations in plane and depth directions and consequently the predicted [Ca2+]i distributions . In the present study , we sought to develop a morphologically correct geometric model of the t-tubule and to use this model for computational studies of the intracellular Ca2+ dynamics . We examined the Ca2+ signaling in rat ventricular myocytes that had been treated with ryanodine and thapsigargin to eliminate Ca2+ release and uptake by the SR . By using published electro-physiological data and laser-scanning confocal [Ca2+]i measurements , we were able to analyze several important spatial and temporal features of the Ca2+ signals in these cells . In this context , our goal was at least three-fold . The first aim was to develop a mathematical model that would be in qualitative or quantitative agreement with published experimental measurements on Ca2+ influx , and Ca2+ buffering and diffusion in rat ventricular cells with SR function inhibited [5] , [26] . Second , to use the model to investigate the importance of t-tubule ultra-structure and membrane Ca2+ flux distribution for the Ca2+ signals . The third task was to simulate the Ca2+ signals in the absence of fluorescent dye and to study the importance of the mobility of endogenous Ca2+ buffers ( ATP and calmodulin ) and altered extracellular Na+ ( [Na+]e ) for the Ca2+ signals . The analysis suggests that the quantitative understanding of the Ca2+ signaling requires more accurate knowledge of the t-tubule microanatomy and Ca2+ flux distribution along the sarcolemma . In agreement with experiment , with 100 µM Fluo-3 , the lack of detectible differences in the depolarization-evoked Ca2+ transients was found when the Ca2+ flux was heterogeneously distributed along the sarcolemma . In the absence of Fluo-3 , the predicted Ca2+ signals were strongly non-uniform . Even at modest elevation of Ca2+ , reached during Ca2+ influx , large and steep Ca2+ gradients may develop in the narrow sub-sarcolemmal space . The model also predicts that branched t-tubule topology and changes in the normal Ca2+ flux density along the cell membrane support Ca2+ waves initiated at the sarcolemma . Preliminary results of this work have been presented to the Biophysical Society in abstract form [27] .
Combining light microscopy ( LM ) and electron microscopy ( EM ) together with 3-D tomographic reconstruction , Hayashi et al . [6] investigated 3-D topologies of important sub-cellular organelles , including dyadic clefts and t-tubules , in mouse ventricular myocytes . In particular , the use of two-photon microscopy ( T-PM ) in their studies had provided data showing detailed spatial organization of t-tubules ( see Fig . 1A upper panel ) that was important for the development of our realistic model for computational studies of intracellular Ca2+ dynamics . The gap between imaging and simulation involves two major steps: ( 1 ) extracting features ( boundary or skeleton ) from imaging data; ( 2 ) constructing geometric models ( represented by meshes ) from the detected features . In addition , image pre-processing is usually necessary for better feature extraction , when the original image is noisy or the contrast between features and background is low . With 3-D T-PM images , Yu and collaborators developed a set of image processing and analysis tools and using the mesh generator called GAMer [28] they were able to generate high-fidelity and quality meshes for 3-D t-tubular systems in mice [29] ( see Fig . 1A lower panel ) . The extreme intricacy of the t-tubule system in mice ( with transverse-axial anatomies and large local variations in t-tubule diameter ) has been observed in rat ventricular myocytes as well [4] , [11] , [30] . Because high-fidelity geometric models representing the realistic t-tubule topology in rats are currently not available , in this study we used the geometric model in mice of Yu's et al . [29] . To investigate the Ca2+ signaling in rat ventricular myocytes , we considered a small compartment containing a single t-tubule and its surrounding half-sarcomeres for two reasons: ( a ) the entire t-tubular system in a ventricular myocyte forms a roughly periodic pattern corresponding to individual sarcomere; and ( b ) a small model contains much fewer number of mesh nodes that would render numerical simulation significantly faster and more feasible in ordinary computers . The surrounding half-sarcomeres were modeled as a rectangular-shaped box of 2 µm×2 µm in the plane of external sarcolemma and 5 . 96 µm in depth ( Fig . 1B left panel and Table 1 ) . As Yu's t-tubule model did not include the realistic cell surface , one of the box faces ( the top red surface in Fig . 1B ) was assumed to be the external cell membrane . The t-tubule inside this compartment was extracted from a sub-volume of the T-PM imaging data corresponding to the region indicated in Fig . 1A lower panel . To make it easier to handle boundary conditions in numerical analysis , we have “closed” the end of each branch , yielding a tree-like t-tubule model ( see the yellow mesh in Fig 1B ) . These added “caps” were treated with the same boundary conditions as for the rest of the t-tubular surface . This simplified treatment clearly could introduce some errors because these “caps” are artificial and no Ca2+-transporting protein should reside there . However , the errors should be negligible as the area of these “caps” is very small , compared to the rest of t-tubular surface . The t-tubule diameter varied from 0 . 19 µm to 0 . 469 µm and the t-tubule depth was 5 . 645 µm . The volume of the model compartment was estimated to be ∼23 . 31 µm3 . The compartment membrane area was ∼9 . 00 µm2 where the percentage of cell membrane within t-tubule was 64% ( ∼5 . 75 µm2 ) and within the external membrane 36% ( ∼3 . 25 µm2 ) , [10] , [11] , [31] , [32] . The accessible volume for Ca2+ was estimated to be ∼35–37% of the total cytosolic volume ( ) ∼12 . 9–13 . 6 pL in adult rat ventricular myocytes [33] , [34] . The sub-cellular aqueous volume of 35–37% assumes that: ( 1 ) myofilaments occupy 47–48% of the cell volume , mitochondria 34–36% , nucleus 0–2% , t-tubule system 0–1 . 2% , and SR lumen 3 . 5%; ( 2 ) 50% of the myofilament space is accessible for Ca2+ ( i . e . contains water ) ; ( 3 ) mitochondria and nuclei are not rapidly accessible for Ca2+; ( 4 ) the SR lumen is not accessible to Ca2+ in the presence of ryanodine and thapsigargin [18] , [33] . Recent immunohistochemical studies have demonstrated that marked variations in the distribution of Ca2+-transporting protein complexes ( L-type Ca2+ channel , Na+/Ca2+ exchanger ) along the cell membrane probably exist [2] , [10] , [12] , [15] , [35]–[43] . The analysis suggests that most of the L-type Ca2+ channels are concentrated in the t-tubules ( from 3 to 9 times more in the t-tubule membrane than on the external sarcolemma ) [2] , [10] , [12] , [38] , [39] and that the concentration of LCC along the t-tubule probably increases toward the center of the cell [36] , [40] . Studies on the distribution of the main Ca2+ efflux pathway , the Na+/Ca2+ exchanger , are more controversial . All studies but one [41] have reported NCX to localize both to the external and t-tubule membrane , but most studies suggest that the NCX is 1 . 7 to 3 . 5 times more concentrated in the t-tubule membrane [15] , [34] , [36] , [42] , [43] . However , Kieval et al . data [43] indicate the NCX is more evenly distributed . In summary , the observed differences in the spatial distribution and molecular architecture of Ca2+ microdomains suggest that significant differences in the excitation-contraction coupling between the cell surface and cell interior may be exist . However how the localization of Ca2+- transporting protein complexes along the sarcolemma regulates the intracellular Ca2+ signaling still remains uncertain . In the current model , the effects of four exogenous and endogenous Ca2+ buffers ( Fluo-3 , ATP , calmodulin , troponin C ) were considered ( Fig . 1C ) . The endogenous stationary buffer troponin C ( TN ) was distributed uniformly throughout the cytosol but not on the cell membrane and in the sub-sarcolemmal space ( ∼40–50 nm in depth ) . The free Ca2+ and mobile buffers ( Fluo-3 , ATP , calmodulin ) diffuse and react throughout the cytoplasm . The cell membrane and sarcomere box faces are subject to reflective boundary conditions . The nonlinear reaction-diffusion equations describing Ca2+ and buffers dynamics inside the model cell are: ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) where: [Bm] represents the concentration of mobile buffer Fluo-3 , ATP or calmodulin; [Bs] is the concentration stationary buffer troponin C . The diffusion coefficients for Ca2+ , CaATP , CaCal and CaFluo as well as the total buffer concentrations and buffer rate constants used in the model are shown in Table 2 . In the model we also assume: ( 1 ) isotropic diffusion for Ca2+ and all mobile buffers [12]; ( 2 ) Ca2+ binds to Fluo-3 , calmodulin , ATP , and TN without cooperativity; ( 3 ) the initial total concentrations of the mobile buffers are spatially uniform; ( 4 ) the diffusion coefficients of Fluo-3 , ATP or calmodulin with bound Ca2+ are equal to the diffusion coefficients of free Fluo-3 , ATP or calmodulin . The total Ca2+ flux ( ) throughout the t-tubule and external membrane is: ( 6 ) where: - total LCC Ca2+ influx; - total NCX Ca2+ flux; - total membrane Ca2+ leak . To describe L-type Ca2+ current , Na+/Ca2+ exchanger , Ca2+ leak current densities we used the following expressions: ( 7 ) ( 8 ) ( 9 ) ( 10 ) Flux parameter values were estimated or taken from the literature ( see Table 3 ) . In this study , the Ca2+ leak is not actually a particular “leak protein” . The Ca2+ leak was included and adjusted so that at rest Ca2+ influx via Ca2+ leak to match Ca2+ efflux via NCX thus no net movement across the cell membrane to occur . In the model , each current density ( Ii ) was converted to Ca2+ flux ( Ji ) by using the experimentally suggested surface to volume ratio ( ∼8 . 8 pF/pL ) in adult rat ventricular myocytes [32] , [33]: ( 11 ) Then , the total compartment Ca2+ flux ( ) was computed by multiplying each total Ji with the model cell volume ( ) , and distributing to the external and t-tubule membrane according to the prescribed Ca2+-handling protein concentration ratio . The voltage-clamp protocol ( holding potential −50mV , electric pulse of 10mV for 70ms ) and whole-cell L-type Ca2+ current were derived from Zahradnikova et al . data with the blocked SR activity [26] . In this study , each simulation started with a basal cytosolic Ca2+ of 100 nM , basal cytosolic Na+ of 10 mM and buffers in equilibrium . The extracellular Ca2+ concentration ( ) was 1 mM and remained constant . Unless specified otherwise in the Figure legends or in the text , the extracellular Na+ concentration ( ) was 140 mM and 3 . 4−6 µM mV−1 ms−1 . In finite element methods , a complex domain needs to be discretized into a number of small elements ( such as triangles or tetrahedra ) . This process is usually referred to as mesh generation [28] , [44] . Although different types of meshes may be generated depending on the numerical solvers to be employed , we restrict ourselves to triangular ( surface ) and tetrahedral ( volumetric ) mesh generation as commonly used in biomedical simulation . In the present simulation , the number of finite element nodes and tetrahedral elements are 50 , 262 and 221 , 076 , respectively . The nonlinear reaction diffusion system was solved by a finite difference method in time and finite element method in space using our SMOL software tool ( Smoluchowski Solver , http://mccammon . ucsd . edu/smol/ ) with the time step of 4 ms . It takes around 20 minutes to run 400 ms snapshots with a single Intel Xeon X5355 processor . The SMOL program utilizes libraries from the finite element tool kit ( FETK ) , which previously has been used in several molecular level studies [45]–[49] . One bottleneck for dynamic 3-D simulation of nonlinear reaction diffusion system is the computing complexity involved in solving the problem . Here we successfully extended SMOL to solve multiple coupled partial differential equations with nonlinear ordinary equations . Multiple tests demonstrate that our SMOL program is quite robust and flexible for various boundary and initial conditions . The simulation results were visualized using GMV mesh viewers [50] . Post-processing and data analyses were implemented by customized Python , MATLAB 2008b ( The MathWorks , Natick , MA ) scripts and Xmgrace software [51] . A version control system , subversion , was used to monitor the development of software [52] .
In agreement with the reported experimental data [2] , [10] , [12] , [36] , [38]–[40] , the spatial patterns of [Ca2+]i were calculated assuming LCC current density: ( 1 ) heterogeneously distributed along the cell surface; ( 2 ) six times higher and uniform in the t-tubule membrane; or ( 3 ) homogeneously distributed along the sacrolemma . In cases ( 1–2 ) the NCX flux density was assumed three times higher in the t-tubule and in case ( 3 ) NCX was evenly distributed along the sarcolemma . In this study , Ca2+ leak density was homogeneously distributed along the cell membrane with respect to all distribution choices of LCC and NCX . In case ( 1 ) , the 3-D distribution of LCC current was computed by combining the cluster density and fluorescent intensity plots , see Fig . 2A . The data were then scaled and fitted by a cubic polynomial: ( 12 ) where: x is the distance from the external cell surface . The parameter values of the polynomial ( pj , j = 1–4 ) are shown in Table 4 . This polynomial was further scaled by a single factor C ( see Table 4 ) such that the total Ca2+ flux along the t-tubule membrane remained unchanged by redistributing the Ca2+ fluxes . To fit the whole-cell LCC current density to the reported data in rat myocytes with SR release inhibited [26] , we used a shape preserving function , ( see Eq . 8 and Fig . 2B ) . Consistent with the Cheng et al . experiment [5] , where the fluorescence signal was recorded along the single scan-lane starting and ending outside the cell and crossing the center of the cell , the model t-tubule was chosen to cross the cell center and the scanned line was located at 200nm away from the t-tubule membrane ( see Fig . 1A and Fig . 3 ) . To gain more detailed insights of how the predicted Ca2+ signals are regulated within this geometrically irregular micro-domain we examined [Ca2+]i at two different line-scan positions: 200 nm , angle 120°; 200 nm , angle 60° ( see Fig . 3 ) . Model results in Figs . 4–5 were computed for conditions approximating those of the experiment by Cheng et al . [5] , who examined Ca2+ signals in voltage-clamped rat myocytes in the presence of 100 µM Fluo-3 and pharmacological blockade of the SR ( see Fig . 4L ) . The computed line-scan images and local Ca2+ time-courses are shown in Figs . 4F–4K . These results demonstrate that with LCC heterogeneous or LCC six times higher in the t-tubule: ( 1 ) predicted Ca2+ concentration profiles were non-uniform when t<100 ms but the variations in [Ca2+]i seem to be within the range of experimental noise in Fig . 4L; ( 2 ) [Ca2+]i was more evenly distributed when t>100 ms , ( Figs . 4F–G , Figs . 4I–J , Video S1 ) . To delineate further the suggested spatial differences in [Ca2+]i ( see Figs . 4F–H ) , we introduced a quantity called ‘spatial Ca2+ heterogeneity’ ( SCH ) . The SCH is defined to be the difference of the maximal and minimal [Ca2+]i values , normalized by the maximal value at given reference point along the line-scan ( 0 . 17 µm , 3 . 09 µm , 5 . 45 µm ) in given moment tj , ( see Figs . 4I–K ) . High SCH suggests non-uniform [Ca2+]i distribution and SCH of zero indicates spatially uniform [Ca2+]i distribution . The histogram in Fig . 4M shows that assuming LCC heterogeneous versus LCC 6 times higher in the t-tubule decreased SCH ( tIca-peak ) by 1 . 6 folds , SCH ( t70 ms ) by 2 . 29 folds , SCH ( [Ca]i-peak ) by 2 . 34 folds , SCH ( t100ms ) by 2 . 87 folds , and SCH ( t200ms ) by 8 . 45 folds . These findings demonstrate that the predicted [Ca2+]i distribution with LCC heterogeneous more closely resembles the Chang et al . experimental findings [5] , ( compare Figs . 4I and 4L ) . Finally , the model predicts strongly non-uniform Ca2+ transients when the LCC , NCX and Ca2+ leak fluxes were uniformly distributed throughout the cell surface ( Figs . 4H , 4K , 4M ) . In addition , Video S1 ( see right panel ) demonstrates that here the Ca2+ signal spreads from the external membrane to the cell center as a continuum wave but after LCC channel closing ( t∼72 ms ) this wave faltered . Predicted global [Ca2+]i transient , Na+/Ca2+exchanger , and Ca2+ leak currents with LCC pathways heterogeneously distributed ( as in Fig . 4F ) are shown in Figs . 4E and 4C–D . Figure 4C demonstrates that: ( 1 ) the depolarization of cell membrane reversed the rest exchanger's direction ( i . e . in the interval 0ms–70ms NCX operated in Ca2+ entry mode ) while when repolarization occurred the flow of Ca2+ through NCX was reversed again ( i . e . in the interval 70ms–400 ms NCX operated in Ca2+ exit mode ) ; ( 2 ) upon returning to resting voltage of −50mV the exchanger's rate rapidly increased ( ) while rate remained unchanged ( note is not voltage-dependent ) thus causing fast extrusion of Ca2+ out of the cell and subsequent sudden drop in the local and global [Ca2+]i . Figures 4F–K illustrate also that the global and all local Ca2+ transients reached the peak after ∼76 ms and that [Ca2+]i levels were higher near the t-tubule mouth because the density of t-tubule branches was higher in this region and close topological proximity of the external membrane additionally increased the relative amount of the entering Ca2+ . Due to the higher [Ca2+]i gradient under the outer cell edge ( t∼70ms ) Ca2+ diffused toward the cell center and when ratio along the cell membrane became approximately equal to ratio a new equilibrium level of [Ca2+]i ( ∼0 . 16 µM ) was reached . Intracellular Ca2+ equilibrated faster when Ca2+ flux was more concentrated in the t-tubule membrane because [Ca2+]i gradient near the t-tubule mouth was lower there than [Ca2+]i gradient with Ca2+transporting complexes distributed homogeneously . Additional reasons for the observed rapid equilibrium of [Ca2+]i may be that [Na+]i was kept constant ( in contrast to existing evidence for the presence of local sub-sarcolemmal [Na+]i gradients on the action potential time-scale [33] , [53] ) or that the realistic distribution of NCX flux may be differ as assumed in the model . Finally , the results demonstrate that the computed average [Ca2+]i peak of 160–185 nM ( see Figs . 4E , 4I–4J ) , is comparable with the measured one of about 163 nM when the SR release and uptake were inhibited [5] . This model is also able to predict how the Ca2+ transients are regulated at different line-scan positions within this geometrically irregular micro-domain . Note , due to the technical limitations the Cheng et al . experiment is not able to suggest where exactly the scanned line is positioned with regard to the specific t-tubule , but the Cheng et al . measurements [5] strongly suggest the similarity of [Ca2+]i at the peripheral and deeper cytoplasm when the SR activity is abolished . For this reason , we examined the Ca2+ profiles ( LCC heterogeneous along the t-tubule , Fig . 4F ) positioning the line-scan at 200 nm and angle 60° ( see Fig . 3 ) or positioning the line-scan at 50 , 100 , 200 , 300 or 400 nm at different angles . No visible differences in the visualized spatial Ca2+ profiles were found ( data not shown ) . The 3-D Ca2+ concentration distributions and spatial Ca2+ profiles at Ca2+ peak ( 76 ms ) are shown in Figs . 5A–D . These model results demonstrate that the Ca2+ concentration near the external membrane decreased while [Ca2+]i in the cell interior increased when Ca2+ transporters were uniformly distributed and after that heterogeneously redistributed . The jumps in Fig . 5D show the predicted local Ca2+ flux ( ) at [Ca2+]i peak in the regions where the scanning line of interest is crossing the t-tubule membrane . Additional interesting model findings are that: ( 1 ) large and steep [Ca2+]i gradients were predicted inside the sub-sarcolemmal 3-D space ( see Video S1 ) ; ( 2 ) the global Ca2+ time-course and time to [Ca2+]i peak did not depend on whether Ca2+ fluxes are distributed homogeneously or heterogeneously along the sarcolemma ( data not shown ) ; ( 3 ) redistributing NCX flux uniformly via the sarcolemma was not able to alter significantly the predicted Ca2+ signals in Fig . 4F ( data not shown ) . In this study the value of 390 µm2 s−1 for diffusion coefficient of free Ca2+ and published buffer diffusion coefficients and parameters were used to compare the calculated Ca2+ signals with the Cheng's et al . fluorescence Ca2+ signals recorded in rats [5] . It has been suggest , however , that the effective diffusion of free Ca2+ in the cytosol ( ) will be slowed down because the exogenous and endogenous Ca2+ buffers and free Ca2+ concentrations are able to affect Ca2+ diffusion strikingly [18] , [19] , [54]–[62] . The measurements of Allbritton et al . [54] report a value of 5–21 µm2 s−1 for at low free when a value of ∼223 µm2 s−1 is assumed for . During Allbritton's et al . in vitro experiments Ca2+ sequestration by the SR , mitochondria and ATP was inhibited and only mobile calmodulin and stationary troponin C were present in the cytosolic extract . It is possible to estimate ( 0 µM Fluo-3 , 0 µM ATP ) , using a simplified equation ( see Eq . 13 ) , because in this study the predicted maximal Ca2+ elevations were sufficiently small ( ) and the diffusion coefficients for Ca2+-bound and free mobile buffer forms were assumed equal [18] , [19] , [58] , [59] . Our calculations predict a value of ∼8 µm2 s−1 for when was 390 µm2 s−1 and ∼6 µm2 s−1 when was 223 µm2 s−1 . Therefore , the estimated value for ( ∼8 µm2 s−1 ) is in reasonable agreement with the experimental observation . ( 13 ) We could not find experimental data suggesting ( 260 µM ATP , 70 µM TN , 24 µM Cal ) or ( 100 µM Fluo-3 , 260 µM ATP , 70 µM TN , 24 µM Cal ) in the solution . Therefore , we used published concentrations of Ca2+ binding proteins and published diffusion and dissociation constants ( ) to estimate the effective diffusion constants of free Ca2+ in the presence of ATP or Fluo-3 and ATP in the cytosol . Our calculations indicate that adding 260 µM ATP in the solution accelerated ( ∼10 . 4 µm2 s−1 ) and that increased additionally when 100 µM Fluo-3 was added ( ∼66 µm2 s−1 ) . Furthermore , our studies suggest that in the presence of 100 µM Fluo-3 and LCC heterogeneous Ca2+ binding and diffusion of ATP and calmodulin could not affected significantly the predicted [Ca2+]i distributions ( data not shown ) . During simulations of SR Ca2+ release into the diadic cleft , a major effect of the stationary phospholipids Ca2+ binding sites has been suggested [17] , [18] . To examine the impact of the phospholipids on the much smaller Ca2+ signals ( arising from Ca2+ influx via Ca2+ current only ) , we included this buffer in our model . The results demonstrated that the phospholipids had only a limited effect on the calculated Ca2+ signals in the sub-sarcolemmal region ( 0 µM Fluo-3 , 260 µM ATP , 24 µM calmodulin ) and that this effect was even smaller when 100 µM Fluo-3 was included ( data not shown ) .
The current study attacks a difficult problem on how to incorporate the structural-based biological information , critical for the subcellular and cellular function , into sophisticated computational investigations . Pursuing this goal we developed a 3-D continuum model of Ca2+ signaling in rat ventricular cells that incorporates the realistic transverse-axial t-tubule topology and considers geometric irregularities and inhomogeneities in the distribution of ion-transporting proteins . The t-tubule micro-architecture was extracted from the Hayashi et al . two-photon imaging data in mice [6] . Because currently high-fidelity geometric models representing the realistic t-tubule micro-architecture in rats are not available , in this study we used the Yu's et al . geometric model in mice [28] , [29] . On the basis of experimental data in rats the aqueous sub-cellular volume , accessible to Ca2+ , was estimated to be ∼35–37% [33] . Since the Ca2+ signaling in cells is largely governed by Ca2+ diffusion and binding to mobile and stationary Ca2+ buffers [17]–[21] , [23] , [54]–[62] , [66] , [67] , the effect of four Ca2+ buffers ( Fluo-3 , ATP , calmodulin , TN ) was considered . The model was validated against published experimental data on Ca2+ influx , membrane protein distributions and Ca2+ diffusion in rat cells treated with ryanodine and thapsigargin to inhibit the SR Ca2+ metabolism [2] , [5] , [10] , [12] , [15] , [26] , [35]–[43] . We found that with 100 µM Fluo-3 the results more closely resemble the Cheng's et al . experimental data [5] when the LCC density increases ∼1 . 7 fold along the t-tubule length and the NCX density is assumed three times higher in the t-tubule . An interesting finding is that with LCC six times and NCX three times higher and uniform in the t-tubule , the predicted fluctuations in the [Ca2+]i profiles were within the range of experimental noise [5] . Strongly non-uniform spatial Ca2+ gradients and propagation of Ca2+ wave are predicted , not observed in Cheng et al . experiment , when the LCC and NCX were uniformly distributed along the sarcolemma . The model studies with 100 µM Fluo-3 indicate also that the [Ca2+]i gradients depend on the diffusion distances in the axial and cell surface directions . Thus , when the LCC were distributed uniformly the local Ca2+ peak in radial depth ( 5 . 96 µm ) decreased from ∼1 . 5 fold while in the other cell directions ( 1 µm×1 µm ) no significant changes were found . Redistributing the amount of Ca2+ pumped via the cell membrane ( i . e . increasing LCC current density along the t-tubule ) while keeping total Ca2+ flux unchanged , lowered Ca2+ gradients near the surface membrane and increased Ca2+ levels in the cell interior ( see Video S1 ) . The results also showed that with 100 µM Fluo-3 and Ca2+ flux heterogeneously distributed along the sarcolemma , the computed average [Ca2+]i peak ( 160–185 nM ) is comparable to the measured of about 163 nM [5] and that the NCX redistribution alone yields to qualitatively similar [Ca2+]i profiles . It should be noted , that in our previous work we used the simplified t-tubule geometry ( assuming cylindrical shape ) to simulate the Ca2+ dynamics in rats [19] . This idealistic t-tubule model also predicts the lack of systematic differences in the fluorescence Ca2+ signal when the Ca2+ transporters were distributed heterogeneously along the sarcolemma . Thus , the following question arises: How these new computational studies based on more realistic t-tubule structural model will further advance our current knowledge on the cell excitability and Ca2+ cycling in rats ? First , in agreement with experiment [2]–[12] current study confirms that due to the branched t-tubule microstructure high and steep sub-sarcolemmal [Ca2+]i gradients could occur throughout the whole cell volume [2]–[6] , [33] , ( see Video S1 ) . Note , Lu et al . idealistic t-tubule model predicts high and steep sub-sarcolemmal [Ca2+]i gradients only in the transverse cell direction [19] . Second , our realistic t-tubule model predicts non-uniformities in [Ca2+]i distribution along the depth of the t-tubule when t<100 ms ( see Fig . 4F and Video S1 left panel ) while this was not the case when the t-tubule geometry is assumed cylindrical ( Fig . 4g in Lu et al . , [19] ) . Third , interesting finding is that no visible differences in the local Ca2+ profiles are predicted when the line-scan was positioned at different . Note , due to the technical limitations the Cheng et al . experiment is not able to suggest where and how exactly the scanned line is positioned with regard to the specific t-tubule [5] . A surprising and important finding of this study is that the spread and buffer capacity of 100 µM Fluo-3 were able to mask completely the pronounced spatial [Ca2+]i non-uniformities that would occur during the Ca2+ influx in the absence of dye ( see Video S2 left panel - SR Ca2+ metabolism inhibited , LCC and NCX transporters heterogeneously distributed ) . Here the simulations demonstrated that with zero Fluo-3 the local and global Ca2+ peaks increased while the time of Ca2+ rise remained unchanged . The predicted sub-sarcolemmal [Ca2+]i gradients were heterogeneous along the cell membrane and larger and steeper compared to those with 100 µM Fluo-3 . The NCX and Ca2+ leak time-courses were also affected due to increased local free [Ca2+]i levels . It is interesting that under these conditions no differences in the local Ca2+ time-courses were found ( as with 100 µM Fluo-3 ) when the line-scan was positioned at different angles and distances . To test further the model we also examined how the mobility of endogenous Ca2+ buffers ( ATP and calmodulin ) and altered extracellular Na+ ( [Na+]e ) would affect the Ca2+ signals in the absence of fluorescence dye when the Ca2+-transportes are heterogeneously distributed . The results showed that when ATP and calmodulin were immobilized Ca2+ diffuses slowly toward the center of the cell , resulting in higher Ca2+ concentrations near the outer cell edge . When the NCX forward mode was inhibited ( assuming [Na+]e = 0 mM ) the local [Ca2+]i peaks increased and this increase was more pronounced near the outer cell edge . New findings are that under these conditions near the outer cell edge Ca2+ wave was initiated while this was not the case when ATP and calmodulin were mobile and [Na+]e 140 mM ( see Video S2 and Video S3 ) . Taken together , these studies provide compelling evidence that ( 1 ) the exogenous Fluo-3 acts as a significant buffer and carrier for Ca2+ , and that ( 2 ) the use of 100 µM Fluo-3 during the experiment can sensitively alter the realistic Ca2+ distribution . A new the question , however , arises: Based on the above model findings what could be the underlying mechanism ( s ) for the predicted heterogeneous Ca2+ concentrations gradients in the absence of Fluo-3 ? A reasonable answer is that the Ca2+ movement and distribution inside the cell rely strongly not only on the specific cell micro-architecture and Ca2+ transporters distribution but also on the presence of endogenous mobile and stationary Ca2+ buffers ( ATP , calmodulin , troponin C - known to affect strikingly the Ca2+ dynamics ) [12] , [17]–[21] , [23] , [27] , [67] . In support of this hypothesis , our simulations studies revealed that in the absence of Fluo-3: ( 1 ) the stationary Ca2+ buffer troponin C imposed stronger diffusion barrier for Ca2+ that resulted in larger and steeper sub-sarcolemmal Ca2+ gradients; ( 2 ) in the cell interior , owing on their sheer buffering capacity , Ca2+ buffers ( troponin C , ATP , calmodulin ) tended to slow down additionally the propagation of Ca2+ so that ATP and calmodulin spreading alone was not able to contribute the spatially uniform Ca2+ profiles to be achieved; ( 3 ) immobilizing the endogenous Ca2+ buffers slowed down the Ca2+ movement from the cell periphery to the center leading to build-up of large sub-sarcolemmal Ca2+ gradients and subsequent initiation of Ca2+ wave . It is important to mention that the Lu et al . idealistic t-tubule model predicts completely different 3-D [Ca2+]i distributions with zero Fluo-3 , SR Ca2+ metabolism inhibited and Ca2+ transporters heterogeneously distributed [19] . Important limitations of the current modeling approach are: ( 1 ) the relatively small size of the model compartment that contains only a single realistic t-tubule shape and spans by just a half-sarcomere inside the ventricular myocyte; and ( 2 ) the assumption that the model compartment is a repeating unit inside the cell . The structural studies , however , provide evidence that in rodent ventricular myocytes the realistic t-tubule network is quite complex , ( see Fig . 1A ) , [6] . The above limitations can be overcome in the future by extending the current geometric model toward more realistic models containing several t-tubules , whole-cell t-tubule network or other sub-cellular organelles ( including mitochondria , SR , nuclei ) . This would allow building an improved geometric models representing more correctly the cell segment of interest and help to gain further insights of how the Ca2+-signaling in rat ventricular myocytes is regulated in the absence or presence of SR Ca2+ release and uptake [20]–[23] , [66] , [67] . However , it is equally important to mention here , that although the limitations ( 1–2 ) this model in a first approximation may yield insights across the whole-cell scale of biological organization . It allows simulating the global Ca2+ signal ( computed from the line-scan image in Fig . 4F ) that roughly would reproduce the whole-cell Ca2+ transient in the Cheng et al . experiment due to observed spatial similarities in [Ca2+]i ( see Fig . 1B–1C in [5] ) . This assumption enables also investigating how the whole-cell Ca2+ signal is regulated by the realistic t-tubule microanatomy , by 3-D distributions of ion-transporting proteins , by mobile dye or endogenous mobile and stationary Ca2+ buffers . It should be noted , that the common pool modeling approaches could not be used to investigate these effects [10] , [24] , [25] . Important limitation of this study is also that we assume that the ion flux pathways are continuously distributed throughout the t-tubule membrane . Immunohistochemical studies , however , suggest that L-type Ca2+ channels appear to be concentrated as discrete clusters in the dyadic clefts ( narrow spaces between LCC and RyR ) distributed regularly along the t-tubule membrane at relatively small distances of ∼0 . 68 µm , [12] , [38] , [68] . It is interesting to mention that in contrast to Soeller and collaborators data in rats [12] , Hayashi et al . data in mice [6] suggest that the dyadic clefts are distributed irregularly along the t-tubule branches . In addition , NCX appears to be absent from the longitudinal tubules [42] . Thus , the above data clearly imply that localized concentration of LCC or NCX flux pathways could result in larger sub-sarcolemmal Ca2+ gradients and local membrane currents that will affect differently the spatial Ca2+ profiles as predicted with the current model . Further extending of our current t-tubule model toward more realistic geometric models containing dyadic cleft topology and L-type Ca 2+ channel clustering could help to better understand how the Ca2+ signaling is regulated in the heart . Finally , in the present model the effects of two endogenous Ca2+ mobile buffers ( ATP , calmodulin ) and one stationary Ca2+ buffer ( troponin C ) were considered only . The ventricular cells , however , contain other stationary Ca2+ buffers ( including phospholipids , myosin , calsequestrin ) or small and high mobile Ca2+ binding molecules ( ADP , AMP ) that were not included in the model ( or may be other stationary and mobile buffers that have not been identified yet ) , [18] , [24] , [69] . Simulations presented in this study demonstrate that the more accurate knowledge of transverse-axial t-tubule microanatomy and protein distributions along the sarcolemma is important to better understand the mechanisms regulating the excitation-contraction coupling in rat ventricular myocytes . The results demonstrate that Ca2+ movement from the cell membrane to the cell interior relies also strongly on the presence of mobile and stationary Ca2+ buffers , including the Ca2+ indicator dye . The key findings are: ( 1 ) the model predicts lack of detectible differences in the fluorescence Ca2+ signals at the peripheral and deep myoplams when the membrane Ca2+ flux is heterogeneously distributed along the sarcolemma; ( 2 ) 100 µM mobile Fluo-3 is able to mask the pronounced spatial non-uniformities in the [Ca2+]i distribution that would occur when the SR Ca2+ metabolism is inhibited; ( 3 ) during the Ca2+ influx alone , large and steep Ca2+ gradients are predicted in the narrow sub-sarcolemmal space ( ∼40–50 nm in depth ) ; ( 4 ) in rodents the branched t-tubule topology , the punctuate spatial distribution of Ca2+ flux along the sarcolemma and the endogenous Ca2+ buffers actually function to inhibit Ca2+ waves . Improved functional and structural computational models are needed to guide the experiments and to test further our understanding of how the t-tubule microanatomy and protein distributions regulate the normal cell function or cell cycle under certain pathologies . To our best knowledge , this study is the first attempt to use the finite element methods to investigate the intracellular Ca2+ responses in physiologically realistic transverse-axial t-tubule geometry .
|
In cardiac muscle cells , calcium ( Ca2+ ) is best known for its role in contraction activation . A remarkable amount of quantitative data on cardiac cell structure , ion-transporting protein distributions and intracellular Ca2+ dynamics has been accumulated . Various alterations in the protein distributions or cell ultra-structure are now recognized to be the primary mechanisms of cardiac dysfunction in a diverse range of common pathologies including cardiac arrhythmias and hypertrophy . Using a 3-D computational model , incorporating more realistic transverse-axial t-tubule geometry and considering geometric irregularities and inhomogeneities in the distribution of ion-transporting proteins , we analyze several important spatial and temporal features of Ca2+ signaling in rat ventricular myocytes . This study demonstrates that the computational models could serve as powerful tools for prediction and analyses of how the Ca2+ dynamics and cardiac excitation-contraction coupling are regulated under normal conditions or certain pathologies . The use of computational and mathematical approaches will help also to better understand aspects of cell functions that are not currently amenable to experimental investigation .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"cardiovascular",
"disorders/arrhythmias,",
"electrophysiology,",
"and",
"pacing",
"mathematics",
"physiology/cardiovascular",
"physiology",
"and",
"circulation",
"biophysics/theory",
"and",
"simulation",
"computational",
"biology/systems",
"biology"
] |
2010
|
Numerical Analysis of Ca2+ Signaling in Rat Ventricular Myocytes with Realistic Transverse-Axial Tubular Geometry and Inhibited Sarcoplasmic Reticulum
|
Genome-wide sequence divergence between populations can cause hybrid sterility through the action of the anti-recombination system , which rejects crossover repair of double strand breaks between nonidentical sequences . Because crossovers are necessary to ensure proper segregation of homologous chromosomes during meiosis , the reduced recombination rate in hybrids can result in high levels of nondisjunction and therefore low gamete viability . Hybrid sterility in interspecific crosses of Saccharomyces yeasts is known to be associated with such segregation errors , but estimates of the importance of nondisjunction to postzygotic reproductive isolation have been hampered by difficulties in accurately measuring nondisjunction frequencies . Here , we use spore-autonomous fluorescent protein expression to quantify nondisjunction in both interspecific and intraspecific yeast hybrids . We show that segregation is near random in interspecific hybrids . The observed rates of nondisjunction can explain most of the sterility observed in interspecific hybrids through the failure of gametes to inherit at least one copy of each chromosome . Partially impairing the anti-recombination system by preventing expression of the RecQ helicase SGS1 during meiosis cuts nondisjunction frequencies in half . We further show that chromosome loss through nondisjunction can explain nearly all of the sterility observed in hybrids formed between two populations of a single species . The rate of meiotic nondisjunction of each homologous pair was negatively correlated with chromosome size in these intraspecific hybrids . Our results demonstrate that sequence divergence is not only associated with the sterility of hybrids formed between distantly related species but may also be a direct cause of reproductive isolation in incipient species .
Separate species are often reproductively isolated by intrinsic postzygotic mechanisms . The diverged genomes from two different parental populations may not interact properly when combined in a hybrid , resulting in reduced fertility , reduced viability , or both [1] . Errors during gamete production , particularly problems associated with altered meiotic recombination between diverged genomes , are becoming recognised as a widespread cause of hybrid sterility [2] . While recent work on the contribution of meiotic recombination to reproductive isolation has focused on individual genes [3] , seminal studies of the baker's yeast Saccharomyces cerevisiae have established genome-wide sequence divergence as a cause of meiotic errors underlying hybrid sterility [4 , 5] . Sequence divergence between the genomes of different species is thought to decrease the rate of meiotic crossing over through the action of the anti-recombination machinery , which prevents recombination between dissimilar sequences . Low rates of recombination between diverged genomes can cause problems during hybrid meiosis since at least one reciprocal exchange ( crossover ) event per homologous pair of chromosomes is necessary to ensure correct segregation [4 , 6] . Failure of homologous chromosomes to segregate correctly ( nondisjunction ) generates meiotic products ( gametic spores ) that either contain too many or too few chromosomes , resulting in reduced hybrid spore viability; all Saccharomyces chromosomes carry essential genes , and failing to inherit any chromosome results in gamete inviability . The magnitude of the contribution of nondisjunction to postzygotic reproductive isolation in Saccharomyces yeasts is not clear because accurate measurements of nondisjunction rates are not currently available . Attempts to quantify the rate of nondisjunction in Saccharomyces hybrids have relied on the analysis of colonies grown from single spores ( obtained by tetrad dissection or random spore analysis ) . These studies have suffered from numerous methodological biases stemming from three principal sources: ( 1 ) chromosome loss during mitotic growth , ( 2 ) limited sensitivity of the techniques used for measuring ploidy , and ( 3 ) the inability to genotype inviable spores . As a result , neither extra copies of chromosomes nor missing copies can be accurately quantified using single-spore–derived colonies . Extra copies of chromosomes ( disomes ) inherited by haploid spores during meiosis are highly unstable and prone to mitotic loss following germination [7] . Disomes generally do not have large effects on the probability of spore germination in S . cerevisiae [8 , 9] but do tend to reduce mitotic growth rates relative to eusomic strains [10 , 11] . As a result , in a growing colony derived from a single disomic spore , mutants that lose one copy of a disomic chromosome will gain a growth advantage over disomic cells and increase in frequency , potentially becoming the dominant type in a heterogeneous colony . Moreover , by generating stoichiometric imbalances in proteins important for mitotic segregation , aneuploidy itself might increase genomic instability [12] . Disome loss has been demonstrated using the meiotic products of triploid yeast [8] . In a triploid meiosis , each chromosome will be inherited as a disome by two of the spores in the resulting tetrad and as a monosome ( single copy ) by the other two spores ( resulting in an average of 8 disomes per spore ) . However , when chromosomes are counted in colonies derived from the four spores of a four-viable–spore tetrad , disomes are underrepresented relative to monosomes . The magnitude of the deficit depends on the sensitivity of the assay used to detect the presence or absence of each chromosome [8 , 13 , 14] . When chromosome loss occurs early in the growth of a colony or is associated with a large increase in growth rate , the lost chromosome may be present in only a small proportion of cells in the colony when analysed , and correct diagnosis of disomy in the founding spore would require a highly sensitive assay . For example , St . Charles and colleagues [8] detected only 853 of 960 expected chromosomes in 10 four-viable–spore triploid-derived tetrads using a comparative genome hybridisation assay , corresponding to an average of 5 . 3 disomes per spore . When a more sensitive PCR-based assay was used , they detected 945 of 960 expected chromosomes , corresponding to an average of 7 . 6 disomes per cell . Chromosome loss during mitotic growth means that measuring disome frequency in single-spore–derived colonies will always underestimate the true magnitude of the problem . The frequencies of disomes in the gametes of yeast hybrids generated by crossing S . cerevisiae and S . paradoxus have been estimated using multiple techniques with inconsistent results . Hunter and colleagues [4] reported highly variable frequencies of disomes for different chromosomes , ranging from 0% ( chromosome [Chr] VI ) to 27 . 2% ( Chr II ) with a mean of 12 . 2% , corresponding to 1 . 95 disomes per cell ( mean number of disomes = 16 × mean frequency of nondisjunction ) . Disome frequency was measured by karyotyping randomly selected single-spore-derived colonies using a pulsed field gel electrophoresis ( PFGE ) technique that allows visual identification of disomes as either two separate bands or a single band twice as bright as that corresponding to a monosome . When disomes are present in only a small proportion of cells in a colony ( because of mitotic loss ) , the sensitivity of detection will depend on whether the interspecific homologs migrate together or as two distinct bands , with the latter case affording much greater sensitivity . Indeed , Hunter and colleagues [4] reported high disome frequencies for the three homologs that migrate as separate bands ( mean for Chr I , II , and VIII = 22 . 3% or 3 . 57 disomes per spore ) compared to six chromosomes that comigrate ( mean for Chr III , VI , IX , X , XI , and XIV = 7 . 1% or 1 . 14 disomes per spore ) . A more sensitive PCR-based approach was used to estimate frequencies of extra copies of all 16 chromosomes in spores derived from the same S . cerevisiae × S . paradoxus cross by Greig and colleagues [15] , who found that S . cerevisiae chromosomes were over-represented in colonies derived from random spores by 31% ( compared to the random-segregation expectation of 33% of viable spores , see below ) . They inferred that this degree of over-representation was evidence of high numbers of disomes ( an average of 4 . 96 disomes per spore ) caused by near-random segregation of homologous chromosomes in hybrid meiosis . Unfortunately , the lack of S . paradoxus genomic sequences at the time of these experiments made it impossible to confirm that the over-representation of S . cerevisiae sequences was actually attributable to the presence of disomes . Instead , Greig and colleagues [15] analysed their hybrid-spore–derived colonies by PFGE—looking only at chromosomes ( I , II , and VIII ) that could be resolved as separate bands—and found similar disome frequencies to the average value of 22 . 3% reported by Hunter and colleagues [4] with an average of 25 . 7% ( or 4 . 10 disomes per spore ) . Thus , PFGE results are consistent with a relatively low overall disome frequency when disomes are scored based on both the number and intensity of bands [4] but also with near-random segregation when only disomes migrating as two distinct bands are counted [15] . More recently , Kao and colleagues [16] used a comparative genome hybridization assay on random-viable-spore–derived colonies and detected an average of 2 . 29 disomes per spore ( 14 . 3% ) , but the sensitivity of their assay is difficult to assess . Nondisjunction frequencies are further biased by the inability to score dead spores . If a chromosome pair segregates randomly , half the time it will segregate correctly and half the time incorrectly . Therefore , under random segregation of a chromosome , 50% of spores should exhibit the consequences of nondisjunction: 25% will contain two copies of the chromosome ( disomes ) and 25% will contain no copies of the chromosome ( nullosomes ) [6] . However , since the 25% of spores that fail to inherit at least one copy cannot be scored , one third of observed spores should contain disomes and two thirds should contain monosomes , resulting in a maximum observed nondisjunction frequency of only 33% . Additionally , any disome ( such as Chr VI [10] ) or combination of disomes that reduces the likelihood of germination will be underrepresented in analyses of viable-spore–derived colonies . However , apart from Chr VI , there is little evidence that the presence of disomes or combinations of disomes has large effects on the likelihood of spore germination [8 , 9] . Mitotic disome loss , insensitive disome detection , and the inability to score dead spores mean that karyotype data from the colonies produced by viable spores can only be used to determine a lower bound for nondisjunction rates , preventing the causes of hybrid sterility from being quantified . Accurate assessment of nondisjunction rates is therefore crucial to understanding the contribution of meiotic mis-segregation to postzygotic reproductive isolation in yeast . Under the simple assumption that spores inheriting zero copies of any chromosome are inviable , but disomes do not affect the likelihood of germination , random segregation of homologous chromosomes is sufficient to explain all of the observed inviability of S . cerevisiae × S . paradoxus hybrid spores: the probability of inheriting at least one copy of each chromosome would be 0 . 7516 or 1 . 0% , equivalent to the measured viability of spores produced by these interspecific hybrids [4 , 17] . Consequently , the high rate of mis-segregation reported by Greig and colleagues [15] would explain nearly all of the infertility observed in interspecific hybrids ( 0 . 76316 = 1 . 3% viability ) . The lower rate reported by Hunter and colleagues [4] would allow much higher spore viability ( 0 . 89116 = 15 . 9% viability ) than observed , meaning that other mechanisms such as toxic disomy or additional forms of hybrid incompatibility must be invoked to explain the low fertility of interspecific hybrids . Accurate assessment requires a sensitive assay that measures the frequency of disomes in spores directly ( preferably in tetrads themselves ) and independently of spore viability . Here , we have used a spore-autonomous fluorescent protein expression assay developed by Thacker and colleagues [18] based on a similar system developed in Arabidopsis [19] to accurately measure meiotic mis-segregation in yeast hybrids .
The spore-autonomous expression system developed by Thacker and colleagues [18] ( received as a kind gift from Scott Keeney ) expresses a fluorescent protein ( GFP , mCerulean , or tdTomato ) under the control of the spore-autonomous YKL050c promoter and the PGK1 terminator . Each fluorescent protein is paired with a promoter/terminator combination from a different Saccharomyces species ( GFP: S . mikatae; mCerulean: S . bayanus; tdTomato: S . kudriavzevii ) . Unfortunately , we found that although these constructs resulted in spore-autonomous expression when integrated into the genome of the S . cerevisiae strain used here ( Y55 ) , when integrated into the genome of our S . paradoxus strain ( N17 ) , they resulted in very weak expression in both spores and vegetative diploids . Swapping the heterologous YKL050c promoter for the species-specific version resulted in some improvement ( see S1 Text and S1 Fig ) , but expression in S . paradoxus and hybrid spores remained weak . We therefore swapped the YKL050c promoter for a different spore-autonomous promoter: the DIT1 ( YDR403w ) promoter [20] . The DIT1 gene is expressed exclusively in sporulating cells , with transcripts first appearing 8–10 hours after transfer to sporulation medium and reaching maximal levels after 12–14 hours , corresponding to the time of prospore enclosure . Expression of both GFP and tdTomato under a strain-specific DIT1 promoter resulted in strong spore-specific and spore-autonomous fluorescence in S . cerevisiae , S . paradoxus , and F1 hybrids . Fluorescent protein expression was placed under control of the DIT1 promoter by replacing the endogenous DIT1 ORF in the desired S . cerevisiae or S . paradoxus haploid ( S4 Fig ) with the entire construct developed by Thacker and colleagues [18] except for the YKL050c promoter ( i . e . , from the start codon of GFP or tdTomato to the end of the URA3 or LEU2 cassette , respectively ) . These constructs , PDIT1_GFP_URA3 and PDIT1_RFP_LEU2 , were then integrated into selected loci on each chromosome to allow analysis of segregation of all chromosome pairs ( S4 Fig ) . Full details of strain construction are provided in the S1 Text . Sporulation was performed as described by Thacker and colleagues [18] . Briefly , prototrophic diploid S . cerevisiae , S . paradoxus , or ( inter- or intraspecific ) F1 hybrids were inoculated into 5 mL of liquid 1% YPA presporulation medium ( 1% Difco yeast extract , 2% Bacto peptone , 1% potassium acetate ) and grown overnight at 30°C . Cells in 1 mL of each overnight culture were harvested by centrifugation , washed in water , resuspended in 2 mL liquid sporulation medium ( 2% potassium acetate ) , and incubated with shaking for 24 h . Slides were prepared by mounting 10 μL of sporulated cultures under a coverslip , sealing with nail polish , and squashing gently to flatten tetrads . We only scored tetrads for which all four spores were present , clearly visible , and easily distinguishable from those of neighbouring tetrads . Fluorescence was manually scored in tetrads using either a Zeiss Axio Scope A . 1 ( Zeiss , Jena , Germany ) equipped with Filter Sets 38HE ( GFP ) and 43HE ( tdTomato ) and a Zeiss EC Plan-NEOFLUAR 100× oil immersion objective or a Zeiss Axio Imager M1 equipped with Filter Sets 38 ( GFP ) and 20 ( tdTomato ) and a Zeiss EC Plan-NEOFLUAR 40×/0 . 75 Ph2 objective . Each tetrad was identified under visible light , scored for GFP and then tdTomato ( or vice versa ) , and then double-checked . The vast majority of tetrads either had 2 green spores and 2 red spores ( correct segregation ) or 2 nonfluorescent spores and 2 spores that were both green and red ( meiosis I nondisjunction , Fig 1 ) . A total of 4 . 0% ( 2 , 062/51 , 112 ) of tetrads exhibited other fluorescence patterns . Of these , the most frequent patterns were 3 spores of one colour and 1 of the other ( likely a gene conversion event between the fluorescent markers ) or 2 spores of one colour , 1 of the other , and 1 nonfluorescent spore ( likely a meiosis II nondisjunction event ) . Because these patterns were not consistent with meiosis I nondisjunction , they were included in the total tetrad counts as having segregated correctly . Our estimates of the total nondisjunction frequency may therefore be slightly conservative .
We observed extremely high levels of chromosomal nondisjunction during meiosis in interspecific hybrids formed between S . cerevisiae strain Y55 and S . paradoxus strain N17 ( Fig 2 ) . On average , each chromosome pair failed to segregate during meiosis I in 40 . 3% of hybrid sporulation events , with a total of 6 , 276 out of 15 , 588 tetrads exhibiting nondisjunction . If spores containing at least one copy of each chromosome are viable , the observed nondisjunction rates would result in a spore viability of 2 . 7% , comparable to approximately 1% observed experimentally [4 , 17] . Thus , the loss of chromosomes due to meiosis I nondisjunction can explain nearly all of the infertility in hybrids between these two species . Nondisjunction was extremely rare in the nonhybrid parental strains , occurring in only 0 . 15% ( 7/4 , 527 ) of tetrads in the S . cerevisiae Y55 parent and 0 . 06% ( 3/5 , 145 ) of tetrads in the S . paradoxus N17 parent . The observed parental nondisjunction frequencies are similar to previously reported values [18 , 21] , indicating our genetic manipulations have not inflated meiotic mis-segregation . We observed similar levels of nondisjunction to those reported above using species-specific YKL050c promoters across five tested chromosomes ( 43 . 8% in N17 × Y55 hybrids compared to 41 . 0% for the same five chromosomes using the DIT1 promoter ) , although this system was more difficult to score ( S1 Fig ) . Most chromosomes exhibited similar levels of mis-segregation , and no relationship was observed between the rate of nondisjunction and chromosome length ( Fig 2: Spearman rank correlation rs = −0 . 085 , P = 0 . 754 ) . One marked exception was Chr XII , which segregated much more reliably than did any other chromosome ( nondisjunction rate of only 31 . 6% ) . To confirm the generality of this lower rate of mis-segregation of Chr XII in interspecific hybrids , we examined the segregation of Chr I , VII , and XII in crosses between S . cerevisiae strain Y55 and the additional S . paradoxus strains YPS138 and N44 and in a cross between S . paradoxus strain N17 and the additional S . cerevisiae strain S288C ( Fig 3 ) . We found that Chr XII had lower rates of nondisjunction than did Chr I and VII in all interspecific hybrids tested . Chr XII contains a single long ribosomal DNA ( rDNA ) tract in Saccharomyces yeasts consisting of an uninterrupted stretch of 150–200 identical or nearly identical approximately 9 , 100 base pair head-to-tail tandem repeats [23] . Recombination is very frequent in the rDNA of S . cerevisiae [24] , and this region has been shown to promote mitotic chromosome segregation fidelity [25] . It is intriguing to speculate that the large size and repetitive nature of the rDNA may promote crossovers between hybrid sister chromosomes , ensuring correct segregation during meiosis . The 18S , 5S , and 25S ribosomal-RNA–encoding regions , but not the intergenic spacers , are almost perfectly conserved between S . cerevisiae and S . paradoxus [26–28] , providing an excellent template for double-strand–break repair . If anti-recombination is responsible for the high rates of nondisjunction in yeast hybrids , it seems reasonable that the long tracts of identity associated with rDNA would help to rescue segregation on Chr XII as shown in Fig 3 . The failure of homologous chromosomes to segregate correctly during meiosis I in hybrids is often attributed to the activity of mismatch repair and anti-recombination systems that prevent recombination between divergent DNA sequences [4 , 5 , 21] . We used our spore-autonomous fluorescent markers to compare the recombination rate between two loci on Chr XI in S . cerevisiae strain Y55 , S . paradoxus strain N17 , and their interspecific hybrid as described by Thacker and colleagues [18] . We calculated the map distance between YKR005c and YKL050c , located approximately 100 kb apart on Chr XI , to be 36 . 1 cM in Y55 and 32 . 9 cM in N17 but only 0 . 4 cM in their hybrid ( S3 Fig ) . Thus , at least in this region of the genome , the recombination rate in the interspecific hybrids is only 1% of that in the parents . Given meiosis in S . cerevisiae typically involves 90 crossovers [29] , extrapolating our observation across the genome suggests that interspecific hybrids experience only about 1 crossover per meiosis—a value comparable to that of 2 . 7 crossovers per viable spore found in hybrids of the same two species [16] . Intraspecific hybrids between closely related parents have previously been shown to suffer from reduced recombination as well . In a genome-wide screen , Martini and colleagues [30] found that hybrids between S . cerevisiae strains S288C and SK1 ( a very close relative of Y55 and about 0 . 7% diverged from S288C ) underwent an average of only 73 crossovers per meiosis , but disrupting the anti-recombination gene MSH2 restored the number of crossovers to the level seen in the parents ( >90 ) . To further investigate the consequences of impaired anti-recombination on meiotic nondisjunction in our interspecific hybrids , we used strains in which expression of the RecQ helicase SGS1 was placed under the control of the CLB2 promoter , which is strongly repressed during meiosis [31] . Sgs1 is required for rejection of homeologous recombination [32] , and the pCLB2_SGS1 construct has previously been found to reduce nondisjunction in partial hybrids between S . cerevisiae and S . paradoxus [33] . Consistent with these results , we found substantially reduced nondisjunction of four tested chromosomes ( II , VII , XII , and XIII ) in interspecific hybrids with impaired anti-recombination relative to wild-type controls ( Fig 4 ) . Chr XII exhibited the largest response to impairing anti-recombination in hybrids , with a 3 . 2-fold improvement in segregation compared to approximately 2-fold for the others . Parental strains exhibited extremely low levels of nondisjunction whether anti-recombination was impaired or intact . Our results add to the mounting evidence that anti-recombination is largely responsible for the sterility of these interspecific hybrids . Indeed , recent work has demonstrated that disabling the activity of multiple components of the anti-recombination machinery during meiosis results in a large improvement in S . cerevisiae × S . paradoxus hybrid spore viability ( from <1% to >30% ) associated with a decrease in disome frequency and an increase in crossover frequency of similar magnitudes [34] . The relevance of meiotic nondisjunction of homologous chromosomes in S . cerevisiae × S . paradoxus hybrids to models of speciation has been questioned because of the high genetic divergence between these species ( 12 . 2% sequence divergence ) ; many reproductively isolated species are much less diverged [1 , 35] . Despite these objections , deletion of the mismatch repair gene MSH2 has been found to improve the fertility—relative to their parents—of intraspecific hybrids formed by crossing different strains of S . paradoxus [21] . Indeed , very small numbers of mismatches can greatly reduce the rate of recombination in Saccharomyces: a single mismatch in a 350-base pair sequence was found to reduce homologous recombination 3- to 4-fold , and two to three additional mismatches reduced the frequency of recombination 9-fold relative to identical sequences [36] . Mismatch-repair–controlled anti-recombination is thought to cause the crossover frequency to fall exponentially with sequence divergence , and the rate of decline is strongest for near-identical sequences [36] . Consequently , meiosis I nondisjunction should be observable in hybrids formed between closely related populations . To determine the importance of meiotic nondisjunction to the fertility of intraspecific hybrids , we examined segregation in the spores of a diploid formed by crossing S . paradoxus strains from the European and Far Eastern clades [37]: N17 and N44 ( 1 . 4% sequence divergence ) . We found this cross produced 63 . 9% viable spores ( 253 viable out of 396 total spores ) , slightly lower than the value of 77 . 1% reported in [38] for a similar cross ( CBS432 × N44 ) . The average nondisjunction rate in intraspecific hybrids was much lower than observed in interspecific hybrids , with only 3 . 4% of homologous pairs failing to segregate ( Fig 5 ) . If , as above , we assume that all spores inheriting at least one copy of each chromosome are viable , then the observed levels of nondisjunction would result in 75 . 7% spore viability . Therefore , as observed in interspecific hybrids , nondisjunction can explain most of the infertility in hybrids formed between much more closely related strains ( S2 Fig ) . We confirmed the occurrence of nondisjunction in intraspecific hybrids by examining segregation of Chr I in a second S . paradoxus cross ( N17 × YPS138 , 3 . 5% genetic divergence ) , which showed a similar nondisjunction frequency ( 10 . 8% ) to that seen in the N17 × N44 hybrids ( 10 . 7% , S1 Data , S2D Fig ) . Even a cross between two very closely related S . cerevisiae strains ( Y55 × S288C , 0 . 6% genetic divergence ) exhibited detectable levels of Chr I nondisjunction ( 1 . 1%; none was observed for the Y55 parent ) . In contrast to the roughly uniform nondisjunction rates for different chromosomes in the interspecific hybrid , we observed a strong negative correlation between chromosome length and the rate of nondisjunction in the intraspecific hybrid ( Spearman rank correlation rs = −0 . 860 , P < 0 . 0001 ) ; the relationship is log–log linear with a slope of −1 . 79 ( Fig 5 ) . This negative relationship was expected since crossover frequency is tightly associated with chromosome size in S . cerevisiae intraspecific hybrids ( S288C × YJM789 , SNP-based genomic distance 99 . 54% identical [37] ) in which the number of crossovers per chromosome increases by approximately 0 . 6 per 100 kb of chromosome length [29 , 39] . If longer chromosomes are more likely to experience at least one crossover event , then they should exhibit lower rates of nondisjunction . Indeed , we found that only Chr I and VI exceeded a nondisjunction frequency of 5% in the intraspecific hybrid; these two chromosomes also have the highest probabilities of exhibiting zero crossovers in homozygous parents [29] . We cannot explain why the negative relationship between nondisjunction and chromosome length is so clear in our analysis of an intraspecific hybrid ( Fig 5 ) but undetectable in our analysis of an interspecific hybrid ( Fig 2 ) . Given the very low frequency of recombination in interspecific hybrids , it is possible that our analysis simply lacked the power to detect a weak negative correlation . Alternatively , it is also possible that rare crossovers in these hybrids are more likely to occur on longer chromosomes but that this is balanced by a bias in crossover-independent mechanisms towards rescuing segregation of shorter nonexchange chromosomes [29] . The levels of nondisjunction we observed in intraspecific and interspecific meiosis are likely attributable to sequence divergence between the two parents . A strong negative relationship between S . paradoxus hybrid spore viabilities and parental genetic distances was reported by Liti and colleagues [38] . After correction for inviability associated with chromosomal rearrangements , they found this relationship was best fit by an exponential decay function . In contrast , examining hybrids formed between strains of S . cerevisiae—a less genetically diverse group than S . paradoxus—Hou and colleagues [40] reported no correlation between hybrid spore viabilities and parental genetic distances . However , if the spore viabilities measured by Hou and colleagues [40] are adjusted to account for the effects of chromosomal rearrangements as done by Liti and colleagues [38] , a strong negative correlation immediately becomes apparent ( S2 Fig ) . Indeed , the relationship is similar to that reported for S . paradoxus hybrids ( and consistent with the results presented here for nondisjunction in the N17 × N44 hybrid ) : spore viability decreases by roughly 1 . 25% for each 0 . 1% parental sequence divergence . We further show a similar negative relationship in a separate S . cerevisiae data set collected for the 100-genomes strains ( S2 Fig , [41] ) . It therefore seems entirely plausible that small amounts of sequence divergence can contribute to reproductive isolation and incipient speciation . We have shown that meiosis I nondisjunction alone can explain nearly all of the sterility of the yeast hybrids studied here , simply because of spores failing to inherit essential chromosomes . If the unexplained spore inviability can be attributed to disomy arising from nondisjunction , then no other mechanism would be necessary to explain yeast hybrid sterility . The magnitude of the effect of meiotic mis-segregation on hybrid sterility has previously been hard to gauge because of technical difficulties in accurately quantifying nondisjunction rates . Moreover , since previous evidence for the association between nondisjunction and postzygotic reproductive isolation was restricted to crosses between highly diverged species ( S . cerevisiae and S . paradoxus ) , a major role of sequence divergence as a cause of hybrid sterility between closely related populations has often been dismissed [1 , 35 , 42 , 43] . Here , we show that nondisjunction can explain most of the hybrid fertility defect observed not only for interspecific crosses but also for an intraspecific cross between much more closely related strains . These results suggest that even small amounts of sequence divergence can directly contribute to postzygotic reproductive isolation , consistent with the negative correlation between hybrid spore viability and parental sequence divergence observed both within and between species ( S2 Fig , [38 , 44 , 45] ) . We do not mean to suggest that anti-recombination is the sole mechanism underlying postzygotic reproductive isolation in yeast . Chromosomal rearrangements can clearly contribute to the genetic isolation of Saccharomyces populations [40 , 46] , and we have intentionally chosen collinear strains to remove these effects . Furthermore , although there is no evidence for lethal genic incompatibilities between the genomes of S . cerevisiae and S . paradoxus [16 , 47] , negative epistasis does reduce hybrid viability between these two species in particular environments [3 , 48] . It is also possible that either the number of spores per tetrad [49] or the efficiency with which diploids sporulate might be affected by negative epistasis between hybrid genomes [50 , 51] , potentially reducing the total number of hybrid gametes produced rather than their viability . Similarly , strong cytonuclear incompatibilities , which reduce hybrid viability and can prevent later generation hybrids from even entering meiosis , have been observed in crosses between species of Saccharomyces yeasts [52] . However , growth and sporulation rates are highly dependent on environmental conditions , and thus the contribution of these factors to reproductive isolation between yeast species is difficult to assess because they cannot be directly observed in nature . Chromosomal mechanisms of speciation ( including both rearrangements and sequence divergence ) are often dismissed as unimportant because they generate underdominance [53] . Mutations that result in nondisjunction in heterozygotes ( or hybrids ) will be initially deleterious since rare mutants will predominantly mate with wild types . Consequently , these mutants should not rise to the frequency necessary to establish a new breeding population in which correct segregation occurs . We agree that underdominance reduces the likelihood of chromosomal mechanisms of speciation acting in obligately sexual organisms , which includes virtually all animals . However , underdominance is not necessarily a problem in facultatively sexual organisms , like yeast , in which a single mutant can rapidly establish a highly inbred population by clonal propagation . Although most speciation research is focused on obligate sexuality , facultative sexuality should not be ignored; many plants , most fungi , and nearly all unicellular eukaryotes are facultatively sexual , and this mode of reproduction is considered the ancestral state of all eukaryotes [54] . Clearly , many mechanisms can contribute to postzygotic reproductive isolation , and how these mechanisms interact to drive speciation in yeast remains an open question [55 , 56] . Nevertheless , the importance—to speciation in facultatively sexual organisms , at least—of small amounts of sequence divergence should not be overlooked .
|
Different species are kept genetically separated from each other by reproductive isolation , which can result in the production of sterile hybrids . Despite the central role of reproductive isolation in evolutionary biology , the mechanisms underlying hybrid sterility remain controversial . We focus on one potential cause: genome-wide sequence divergence . Sequence divergence can lead to hybrid sterility by interfering with the proper segregation of chromosomes during gamete production , leading to most gametes lacking at least one essential chromosome . While hybrid sterility is known to be associated with mis-segregation in Saccharomyces yeasts , quantification of the frequency of these errors has been elusive until now . We use the recently developed technique of spore-autonomous fluorescent protein expression to quantify chromosome segregation errors in gametes produced by both interspecific and intraspecific yeast hybrids . By doing so , we show that segregation is near random in interspecific hybrids , accounting for nearly all of the observed gamete inviability . We also demonstrate that sequence divergence is associated with sterility of hybrids both between distantly related species and between different strains of the same species , providing a general mechanism for reproductive isolation in yeast .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"and",
"discussion"
] |
[
"fungal",
"spores",
"meiosis",
"homologous",
"chromosomes",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"fungi",
"protein",
"expression",
"model",
"organisms",
"experimental",
"organism",
"systems",
"molecular",
"biology",
"techniques",
"dna",
"fungal",
"reproduction",
"homologous",
"recombination",
"saccharomyces",
"research",
"and",
"analysis",
"methods",
"mycology",
"animal",
"studies",
"chromosome",
"biology",
"autosomes",
"chromosome",
"pairs",
"molecular",
"biology",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"yeast",
"gene",
"expression",
"and",
"vector",
"techniques",
"biochemistry",
"eukaryota",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"dna",
"recombination",
"organisms",
"chromosomes"
] |
2018
|
Spore-autonomous fluorescent protein expression identifies meiotic chromosome mis-segregation as the principal cause of hybrid sterility in yeast
|
Point process generalized linear models ( PP-GLMs ) provide an important statistical framework for modeling spiking activity in single-neurons and neuronal networks . Stochastic stability is essential when sampling from these models , as done in computational neuroscience to analyze statistical properties of neuronal dynamics and in neuro-engineering to implement closed-loop applications . Here we show , however , that despite passing common goodness-of-fit tests , PP-GLMs estimated from data are often unstable , leading to divergent firing rates . The inclusion of absolute refractory periods is not a satisfactory solution since the activity then typically settles into unphysiological rates . To address these issues , we derive a framework for determining the existence and stability of fixed points of the expected conditional intensity function ( CIF ) for general PP-GLMs . Specifically , in nonlinear Hawkes PP-GLMs , the CIF is expressed as a function of the previous spike history and exogenous inputs . We use a mean-field quasi-renewal ( QR ) approximation that decomposes spike history effects into the contribution of the last spike and an average of the CIF over all spike histories prior to the last spike . Fixed points for stationary rates are derived as self-consistent solutions of integral equations . Bifurcation analysis and the number of fixed points predict that the original models can show stable , divergent , and metastable ( fragile ) dynamics . For fragile models , fluctuations of the single-neuron dynamics predict expected divergence times after which rates approach unphysiologically high values . This metric can be used to estimate the probability of rates to remain physiological for given time periods , e . g . , for simulation purposes . We demonstrate the use of the stability framework using simulated single-neuron examples and neurophysiological recordings . Finally , we show how to adapt PP-GLM estimation procedures to guarantee model stability . Overall , our results provide a stability framework for data-driven PP-GLMs and shed new light on the stochastic dynamics of state-of-the-art statistical models of neuronal spiking activity .
Point-process generalized linear models ( PP-GLMs ) have become an important approach in the statistical modeling of neurophysiological responses from single nerve cells and their interactions in neural circuits [1–8] . A specific class of PP-GLMs are nonlinear Hawkes processes [9 , 10] . In this case , each action potential ( spike ) modulates the firing intensity of the neurons in the future . Nonlinear Hawkes PP-GLMs can capture the major canonical dynamics of single neurons [11–13] and as phenomenological models avoid the many issues that arise in the specification of biophysically detailed neuronal models [14 , 15] . In this way , nonlinear Hawkes PP-GLMs are also important phenomenological models for the simulation and study of large-scale neuronal network models of brain function . However , nonlinear Hawkes PP-GLMs also lead to non-renewal point process spike train statistics because contributions to the intensity from many previous spikes can accumulate over arbitrary time scales [16 , 17] . This raises the question of whether such models will produce stable , stationary dynamics in simulations , or whether firing rates will diverge or settle into unphysiological rates depending on a specified absolute refractory period . For the linear Hawkes point process model , stability can be assessed by calculating the integral of the spike-history kernel , i . e . , the effect that each spike has on subsequent activity of the same cell [18–20] . However , for the prevalent nonlinear case , no such practical criterion is currently available . Main stability results established by Brémaud and Massoulié are too restrictive for our applications [9 , 21] . Furthermore , model parameters are typically estimated from data using maximum-likelihood methods [2] . For linear autoregressive processes , it is well known that maximum-likelihood estimates can lead to unstable dynamics [22 , 23] . We expect this to be even more severe in nonlinear models . Here , we first show that PP-GLMs estimated from physiological data might not generate spike train realizations that match even simple statistics such as mean firing rates of the original data . Instead , firing rates tend to diverge to the maximum firing rate that is allowed in the presence of an absolute refractory period . Firing patterns like this would typically be considered unphysiological . This can happen in spite of the models passing commonly used goodness-of-fit tests based on , for example , the time-rescaling theorem [24 , 25] . To address the above stability issues , we propose an approximative framework to derive stochastic stability conditions for PP-GLMs . For a stochastically stable point process , the state of the point process stochastically evolves in time , but can be described by a stable and time-invariant distribution of “states” ( sample paths ) , resulting in a stationary point process [9] . In contrast , a process that is not stochastically stable may show similar stochastic dynamics for some time , but eventually its state may diverge and never return . In neural point processes , this scenario is typically associated with a divergence of the firing rate . In case of an actual divergence there is no stationary distribution of states . Note , however , that when considering absolute refractory periods , the divergence of the firing rate and associated internal states of the point process are limited . Firing with an inter-event interval equal to the refractory period is nonphysiological , and the state distribution in this mode of firing is singular . We call this a diverged state in slight deviance from the usual terminology . Our use of the term stochastic stability , however , is in line with the definition from stochastic dynamical systems , but generalized to stochastic point processes ( see also [9 Remark 4] ) . Our approach to derive stability conditions for PP-GLMs is based on a recently developed mean-field theory of neural dynamics [26 , 27] . The approach relies on the following steps . First , we use a quasi-renewal ( QR ) approximation that decomposes spike history effects into the contribution of the last spike and an average of the conditional intensity function ( CIF ) over all spike histories prior to the most recent spike . Second , after truncation of a moment-based expansion , this decomposition leads to a tractable expression for the approximated CIF . Third , under stationarity conditions , fixed points can be derived as self-consistent solutions of an integral equation , which correspond to expected steady-state firing rates of the neuron . Fourth , depending on the number and stability of these fixed points , each single-neuron model can be unambiguously classified into one of three types: stable , divergent , or fragile . The latter corresponds to metastability which results from stochastic fluctuations perturbing the dynamics in the presence of multiple stable fixed points and when the upper fixed point corresponds to a stable but unphysiologically high firing rate . Fluctuations around the low-rate fixed point of the network dynamics predict an expected time horizon until rates will converge to the high-rate state . This expected time metric can be used to estimate the probability of firing rates to remain finite for a given time period . Examination of the stability of the fixed points and how it depends on the shape of the spike-history filter not only determines the stability of the stochastic dynamics , but also leads to general stability constraints on PP-GLM parameters . In the following sections , we present our framework to assess the stability of a specific neuron model in detail . We validate the QR approximation in comparison to results of numerical simulations for a large range of artificial neuron models that are neurophysiologically plausible . We then apply the method to real-world data sets . Finally , we demonstrate and discuss how parameter estimation procedures could be adapted to ensure stability of estimated models . As stated above , stability of neuron models is particularly important when numerical simulations are desired or spike trains are to be generated from the model . Generated spike trains can be used to assess model goodness-of-fit and to perform forecasting of neural activity over longer time scales . Our results are a first step towards ensuring stability for recurrently connected neural network models . These models can be put in the framework of multivariate nonlinear Hawkes models , and our classification framework conceptually translates to the multivariate case . Stable ( or stabilized ) neuron models play an important role in the field of computational neuroscience , especially in the simulation of large-scale models of brain function . They are also important in neuroengineering , where neuron models are embedded in hybrid and closed-loop applications [28 , 29] .
The nonlinear Hawkes process is a point process model that is commonly used to describe neurophysiological responses . It defines the conditional intensity function ( instantaneous firing rate ) of a neuron as a nonlinear function of previous spiking activity ( Fig 1A ) : λ ( t | H t ) = ϕ h ( t ) , ( 1 ) where H t denotes the spiking history up to time t and ϕ ( x ) is a non-negative nonlinear function . The term h ( t ) consists of a constant offset I0 and a convolution of the spike train S ( t ) with ( temporal ) spike-history kernels or filters η ( s ) : h ( t ) = I 0 + [ η * S ] ( t ) = I 0 + ∑ k = 1 K η ( t - t k ) , ( 2 ) where the { t k } ∈ H t correspond to the previous spike times ( see “Materials and Methods” for details ) . Based on both theoretical and empirical arguments [2 , 7] , we set ϕ ( x ) = exp ( x ) to arrive at: λ ( t | H t ) = exp I 0 + ∑ k = 1 K η ( t - t k ) = c exp ∑ k = 1 K η ( t - t k ) , ( 3 ) with c = e I 0 > 0 . Every previous spike contributes a spike-history kernel , and effects of all previous spikes accumulate . This leads to , in general , a non-renewal point process model . The model parameters that describe the kernel η ( s ) and the baseline firing rate c can be estimated using maximum-likelihood optimization within the framework of generalized linear models ( GLMs ) [2 , 4] . As stated earlier , these point-process GLMs ( PP-GLMs ) were recently shown to be able to describe all major canonical dynamics of single neurons [11–13] and , thus , can serve as a canonical class of mathematically tractable models to describe general single-neuron spiking activity . For example , in [6] , we analyzed spiking data from the stomatogastric nervous system of the crab . Neurons that are part of the pyloric network fire in stereotypical , rhythmic activity patterns ( Fig 2A ) . Estimated PP-GLMs from this physiological dataset pass common goodness-of-fit tests such as based on residual analysis or the time-rescaling theorem . We created stochastic realizations of spike trains based on the model . These spike trains reproduce the observed burst pattern of the training data , and in a complete network simulation , the relative phases of the overall pyloric rhythm [6] . However , such simulations of spike trains from data-driven PP-GLMs do not always result in physiological spiking patterns . For example , when estimating PP-GLMs from single-unit data recorded from the neocortex of human epileptic patients [30] , goodness-of-fit tests are generally passed , but simulated activity tends to diverge to unphysiologically high firing rates . One data set and estimated model are shown in Fig 2B . In a more comprehensive analysis , we fitted PP-GLMs to spiking data from neurons recorded from motor-related cortical areas in the monkey [31] . We restricted data to a one-second steady-state movement preparation period of the trial . This period was roughly stationary since it did not include , by design , firing rate transients driven by sensory stimuli or movement execution . For 35 out of the 99 data-driven models , we find that simulated spike trains have finite divergence times . Some of these models diverge in simulations even when goodness-of-fit tests are passed . We show two examples in Fig 2C and 2D . Qualitatively similar results are obtained for all other models . Instability is also observed when the simulation is performed using a nonlinearity that grows less rapidly than the exponential . We generated stochastic realizations of spike trains using two additional nonlinearities ( a linear rectifier , and f ( x ) = log ( 1 + ex ) , a smooth interpolation between an exponential and linear function , S1 Fig ) . Both functions are globally dominated by the exponential function ( see S1A Fig ) . Potentially , nonstable behavior could be observed when simulating with the exponential nonlinearity but not with the two less rapidly accelerating nonlinearities . However , we find that even in these cases , firing rates diverged for the same data sets as presented in Fig 2B and 2C . In summary , while PP-GLMs estimated from data may serve well in encoding and decoding analysis that require one-step spike prediction conditioned on actually observed spike history and may pass goodness-of-fit tests , they tend to be poor generative models because of the lack of stochastic stability . The use of PP-GLMs as generative models , however , is essential when statistical analyses of spike trains generated by the model are required , or when long-term prediction of future spiking states in single-neuron and neuronal networks is used in neural decoding or closed-loop interventions . To our knowledge , the stability of PP-GLMs estimated from data has not been systematically examined before . In the next sections , we will develop a framework to assess the dynamics and stability of stochastic spiking neuron models . PP-GLMs have post-spike filters which typically make the spike train probability depend on many previous spikes . These dynamics are in contrast to the conditional intensity function of a renewal point process which depends only on the very last spike time . Therefore , PP-GLMs are generally not renewal processes . When assessing dynamics and stability of PP-GLMs , we are interested in the behavior of the corresponding firing rates . For such general PP-GLMs , however , there are no closed-form solutions for even simple statistical features , such as expected mean firing rates or second-order statistics . Here , to obtain estimates of such statistical features for a given nonlinear Hawkes process , we employ an approximation based on a recently introduced quasi-renewal approximation [26 , 27] . The quasi-renewal approximation ( Fig 1B ) consists of approximating the ( non-renewal ) PP-GLM by a process which is nearly a renewal-process ( hence , “quasi-renewal” ) that depends on the last spike time and on the average firing rate in the past [26] . Specifically , we consider the steady-state conditional intensity λ 0 ( t , t ^ ) at time t as the average intensity over all possible spike histories that share the most recent spike at time t ^: λ 0 ( t , t ^ ) = 〈 λ ( t | H t ) 〉 S ( t < t ^ ) = c exp ( η ( t − t ^ ) ) 〈 exp ( [ η * S ] ( t ) ) 〉 S ( t < t ^ ) . ( 4 ) The first term explicitly models the effect of the most recent spike only , and the second term represents the average of the spiking activity prior to the time of the last spike in the steady-state regime . It can be approximated by ( see “Materials and Methods” for details ) : 〈 exp ( [ η * S ] ( t ) ) 〉 S ( t < t ^ ) ≈ exp A 0 ∫ t - t ^ ∞ ( e η ( u ) - 1 ︸ γ ( u ) ) d u , ( 5 ) with γ ( u ) = eη ( u ) − 1 for the exponentiated spike-history kernel . Here , A0 is the steady-state firing rate of the process . Intuitively , the convolution of the actual spike train S ( t ) with η ( s ) is replaced by the convolution of a homogeneous Poisson process spike history of intensity A ( t − u ) ≡ A0 with an effective filter γ ( u ) ( Fig 1B ) . Since the convolution is applied to a constant A0 , the term reduces to a product of A0 and the integral of γ ( u ) with the lower bound dependent on τ = t - t ^ . Using this approximation , we obtain a quasi-renewal CIF by combining Eqs ( 4 ) and ( 5 ) . Given the QR-CIF , Eq ( 4 ) , we can then as for ordinary renewal processes , derive the steady-state survivor function S 0 as: S 0 ( τ ) = exp - ∫ 0 τ λ 0 ( u ) d u . ( 6 ) S 0 then yields the steady-state probability density P0 of the inter-spike intervals: P 0 ( τ ) = S 0 ( τ ) λ 0 ( τ ) . ( 7 ) The inverse of the expected inter-spike interval must equal the firing rate f which thus is an implicit function of A0 through Eq ( 5 ) : f ( A 0 ) = ∫ 0 ∞ τ P 0 ( τ ) d τ - 1 . ( 8 ) In effect , the QR theory derives a transfer function f ( A0 ) > 0 that links an assumed average spike history to a predicted firing rate . Assuming stationarity , f ( A0 ) has to match A0 which leads to a fixed-point equation . Intersections of f ( A0 ) with the identity correspond to expected fixed points of the dynamics . Stable fixed points in the quasi-renewal approximation predict steady-state firing rates of nonlinear Hawkes processes ( PP-GLMs ) . We first show how the number and stability of fixed points of the derived transfer function f ( A0 ) for the nonlinear Hawkes process endowed with an absolute refractory period can be used to classify the dynamical behavior of the single-neuron model ( Fig 3 ) : Overall , the above classification of the qualitative stochastic dynamics suggests a general framework to assess stability and dynamics of stochastic spiking neuron models ( Fig 4 ) . In the particular case of data-driven models , training data are used to estimate parameters of a nonlinear Hawkes model ( PP-GLM ) through ( regularized ) maximum-likelihood optimization . As shown above , we find empirically that simulating spike trains from these models often yields unphysiological spiking patterns , and firing rates may diverge ( Figs 2 and 4 , top ) . We can use the quasi-renewal approximation to analyze the stability of the estimated neuron model . This approximation predicts the dynamics of the neuron model and distinguishes three qualitatively different dynamical behaviors . Stable or fragile models with high expected divergence times can be safely used to generate stochastic samples from the model . For divergent models or fragile models with low expected divergence times , stabilization constraints can be added to the maximum-likelihood optimization problem to constrain the feasible parameter space to non-divergent models . In any of the three cases , the model ( or its stabilized variant ) is evaluated based on standard model selection and goodness-of-fit tests before any inference is made ( Fig 4 , bottom ) . In the next sections , we present the application of this framework to simulated and real data to demonstrate its validity and utility in modeling electrophysiological responses . First , we demonstrate the proposed method for PP-GLMs that have a spike-history filter that consists of either a single exponential or a sum of two exponentials before moving on to filters estimated from neurophysiological data . We start with the analysis of a simple PP-GLM with a spike-history filter given by a single exponential and an absolute refractory period . The complete model is given by: λ ( t | H t ) = c e [ η * S ] ( t ) , η ( s ) = J θ ( s ) e - s / τ + J ref θ ( s ) θ ( τ ref - s ) , with Heaviside function θ ( x ) and parameters τref = 2 ms , Jref = −∞ ( −1012 in numerical simulations ) , τ = 20 ms , and amplitude J . We scanned the two-dimensional parameter space given by the amplitude J of the filter and the baseline firing rate c ( Fig 5A; −2 ≤ J ≤ 4 with 121 equally spaced samples and 0 . 1 ≤ c ≤ 6 . 0 s−1 with 60 equally-spaced samples ) . The QR approximation predicts three regimes of dynamical activity: For slightly positive and negative kernel amplitude J , the model is stable . Indeed , in simulations , we observe finite and stable rates ( Fig 5B , top row ) . For higher amplitudes , the dynamics are predicted to be fragile and ultimately divergent ( for large J and c ) . As expected , the average divergence time estimated from numerical simulations gradually decreases with increasing J and c ( color-coded in Fig 5A ) . Divergent models are almost instantly diverging ( Fig 5B , bottom row ) , while for fragile models , a whole spectrum of divergence times is observed ( Fig 5B , middle row ) . Within the variance given by the finite number of simulated models , we did not observe any discrepancies between the behavior predicted by the QR approximation and the numerical results . For all models of the parameter range that were classified as stable , we compared the predicted steady-state firing rate to the one observed in numerical simulations ( Fig 6A ) . In this case , the QR approximation provides an excellent prediction of mean firing rates ( Pearson’s correlation coefficient ρ > 0 . 999 ) . A major feature of the QR approximation is to predict ( an upper bound on ) the expected divergence times for fragile models . In practice , this is relevant for model sampling via simulation where it is important to classify fragile models as “effectively stable” or divergent ( see Fig 4 ) . For high firing rates close to A 0 = τ ref - 1 , the regular spike train with inter-spike intervals around τref is the only possible spike train realization . Therefore , one way to estimate E[Tdiv] is to consider periodic spike histories with different frequencies that would lead to self-sustained periodic firing at maximal rate with high probability ( see “Materials and Methods” for details ) . Fig 6B compares predicted versus observed divergence times for all fragile models . The predicted Tdiv provide an upper bound on the observed divergence times . The divergence time of a simulation was defined as the end of the first two-second interval in which the average rate exceeded λthr . For this reason , estimated divergence times cannot be below 2 s ( Fig 6B , gray area ) . Therefore , small estimated divergence times do not obey the predicted bound . However , there seems to be a reasonable ( power-law ) dependence between predicted and observed Tdiv ( Pearson’s correlation coefficient ρ = 0 . 925 ) . We now look at more complex PP-GLMs to test the validity of our proposed framework . We consider spike-history filters consisting of a sum of two exponentials with amplitudes Jr , Ja , and corresponding time constants τr and τa: λ ( t | H t ) = c e [ η * S ] ( t ) , η ( s ) = J r θ ( s ) e - s / τ r + J a θ ( s ) e - s / τ a + J ref θ ( s ) θ ( τ ref - s ) . Depending on the signs of the amplitudes , this model resembles many plausible single-neuron behaviors: Jr < 0 indicates a relative refractory period beyond the absolute 2 ms refractory period while Jr > 0 promotes bursty dynamics . Similarly , Ja ≠ 0 can be interpreted as inhibitory or facilitating adaptation ( e . g . , spike-frequency adaptation [16 , 17 , 33] ) . We evaluated models on a wide range of combinations of amplitudes Jr and Ja ( Fig 7A; −11 ≤ Jr ≤ 11 with 100 equally spaced samples and −3 ≤ Ja ≤ 3 with 75 equally-spaced samples ) , for τr = 20 ms and τa = 100 ms , respectively , and for fixed c = 5 s−1 . As expected , negative and positive but small values of Jr and Ja lead to stable dynamics . For a narrow band , models are expected to be fragile ( Fig 7B , top row ) . This observation is consistent with divergence times estimated from numerical simulations . Finally , larger values of either Ja or Jr lead to divergent models , although in an asymmetric way ( Fig 7A and 7B , bottom row ) . Qualitatively similar results are obtained for other values of the baseline c ( S2 and S3 Figs ) . Estimated divergence times are generally consistent with the qualitative prediction of the QR approximation with one exception: For multiphasic spike-history filters , i . e . either strongly refractory neurons ( Jr ≪ 0 ) with excitatory rebounds ( Ja ≫ 0 ) or the opposite ( Jr ≫ 0 and Ja ≪ 0 ) , the QR approximation predicts divergent models , but simulations indicate that rates remain below the threshold λthr to be classified as divergent ( Fig 7A , upper left and lower right corners; Fig 7B , middle row ) . Spike trains generated from models with these parameters tend to produce intermittent bursts . This is a condition for which the quasi-renewal approximation is known to become invalid [26] . The dynamics lead to the divergent state where the model neuron fires initially at maximally allowed firing rate ( hence , unphysiological ) . However , the dynamics escape this high-firing rate fixed point after a finite number of such high-rate bursts , and the activity reverts back to the low-rate state . Averaged over a longer time period , the mean activity stays well below λthr and therefore , T ^ div = ∞ in contrast to the dynamics predicted by the QR approximation . The reason for this discrepancy is that the QR approximation assumes homogeneous , Poisson-like firing prior to the last spike time , while the only way to achieve firing rates close to the maximally allowed rate by the absolute refractory period is to have a highly regular spike train . For highly regular spike histories , the QR approximation does not provide valid fixed points of the dynamics . However , these cases can be captured by an analysis of the regular spiking limit ( see “Bursting and the regular spiking limit” ) which does not explicitly depend on the QR approximation . A less severe limitation of the QR approximation is visible in the comparison between predicted and simulated steady-state firing rates for models classified as stable ( Fig 8A ) . While most rates are accurately estimated ( points near the diagonal line ) , very bursty neurons have higher firing rates in simulations than predicted by the QR approximation ( stable models with Jr ≫ 0 and Ja ≪ 0 ) . This is due to dependencies beyond the last spike that are ignored in the approximation but are non-negligible for burst firing . Loosely speaking , in this case the QR approximation predicts the rate of isolated spikes and bursts , but not the number of total spikes . The burst duration can be predicted from a simple criterion based on the conditional intensity function of the PP-GLM , independent of the QR approximation ( see “Bursting and the regular spiking limit” ) . Finally , predicted and observed divergence times for all fragile models are well approximated by the QR approximation ( Fig 8B ) except for small divergence times whose estimation is biased due to the finite time window to detect divergence in numerical simulations . In summary , the QR approximation yielded remarkably accurate predictions of the dynamical behavior of PP-GLMs for most parameter settings . When the steady-state rate was not accurately predicted for bursty neurons , the qualitative prediction was still consistent with simulations . For extreme parameter values , we observed intermittent burst activity that was incorrectly predicted to be divergent . Although not divergent according to the definition of Fig 3 , the resulting spiking pattern would nevertheless be considered unphysiological and undesirable in modeling applications . Thus , in this case , the discrepancy between the QR prediction and the simulation does not play a significant role in practice . So far , we have studied parametric spike-history filters in the form of a single exponential or sum of two exponential terms . In the next section , we will show that the validity of the QR approximation extends to physiological PP-GLM spike-history filters as they are typically obtained in the context of data-driven model estimation . We applied the QR approximation to models estimated from actual neuronal recordings , specifically , multi-electrode single-unit recordings in monkey cortex ( see “Materials and Methods” ) . Of the nonlinear Hawkes PP-GLMs estimated from 99 recorded single units , 11 were predicted to be stable , 86 were predicted as “fragile” with varying degrees of expected divergence times , and 2 were predicted to be divergent . In all cases , the predictions were consistent with numerical simulations: For all models predicted to be stable , none of the N = 48 simulations of length T = 1000 s diverged ( T ^ div = ∞ ) , and both divergent models showed finite divergence times ( T ^ div = 86 s and T ^ div = 2 s , respectively ) . Fragile models did not diverge in our simulations of length T = 1000 s in 53 out of the 86 cases , while the other maximum-likelihood models diverged with varying degrees of observed divergence times ( T ^ div = 4 - 45 000 s ) . We examined in detail the stability predictions based on the QR approximation for a divergent neuron model in Fig 9 . Spike-sorted single-unit activity ( Fig 9A ) was used to estimate a nonlinear Hawkes process with ten basis functions for the spike-history filter consisting of raised cosines [4 , 5] . The resulting maximum-likelihood estimate ( MLE ) displays a relative refractory period followed by an excitatory rebound ( Fig 9B ) . The corresponding transfer function of the QR approximation shows a single stable fixed point close to the maximally allowed firing rate ( Fig 9C ) . Therefore , this model is classified as divergent . We then explored the neighboring parameter space by varying the baseline rate parameter c and using scaled versions of the MLE spike-history filter . The qualitative predictions ( separated by thick lines in Fig 9D ) were overall consistent with numerical simulations of the model ( Fig 9D ) . Here , the color scale represents the estimated divergence time in simulations based on 48 independent simulations of T = 1000 s each . In sum , the QR approximation not only predicted correctly the stability of the data-driven neurophysiological models , but also the stability of parameter variations around the MLE model . We have shown that fitting PP-GLMs to electrophysiological data can lead to divergent and fragile models ( Figs 2 and 9 ) . The QR approximation not only offers a way to predict stability , but also to find stable models . As stated before , stability is an important feature when the goal is to sample from the model or to obtain data-driven models for simulations . Conceptually , we can constrain the parameter search for the maximum-likelihood solution to the parameter space for which the QR approximation predicts stable models . We will now provide a proof of concept of this approach by demonstrating how this constraint can be implemented in practice . We search for the maximum-likelihood estimate ( MLE ) under the additional constraint that the model is predicted to be stable by the QR approximation . This can be implemented by minimizing the cost function consisting of the negative log-likelihood of the data plus a penalty term that is infinity whenever the model is predicted to be fragile or divergent and zero otherwise . We use a gradient-free numerical optimization scheme ( see “Materials and Methods” ) . We initialize the parameter values with the unconstrained MLE for which all positive coefficients are set to zero . This corresponds to a non-positive spike-history filter and ensures that the initial evaluation of the cost function is finite . We call the constrained solution the “stabilized MLE” . For illustration , in Fig 10A , we use the same data from monkey electrophysiological recordings as in Fig 9 . We find that the spike-history filter of the stabilized MLE resembles a regularized version of the ( unconstrained ) MLE ( Fig 10A ) , i . e . , its coefficients are slightly biased towards zero . The MLE itself passes a goodness-of-fit test on training and test data and yields a substantial power in predicting spiking in 1 ms time bins ( Fig 10B–10D ) . However , sampling from the MLE leads to divergent and unphysiological spike trains ( Fig 10E ) . In contrast , sampling from the stabilized model yields firing rates comparable to the training data , and spike trains that are qualitatively similar ( Fig 10E ) . We quantified similarity of spike train statistics up to second order using the inter-spike interval ( ISI ) distribution ( Fig 10F ) and spike train auto-correlation function ( Fig 10G ) . We evaluated the spiking pattern using three metrics ( local ISI variability lv , shape ( log κ ) and scale parameter ( log α ) of the best-fitting Gamma distribution , following the methodology presented in [35 , 36] . Our data from monkey region PMv are consistent with reported values in [36 , Table 1 , row 15]: lv ( data ) = 0 . 54 ± 0 . 18 , log κ ( data ) = 0 . 80 ± 0 . 38 , log α ( data ) = 3 . 16 ± 0 . 26 ( mean ± standard deviation over segments of 20 consecutive ISIs ) . Metrics for spike trains generated from the stabilized MLE model are well aligned to those of the physiological data: lv ( stab . MLE ) = 0 . 66 ± 0 . 22 , log κ ( stab . MLE ) = 0 . 69 ± 0 . 35 , and log α ( stab . MLE ) = 3 . 26 ± 0 . 26 . In addition , the spike trains generated from the stabilized MLE also reproduce the serial correlations between consecutive ISIs in the physiological data ( Fig 10H ) . Although this was not a direct optimization criterion , the stabilized MLE is almost as good in predicting spiking activity in 1 ms time bins conditioned on observed spiking history as the MLE ( Fig 10B ) and scores only marginally worse on the goodness-of-fit tests ( Fig 10D ) . The cross-validated log-likelihood of the stabilized model is within 80% of the log-likelihood score of the MLE ( 2 . 43 bits/s versus 2 . 90 bits/s relative to the prediction of a homogeneous Poisson process with correct spiking rate , Fig 10C ) . Spike train statistics were not shown for the MLE because physiological spike trains could not be obtained . However , we explored the possibility of a simple modification in the simulation procedure for the MLE solution that guarantees stability of the generated spike train . In the “reset condition” [34] , the conditional intensity of the point process at time t is not calculated using the original CIF ( Eq ( 3 ) ) but instead using only the spike-history effect stemming from the most recent spike at t ^ : λ ( t | H t ) = c exp ( η ( t - t ^ ) ) . This yields a renewal process with guaranteed finite rate in simulations . We found that models simulated with the reset condition do produce finite firing rates and spike trains that look plausible on first sight ( Fig 10D ) . However , the firing rate and autocorrelation function are not matched with the training data . Most notably , neither ISI shape ( Fig 10F and 10G , green line; lv ( reset ) = 0 . 72 ± 0 . 21 , log κ ( reset ) = 0 . 51 ± 0 . 31 , log α ( reset ) = 2 . 57 ± 0 . 20 ) nor ISI correlations are reproduced due to the renewal property of the modified model ( Fig 10H , green line ) .
We now discuss possible relaxing assumptions and extensions to the quasi-renewal framework to assess stability and dynamics of neural point process models . First , the ability to separate the effect of the most recent spike from all previous spikes and to use the moment-generating functional are unique to the exponential nonlinearity . Possibly , for other specific choices of the nonlinearity a similar manipulation or approximation of Eq ( 11 ) may be conceivable , or the firing rate transfer function may be estimated by numerical simulation of the model . However , we consider such an extension of the theory beyond the scope of the present work . Using a different nonlinearity will likely require to invoke some type of Lipschitz condition and alternative ways of studying the stability properties ( see , e . g . , [9] ) . Second , the nonlinear Hawkes process can be formulated as a multivariate process to describe an ensemble of coupled neurons [4 , 5 , 21] . The corresponding extension of the quasi-renewal framework is possible [27] and may be used to study stability of such networks of heterogeneous neurons . Local linear stability analysis of derived fixed points for the neuronal network dynamics can then be readily implemented based on the spectral radius of a coupling matrix ( obtained from coupling coefficients , history filters’ integrals , and the nonlinearity’s first derivative ) computed at the fixed point locations [9 , 21] . Third , we assumed no ( time-varying ) exogenous input . The framework can be easily extended to accommodate non-stationary inputs , such as stimulus drive , by allowing the baseline firing rate c in Eq ( 3 ) to be time-dependent and performing the QR stability analysis for the supremum of c as long as such a bound exists . If stability is predicted for this dominating model , the model with time-varying exogeneous input will be stable as long as the exogeneous drive is independent of the firing rate of the neuron itself [21] . This generalization allows the stability analysis to be performed , e . g . , for state-space models such as linear dynamic systems with conditionally Poisson observations , where spike-history effects are combined with neural couplings to a low-dimensional latent state whose dynamics is stable itself [2 , 38–42] . Fourth , throughout this study , we assumed an absolute refractory period of 2 ms . Because the absolute length of such a refractory period does not impact the theoretical analysis , a refractory period of arbitrary length may be used . Although the existence of refractory periods is pervasive in most physical applications , the assumption of an absolute refractory period might not be justified in some cases . The refractory period leads to a finite support of the transfer function and a bounded firing rate . In the limit of a vanishing refractory period , there might be additional alignments of fixed points to those outlined in Fig 3 , such as a single stable and a single unstable fixed point , which can be similarly classified into stable , fragile , and divergent dynamics . Finally , we provided an approximation of the divergence time for fragile , metastable models . A more direct estimation of the divergence time seems desirable . One alternative to estimate the divergence time would be a fluctuation analysis in analogy to Brownian motion in a potential . In this case , the divergence time would correspond to the escape time of the particle from the potential given around the low-rate stable fixed point . In first order , noise in this process may be described by white noise , scaled with an intensity that should be proportional ( if not equal ) to the square root of the rate ( Poisson statistics ) . A more insightful or useful estimate of the fluctuations may also be derived from the power spectral density of the activity in the metastable state [27 , 43] . Alternatively , a periodic perturbation approach as in [7] may be attempted to calculate the time scales of expected divergences . We provided a proof-of-principle of how our proposed quasi-renewal framework can be incorporated into a maximum-likelihood parameter estimation procedure in order to guarantee the stability of models estimated from physiological data . We saw that the ordinary MLE provided unphysiological spike trains while the stabilized version matched first- and second-order statistics of the training data with only marginal loss in predictive power and goodness-of-fit scores . In our current implementation , we maximize the likelihood of the model under the constraint that the model is predicted to be stable by the QR approximation . The determination of stability based on the number and location of fixed points—essentially a bifurcation analysis—consists of multiple steps and is highly nonlinear . We were not able to differentiate the penalty term to exploit more efficient gradient-based optimization schemes . However , empirically , we have observed that the space of admissible parameters seems to form a single connected , possibly even convex , set . From a computational point of view , the determination of stability involves one-dimensional ( scalar ) arithmetic that allows fast evaluation of many candidate parameter sets during the optimization procedure . If necessary , additional speed-ups could be obtained by parallelizing the computation of the penalty term when evaluating different local search directions . We restricted the attainable parameter space to all models that are classified as stable , thereby explicitly excluding fragile models with long expected divergence times . In practice , stable models and fragile models are both candidates for physiological dynamics as long as simulation times are shorter than typical divergence times . Therefore , a more refined cost function than the one used here ( see Eq ( 35 ) ) could involve a penalty proportional to the expected divergence rate E[Tdiv]−1 , weighted by a corresponding regularization parameter . The calculation of the expected divergence rate is more computationally expensive than the determination of stability itself but would provide a continuous and potentially smooth penalty function that could be superior to the all-or-nothing penalty term of the proposed optimization scheme . Alternatively , a computationally efficient regularization is the L1-regularized maximum-likelihood estimate . It provides a convex optimization problem that can be efficiently solved [44–46] . For stronger regularization parameters , estimated coefficients tend towards zero . Nonlinear Hawkes processes with vanishing spike-history filter ( coefficients tending towards zero ) are always stable . This implies the existence of an optimally L1-regularized solution that is predicted to be stable . Therefore , strong L1-regularization might be an alternative approach to model stabilization . In addition , a more parametric description of the spike-history kernel could facilitate stability . For example , the addition of an adapting ( inhibitory ) power-law component is likely to prevent any runaway-excitation and has been observed experimentally under certain conditions [17] . A parametric form like this will have to be accompanied by model selection and appropriate goodness-of-fit tests . A simple way to ensure stability of nonlinear Hawkes processes in simulations is to implement a “reset condition” following each spike when the previous spiking history is forgotten . This leads to a renewal process with a well-defined stationary and unique solutions [47] . However , by definition , second-order statistics such as ISI correlations cannot be reproduced and spike trains generated with this condition are not realizations from the original nonlinear Hawkes model . Nevertheless , it might be an easy fix in certain applications where stable network simulations are desired without explicitly reproducing physiological spike train statistics [34] . Finally , we note that while previous work in the mathematical domains covers uniqueness and existence theorems for stationary point processes [9 , 21] , it does not provide predictions of dynamics , metastability , or whether steady-state rates are in a physiological regime . In the case of Lipschitz-continuous nonlinearities , existing conditions for stability are overly conservative and of little relevance for neurophysiologically plausible spike-history filters . We are currently working on relaxations of these conditions that would allow easier characterizations of stability of neuron models and neuronal networks , and we hope to report such results in the future . The ability to predict stochastic stability of a given point process model has merit in its own right and is therefore a main contribution of our study . While stability in simulations is neither a necessary nor sufficient condition for the model fitting procedure itself , and a trade-off between stability constraints and other standard regularized MLE approaches should be considered case by case , data-driven models that are guaranteed to be stable are of major importance for many applications . We will conclude by giving a few examples . First , to understand brain circuits may mean to be able to rebuild them using artificial components . PP-GLMs offer a direct and relatively well-understood method to derive neuron models from data . But their use in rebuilding brain circuits in simulation is limited if stochastic stability is uncertain . For the simple example of a PP-GLM fit of a neuronal network with an embedded “synfire chain” , Zaytsev et al . ensured stability of the network by adding a reset mechanism of the membrane after each spike to prevent run-away excitation [34] . Second , PP-GLMs are used in the context of ( closed-loop ) brain machine interfaces . Present-day experiments may interface brain tissue to virtual actuators that in turn provide feedback signals to the brain ( e . g . , [48] ) . Future applications of such technology may replace damaged neural tissue by simulated neural circuits which are connected bidirectionally to the brain . Such circuits could potentially be made using PP-GLMs or related models , fitted to the system that is being substituted . Stability of the model is essential in this case to exclude major system malfunction . A final application is to make spike-timing predictions from neuron models: Given parallel recordings of neuronal activity , precise timing of single spikes can already be predicted using PP-GLMs [5] . However , such predictions have been limited to the very near future ( on the order of milliseconds ) , but typically cannot be used for extended time differences into the future because of model instability . Ensuring stability of PP-GLMs potentially extends the time horizon of spike timing prediction . This may enable applications in epilepsy treatment , seismology , finance , and other fields that study self-exciting dynamical processes .
For the non-human primate data , all procedures were in accordance with Brown University Institutional Animal Care and Use Committee approved protocols and the Guide for the Care and Use of Laboratory Animals . Approval for the human studies was granted by local Institutional Review Boards ( Partners Human Research Committee , Massachusetts General Hospital , Harvard Medical School ) , and the participant was enrolled after obtaining informed consent . Full experimental details for the electrophysiological data from the crab can be found in [6] , for the non-human primate data in [31] , and for the human data in [30 , 49] . The nonlinear Hawkes process is a flexible class of self-exciting and/or self-inhibiting point process models [9] . For a stochastic point process , the conditional intensity function is given by [10]: λ ( t | H t ) = lim Δ → 0 P ( spike in ( t , t + Δ ] | H t ) Δ , ( 9 ) where H t is the history of the process ( i . e . , all K spikes at times tk up to time t ) . In the nonlinear Hawkes model , λ ( t | H t ) is assumed to take the following form: λ ( t | H t ) = ϕ [ η * S ] ( t ) = ϕ ∑ k = 1 K η ( t - t k ) , ( 10 ) where ϕ ( x ) : R → R + is a nonlinearity that maps the convolution of the spike train S with a causal auto-history kernel η ( s ) ∈ R onto a non-negative conditional intensity λ ( t | H t ) . Here , we consider ϕ ( x ) = c exp ( x ) = exp ( I0 + x ) with c = exp ( I0 ) > 0 . The exponential nonlinearity implies that modulations from previous spikes interact in a multiplicative way [50] . The choice for the exponential nonlinearity has both theoretical [2 , 7] and empirical support , e . g . , from electrophysiological experiments [51 , 52] . We only consider the single-neuron ( univariate ) case although nonlinear Hawkes processes can be similarly defined for coupled neuronal ensembles with the corresponding matrix of auto- and cross-history kernels ηij ( s ) . Furthermore , we assume constant exogenous input , i . e . , c ≡ const . For time-varying inputs c ( t ) or other ( potentially non-stationary ) exogenous inputs , a conservative stability analysis can be performed by using csup = sup c ( t ) as long as such a bound exists . Hence , the stochastic process is completely determined by two parameters: c ( or equivalently I0 ) and the causal auto-history kernel η ( s ) . To simplify analysis , we make certain assumptions about η ( s ) . One is the introduction of an absolute refractory period τref which indicates that the next spike can occur no closer than τref to the last spike . It can be modeled by setting η ( s ) = −∞ for s < τref . Given that action potentials , the events that are modeled with the point process model , have an extent of around 1 ms , we assume τref = 2 ms . In the nonlinear Hawkes model , the effects of previous spikes can accumulate . This leads , in general , to a non-renewal process . For this specific class of point process models , there are no closed-form formulas to predict mean intensities , inter-spike interval ( ISI ) distributions , power spectra , or other properties of the process . We are interested in whether a nonlinear Hawkes process with given parameters permits stable , finite steady-state firing rates . To make progress , we need to approximate Eq ( 10 ) . Our approach is based on the recently introduced quasi-renewal approximation [26 , 27] . To obtain an estimate of the mean rate of the process A0 , we average λ ( t | H t ) over all possible spike trains S ( t ) prior to the last spike t ^: λ 0 ( t , t ^ ) = 〈 λ ( t | H t ) 〉 S ( t < t ^ ) = c exp ( η ( t - t ^ ) ) 〈 exp ( [ η * S ] ( t ) ) 〉 S ( t < t ^ ) . We identify the second term with the moment-generating functional of S that can be expanded in a series of moments [26 , 53] which we truncate after the first order: λ 0 ( t , t ^ ) ≈ c exp ( η ( t - t ^ ) ) exp ∫ - ∞ t ^ ( e η ( t - t ′ ) - 1 ) 〈 S ( t ′ ) 〉 S ( t < t ^ ) ︸ A 0 d t ′ , ( 11 ) with A 0 = 〈 S ( t ′ ) 〉 S ( t < t ^ ) being the first moment of the averaged spike train , which corresponds to a constant by assuming stationarity . A0 is the steady-state firing rate which for now remains unknown . We can rewrite the integration bounds and introduce τ = t - t ^ as the time since the last spike to obtain: λ 0 ( τ ) = c exp η ( τ ) + A 0 ∫ τ ∞ ( e η ( u ) - 1 ︸ γ ( u ) ) d u , ( 12 ) with γ ( u ) = eη ( u ) − 1 for the exponentiated kernel and Γ = ∫ 0 ∞ γ ( u ) d u . Γ is used in the rescaling of the filter for the analysis presented in Fig 9 . Using the quasi-renewal ( QR ) conditional intensity of Eq ( 12 ) , we obtain the steady-state survivor function S 0 and inter-spike interval ( ISI ) density P0 as: S 0 ( τ ) = exp - ∫ 0 τ λ 0 ( u ) d u , ( 13 ) P 0 ( τ ) = S 0 ( τ ) λ 0 ( τ ) , ( 14 ) which , in turn , predict the firing rate: f ( A 0 ) = ∫ 0 ∞ τ P 0 ( τ ) d τ - 1 . ( 15 ) Because d d τ S 0 ( τ ) = - P 0 ( τ ) and integrating by parts: f ( A0 ) =[ −∫0∞τddτS0 ( τ ) dτ ]−1=[ −[ τS0 ( τ ) ]0∞+∫0∞S0 ( τ ) dτ ]−1=[ ∫0∞S0 ( τ ) dτ ]−1 , ( 16 ) which is continuous and differentiable in A0 . Eq ( 16 ) defines an average input-output mapping for the single neuron , known as transfer or gain function , which maps an assumed mean input rate A0 to the mean output rate of the process f ( A0 ) . The dynamics of the model can be characterised based on the properties of the transfer function . Fixed points of this map , A ^ 0 = f ( A ^ 0 ) , can be locally stable or unstable . To detect the fixed points reliably , we search for the zero crossings of the function g ( A0 ) = f ( A0 ) − A0 . A fixed point A ^ 0 is locally stable if g ′ ( A ^ 0 ) = d d A 0 g ( A 0 ) | A 0 = A ^ 0 < 0 and unstable if g ′ ( A ^ 0 ) ≥ 0 . Based on the number and location of fixed points of the mean firing rate map ( Eq ( 16 ) ) we can then classify the model ( see Results ) . For fragile ( metastable ) models , we may ask whether there is an inter-spike interval x , which if several spikes occur repeatedly with this interval , causes a divergence of the firing rate . If such an x exists , we can compute the probability of this event . Although other routes to a divergent rate are possible , this one yields an explicit value for its rate of occurrence and can be used as a lower bound for the divergence rate of a metastable model . Let t1 be the time of a spike of the process . Then the next spike occurs within the interval x with probability 1 - S ( t 1 + x , t 1 ) , where S is the survivor function . We may now iterate this argument to compute the probability that , following t1 , there is a sequence of K spikes with intervals smaller or equal to x , as: p reg ( x ) = ( 1 − S ( t 1 + x , t 1 ) ) ( 1 − S ( t 1 + 2 x , t 1 + x ) ) … = ∏ k = 1 K [ 1 − S ( t 1 + k x , t 1 + ( k − 1 ) x ) ] ︸ p ( x , k ) ( 17 ) To evaluate Eq ( 17 ) and compute S , we need to approximate the intensity function λ ( t | H t ) for the case that up to t1 we do not have information about the spike history apart from the rate A0 , but from t1 on it is defined as the regular firing case with spike times tk = t1 + ( k − 1 ) x for k ≥ 1 . This gives rise to a similar quasi-renewal approximation as for Eq ( 12 ) : λreg ( t , x ) =〈λ ( t|Ht ) 〉S ( t<t1 ) =c exp{ ∑k=1∞ η ( t−tk ) } 〈exp{ [ η*S ] ( t ) }〉S ( t<t1 ) ≈c exp{ ∑k=1∞η ( t−t1− ( k−1 ) x ) } exp{ ∫−∞t1 ( eη ( t−t′ ) −1 ) A0dt′ }=c exp{ ∑k=1∞η ( t−t1− ( k−1 ) x ) } exp{ A0∫t−t1∞γ ( s ) ds } . ( 18 ) Inserted into Eq ( 17 ) , we then first check whether the sequence p ( x , k ) increases monotonically towards 1 , setting t1 = 0 . If for a given k we have p ( x , k ) > p ( x , k + 1 ) , we terminate the iteration because x does not seem to lead to the regular divergence and return preg = 0 . If , in contrast , for some value of k , p ( x , k ) is close to 1 , we have found a divergent case that occurs with probability preg ( x ) , as given by Eq ( 17 ) with K = k . This procedure is performed for all x ∈ [ τ ref , A 0 - 1 ] , and the maximum maxx preg ( x ) is returned . As preg ( x ) is the probability of the regular divergence with intervals x or shorter to occur after any spike of the process , the rate of divergence is thus bounded from below by: r div ≥ A 0 max x p reg ( x ) . ( 19 ) This provides an upper bound for T div = r div - 1 . Apart from the dynamic stability that we have discussed so far , a particular limit of the space of possible spike trains is of special interest . In case of a divergent firing rate , which occurs in unstable or fragile models , the analysis of the gain function predicts that the firing rate saturates at the limit given by the inverse of the refractory period , A 0 = τ ref - 1 . However , there is only one spike train that can realize this firing rate , which is the regular spike train: S x ( t ) = ∑ k = - ∞ ∞ δ ( t - k x ) , ( 20 ) with inter-spike-interval x = τref . Here θ denotes the Heaviside function and δ denotes the Dirac delta function . For the regular spike train Sx , with x > τref being close to the refractory period , to be a possible mode of firing of the model , it is necessary that the conditional intensity of the neuron ( Eq ( 10 ) ) , evaluated at time x after the last spike ( at t − x ) , λ reg K ( x ) = c exp ∑ k = 1 K η ( k x ) , ( 21 ) reaches a sufficiently high value , so that the rate x−1 can be maintained , when the regular spike train extends into the past forever , λ reg ( x ) = lim K → ∞ λ reg K ( x ) . But what precisely is that sufficiently high value of λreg ? Since after τref the refractory period is over , and because the conditional intensity changes approximately on the time scale of the filter η that is much greater than the remaining interval x − τref , we may approximate the mean output inter-spike-interval of the process by: μ reg ( x ) = τ ref + λ reg - 1 ( x ) . ( 22 ) Now we can formulate a condition on the divergent firing model class: If the expected interval μreg in the regular firing case is smaller or equal to the input interval x , the regular firing state can be maintained . From Eqs ( 22 ) and ( 21 ) , we obtain the condition: ∑ k = 1 ∞ η ( k x ) ≥ - ln c ( x - τ ref ) . ( 23 ) This condition is instructive in two ways: First , to maintain the interval x = τref , the series on the left hand side ( LHS ) has to diverge to positive infinity quicker than the logarithm on right hand side ( RHS ) . Second , in case the series on the LHS of Eq ( 23 ) converges for all x ≥ τref , there is a minimum value of x for which Eq ( 23 ) is still fulfilled . Then x−1 is the peak firing rate that this model can maintain close to the regular spiking limit . Many models , even ones with an upper unstable fixed point , might not fulfill Eq ( 23 ) . These models are fragile , but also cannot maintain the regular firing mode . Nonetheless we may ask for how many regular spikes they can maintain the tonic activity . This can be addressed by a modified condition like Eq ( 21 ) considering K < ∞ . By analogous reasoning as above , we arrive at the condition: ∑ k = 1 K η ( k x ) ≥ - ln c ( x - τ ref ) . ( 24 ) For a given tonic firing interval x close to τref ( e . g . , defined as x - 1 = 0 . 9 × τ ref - 1 ) , the maximum Kmax for which Eq ( 24 ) is fulfilled yields a good approximation of the duration Kmax x of the intermittent regular spiking episodes of the model . If the spike-history filter is a sum of two exponential terms , the condition in Eq ( 24 ) takes the specific form of a geometric series: − ln [ c ( x − τ ref ) ] ≤ ∑ k = 1 K η ( k x ) = ∑ k = 1 K [ J r e − k x / τ r + J a e − k x / τ a ] = J r e − x / τ r ∑ k = 0 K e − k x / τ r + J a e − x / τ a ∑ k = 0 K e − k x / τ a = J r e − x / τ r 1 − e − K x / τ r 1 − e − x / τ r + J a e − x / τ a 1 − e − K x / τ a 1 − e − x / τ a = J r 1 − e − K x / τ r e x / τ r − 1 + J a 1 − e − K x / τ a e x / τ a − 1 , ( 25 ) which for K → ∞ becomes condition Eq ( 23 ) , which here is: - ln c ( x - τ ref ) ≤ J r ( e x / τ r - 1 ) - 1 + J a ( e x / τ a - 1 ) - 1 . ( 26 ) The boundary defined by Eq ( 26 ) in the ( Jr , Ja ) space is a line: For every Jr there is a maximum J a max from which on Eq ( 26 ) is true . For J a < J a max , in contrast , we are assured that the regular firing mode with interval x is unstable . Models for which Eq ( 26 ) is not fulfilled will show intermittent bursting activity ( compare with Fig 7 ) . We estimate the average divergence time of a given neuron model by simulating N = 48 independent neurons for T = 1000 s each . A neuron is said to have diverged at time t if its average firing rate in the interval [t − 1 , t + 1] seconds exceeds λ t h r = 0 . 9 × τ ref - 1 . Alternatively , the firing rate may stay below λthr until the end of the simulation ( “censored observation” ) . We can now derive the maximum-likelihood estimate of the divergence time . We assume that neurons diverge randomly with rate r = 1/Tdiv . This seems to be justified in practice based on our simulations . Then , the likelihood of observing a divergence time y smaller than T is 1 T div e - y / T div and the probability to observe a censored observation of length T is given by: ∫ T ∞ 1 T div e - y / T div d y = e - T / T div . ( 27 ) If we denote the observed divergence times as y1 , y2 , … , yk , and we have Nc = N − k censored observations , the overall log-likelihood function is given by: log L = ∑ i = 1 k log ( 1 T div e - y i / T div ) + N c log ( e - T / T div ) ( 28 ) = - ( N - N c ) log ( T div ) - N c T T div - ∑ y i T div . ( 29 ) At the maximal ( log- ) likelihood estimate T ^ div , the gradient with respect to Tdiv has to vanish: ∂ log L ∂ T div | T div = T ^ div = - N - N c T ^ div + N c T T ^ div 2 + ∑ y i T ^ div 2 = ! 0 , ( 30 ) T ^ div ( N - N c ) = N c T + ∑ y i , ( 31 ) T ^ div = N c T + ∑ y i N - N c . ( 32 ) This assumes that there was at least one non-censored observation ( Nc < N ) . Otherwise , we set T ^ div = ∞ . Note that if Nc > 0 , T ^ div may be larger than T . If η ( s ) in Eq ( 10 ) is parameterized through a set of basis functions {Bi ( s ) } with linear coefficients {βi} , then η ( s ) = ∑i βi Bi ( s ) . All model parameters { I 0 , β → } can be estimated via the statistical framework of generalized linear models ( GLMs ) [2] . We discretize the spike train to obtain a series of spike counts ni in each time window of length Δ = 1 ms . The expected spike count is given by the discrete-time approximation of Eq ( 10 ) as E[ni] = λiΔ . The log-likelihood is then proportional to: log L ∝ ∑ i ( n i log ( λ i Δ ) - λ i Δ ) . ( 33 ) For the estimation of physiologically plausible model parameters ( Fig 9 ) , we used 10 raised cosine functions [4] with logarithmically spaced peaks up to 400 ms as basis functions {Bi ( s ) } for the spike-history filter . In addition , an absolute refractory period of τref = 2 ms was enforced . To improve numerical convergence and to ensure finite parameters for very sparse data sets , we added a small L2-penalty term to the log-likelihood function so that the maximum-likelihood estimate ( MLE ) corresponds to the minimum of the cost function: C ( I 0 , β → ) = - log L ( I 0 , β → ) + α ∑ i = 1 10 β i 2 , ( 34 ) with regularization parameter α = 5 ⋅ 10−4 . For the stabilization procedure ( Fig 10 ) , we performed the maximum-likelihood estimation ( see previous section ) under the additional constraint that the model is predicted to be stable by the QR approximation . That is , we optimized the cost function: C ( I 0 , β → ) = - log L ( I 0 , β → ) + α ∑ i = 1 10 β i 2 if the model { I 0 , β → } is predicted to be stable , ∞ otherwise . ( 35 ) We used a gradient-free optimization algorithm ( Nelder-Mead ) with a convergence criterion on the change in parameter values ( ∥ Δ β → ∥ ∥ β → ∥ < 10 - 4 ) . The initial condition was chosen as the MLE solution for which positive parameter coefficients were set to zero . Because basis functions Bi ( s ) are non-negative , this corresponds to a non-positive spike-history filter and ensures that the initial evaluation of the cost function is finite . Then , the optimization starts from a region with finite cost and allows the algorithm to descend to a ( local ) minimum . All simulations with the spike-history filter consisting of one or two exponentials ( Figs 5 to 8 ) were performed using NEST [54] , with neuron model “pp_psc_delta” in time steps of 0 . 5 ms . All other spike train simulations were performed with custom-written MATLAB software with a time discretization of 0 . 2 ms . Analysis and optimization were performed in MATLAB and Python .
|
Earthquakes , gene regulatory elements , financial transactions , and action potentials produced by nerve cells are examples of sequences of discrete events in space or time . In many cases , such events do not appear independently of each other . Instead , the occurrence of one event changes the rate of upcoming events ( e . g , aftershocks following an earthquake ) . The nonlinear Hawkes process is a statistical model that captures these complex dependencies . Unfortunately , for a given model , it is hard to predict whether stochastic samples will produce an event pattern consistent with observations . In particular , with positive feedback loops , the process might diverge and yield unrealistically high event rates . Here , we show that an approximation to the mathematical model predicts dynamical properties , in particular , whether the model will exhibit stable and finite rates . In the context of neurophysiology , we find that models estimated from experimental data often tend to show metastability or even unstable dynamics . Our framework can be used to add constraints to data-driven estimation procedures to find the optimal model with realistic event rates and help to build more robust models of single-cell spiking dynamics . It is a first step towards studying the stability of large-scale nonlinear spiking neural network models estimated from data .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"action",
"potentials",
"medicine",
"and",
"health",
"sciences",
"membrane",
"potential",
"electrophysiology",
"neuroscience",
"simulation",
"and",
"modeling",
"optimization",
"systems",
"science",
"mathematics",
"computational",
"neuroscience",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"transfer",
"functions",
"mathematical",
"functions",
"animal",
"cells",
"mathematical",
"and",
"statistical",
"techniques",
"nonlinear",
"dynamics",
"approximation",
"methods",
"cellular",
"neuroscience",
"cell",
"biology",
"physiology",
"neurons",
"single",
"neuron",
"function",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"computational",
"biology",
"neurophysiology"
] |
2017
|
On the stability and dynamics of stochastic spiking neuron models: Nonlinear Hawkes process and point process GLMs
|
In many cases , the computation of a neural system can be reduced to a receptive field , or a set of linear filters , and a thresholding function , or gain curve , which determines the firing probability; this is known as a linear/nonlinear model . In some forms of sensory adaptation , these linear filters and gain curve adjust very rapidly to changes in the variance of a randomly varying driving input . An apparently similar but previously unrelated issue is the observation of gain control by background noise in cortical neurons: the slope of the firing rate versus current ( f-I ) curve changes with the variance of background random input . Here , we show a direct correspondence between these two observations by relating variance-dependent changes in the gain of f-I curves to characteristics of the changing empirical linear/nonlinear model obtained by sampling . In the case that the underlying system is fixed , we derive relationships relating the change of the gain with respect to both mean and variance with the receptive fields derived from reverse correlation on a white noise stimulus . Using two conductance-based model neurons that display distinct gain modulation properties through a simple change in parameters , we show that coding properties of both these models quantitatively satisfy the predicted relationships . Our results describe how both variance-dependent gain modulation and adaptive neural computation result from intrinsic nonlinearity .
An f-I curve , defined as the mean firing rate in response to a stationary mean current input , is one of the simplest ways to characterize how a neuron transforms a stimulus into a spike train output as a function of the magnitude of a single stimulus parameter . Recently , the dependence of f-I curves on other input statistics such as the variance has been examined: the slope of the f-I curve , or gain , is modulated in diverse ways in response to different intensities of added noise [1]–[4] . This enables multiplicative control of the neuronal gain by the level of background synaptic activity [1]: changing the level of the background synaptic activity is equivalent to changing the variance of the noisy balanced excitatory and inhibitory input current to the soma , which modulates the gain of the f-I curve . It has been demonstrated that such somatic gain modulation , combined with saturation in the dendrites , can lead to multiplicative gain control in a single neuron by background inputs [5] . From a computational perspective , the sensitivity of the firing rate to mean or variance can be thought of as distinguishing the neuron's function as either an integrator ( greater sensitivity to the mean ) or a differentiator/coincidence detector ( greater sensitivity to fluctuations , as quantified by the variance ) [3] , [6] , [7] . An alternative method of characterizing a neuron's input-to-output transformation is through a linear/nonlinear ( LN ) cascade model [8] , [9] . These models comprise a set of linear filters or receptive field that selects particular features from the input; the filter output is transformed by a nonlinear threshold stage into a time-varying firing rate . Spike-triggered covariance analysis [10] , [11] reconstructs a model with multiple features from a neuron's input/output data . It has been widely employed to characterize both neural systems [12]–[15] and single neurons or neuron models subject to current or conductance inputs [16]–[19] . Generally , results of reverse correlation analysis may depend on the statistics of the stimulus used to sample the model [15] , [19]–[25] . While some of the dependence on stimulus statistics in the response of a neuron or neural system may reflect underlying plasticity , in some cases , the rapid timescale of the changes suggests the action of intrinsic nonlinearities in systems with fixed parameters [16] , [19] , [25]–[29] , which changes the effective computation of a neuron . Our goal here is to unify the f-I curve description of variance-dependent adaptive computation with that given by the LN model: we present analytical results showing that the variance-dependent modulation of the firing rate is closely related to adaptive changes in the recovered LN model if a fixed underlying model is assumed . When the model relies only on a single feature , we find that such a system can show only a single type of gain modulation , which accompanies an interesting asymptotic scaling behavior . With multiple features , the model can show more diverse adaptive behaviors , exemplified by two conductance-based models that we will study .
Recently , Higgs et al . [3] and Arsiero et al . [4] identified different forms of variance-dependent change in the f-I curves of various neuron types in avian brainstem and in cortex . Depending on the type , neurons can have either increasing or decreasing gain in the f-I curve with increasing variance . These papers linked the phenomenon to mechanisms underlying spike rate adaptation , such as slow afterhyperpolarization ( sAHP ) currents and slow sodium channel inactivation . We recently showed [7] that a standard Hodgkin–Huxley ( HH ) neuron model , lacking spike rate adaptation , can show two different types of variance-dependent gain modulation simply by tuning the maximal conductance parameters of the model . These differences in gain modulation correspond to two different regimes in the space of conductance parameters . In one regime , which includes the standard parameters , a neuron periodically fires to a sufficiently large constant input current . In the other regime , a neuron never fires to a constant input regardless of its magnitude , but responds only to rapid fluctuations . This rarely discussed property has been termed class 3 excitability [30] , [31] . Higgs et al . [3] proposed that the type of gain modulation classifies the neuron as an integrator or differentiator . Here , we examine two models that show these different forms of variance-dependent gain modulation without spike rate adaptation , and study the resulting LN models sampled with different stimulus statistics . We show that these fixed models generate variance-dependent gain modulation , and that this gain modulation is well predicted by aspects of the LN models derived from white noise stimulation . The two models are both based on the HH [32] active currents; one model is the standard HH model , and the other ( HHLS ) has lower Na+ and higher K+ conductances . The HHLS model is a class 3 neuron and responds only to a rapidly changing input . For this reason , the HHLS model can be thought of as behaving more like a differentiator than an integrator [3] , [7] . Figure 1 shows the different gain modulation behaviors of the HH and HHLS conductance-based models . For the HH model , Figure 1A , the f-I curves in the presence of noise are similar to the noiseless case except that they are increasingly smoothed at the threshold . In contrast , Figure 1C shows that the f-I curves of the HHLS model never converge toward each other as the noise level increases . This case resembles that of layer 5 pyramidal neurons in rat medial prefrontal cortex [4] , as well as nucleus laminaris ( NL ) neurons in the chick auditory brainstem and some pyramidal neurons in layer 2/3 of rat neocortex [3] . While for these layer 2/3 neurons , there is evidence that this change in f-I curve slope may be related to the sAHP current [3] , at steady state this effect can be obtained in general by tuning the maximal conductances without introducing any mechanism for spike rate adaptation [7] . For a system described by an LN model with a single feature , we derive an equation relating the slopes of the firing rate with respect to stimulus mean and variance . We then consider gain modulation in a system with multiple relevant features and derive a series of equations relating gain change to properties of the spike-triggered average and spike-triggered covariance . Throughout , we assume that the underlying system is fixed , and that its parameter settings do not depend on stimulus statistics . For example , if the model has a single exponential filter with a time constant τ , we assume that τ does not change with the stimulus mean ( I0 ) or variance ( σ2 ) . However , this does not mean that the model shows a single response pattern regardless of the statistical structure of stimuli . The sampled LN description of a nonlinear system with fixed parameters—even when the underlying model is an LN model [25]—can show interaction with the input statistics , leading to different LN model descriptions for different input parameters [19] , [25] , [27]–[29] . We refer to this as intrinsic adaptation . An LN model is composed of its relevant features {εμ ( t ) } ( μ = 1 , 2 , … , n ) ) , which act as linear filters on an incoming stimulus , and a probability to spike given the filtered stimulus , P ( spike|filtered stimulus ) . For a Gaussian white noise stimulus with mean I0 and variance σ2 , the firing rate is ( 1 ) where is the time-integrated filter and x is the mean-subtracted noise stimulus filtered by the n relevant features . p ( x ) is an n-dimensional Gaussian distribution with variance σ2 . We refer to the Materials and Methods section for a more detailed account of the model . For a one-dimensional model n = 1 , Equation 1 can be rewritten with change of variables ( 2 ) Since p ( x ) is Gaussian , it is also the kernel or Green's function of a diffusion equation in terms of ( x , σ2 ) and therefore so is p ( x−I0ε̅ ) in terms of ( I0 , σ2 ) . In other words , we haveNow operating with on both sides of the equation , p ( x−I0ε̅ ) is the only term on the left hand side of Equation 2 that depends on ( I0 , σ2 ) and therefore the right hand side of Equation 2 vanishes . Thus one finds ( 3 ) The boundary condition is given by evaluating Equation 2 as σ2→0; here the Gaussian distribution becomes a delta functionand the boundary condition is given by the zero-noise f-I curve . Thus , when a model depends only on a single feature , ε ( t ) , the f-I curve with a noisy input is given by a simple diffusion-like equation , Equation 3 , with a single parameter , the time integrated filter , , determining the diffusion constant 1/2ε̅2 . Equation 3 states that the variance-dependent change in the firing rate is simply determined by the curvature of the f-I curve . Thus , a one-dimensional system displays only a single type of noise-induced gain modulation: as in diffusion , an f-I curve is gradually smoothed and flattened as the variance increases . Given a boundary condition , such as an f-I curve for a particular variance , the family of f-I relations can be reconstructed up to a scale factor by solving Equation 3 . For example , one can predict how the neuron would respond to a noise stimulus based on its output in the absence of noise . Note that the solution of Equation 3 generalizes a classical result [33] based on a binary nonlinearity to a simple closed form which applies to any type of nonlinearity . Figure 2A and 2B show a solution of Equation 3 . While this one-dimensional model is based on the simplest and most general assumptions , it provides insights into the structure of variance-dependent gain modulation . The boundary condition is an f-I curve with no noise , f = ( I+0 . 1 ) 1/2 for I>0 and f = 0 for I≤0 , which imitates the general behavior of many dynamical neuron models around rheobase [34]–[36] . Compared with the HH conductance-based model , Equation 3 captures qualitative characteristics of the HH f-I curve despite differences due to the increased complexity of the HH model over a 1D LN model: in Figure 2A and 2B , there is a positive curvature ( second derivative of firing rate with respect to current ) of the f-I curve below rheobase related to the increase of the firing rate with increasing variance . In contrast , the behavior of the HHLS model cannot be described by Equation 3 . Even though the f-I curves in Figure 1C mostly have negative curvature , the firing rate keeps increasing with variance , implying that the HHLS model cannot be described by a one-dimensional LN model . We also compared Equation 3 with the f-I curves from two commonly used simple neuron models , the leaky integrate-and-fire ( LIF ) model ( Figure 2C ) , and a similar model with minimal nonlinearity , the quadratic integrate-and-fire ( QIF ) model [37] , [38] ( Figure 2D ) . The f-I curves of the two models are similar but have subtle differences: in the LIF model , firing rate never decreases with noise , even though parameters were chosen to induce a large negative curvature , as shown analytically in Text S1 . The QIF model behavior is much more similar to the 1D LN model , marked by a slight decrease in firing rate at large I0 . From this perspective , the QIF is a simpler model in terms of the LN description despite the dynamical nonlinearity . It is interesting to note that for one-dimensional models , the gain modulation given by Equation 3 depends only on the boundary condition , which implicitly describes how an input with a given mean samples the nonlinearity , but not explicitly on the details of filters or nonlinearity . An ideal differentiator , where firing rate is independent of the stimulus mean , is realized only when the filter has zero integral , ε̅ = 0 . This is also the criterion that would be satisfied if the filter itself were ideally differentiating . We will return to the relationship between the LN model functional description and that of the f-I curves in the Discussion . Here we examine gain modulation in the case of a system with multiple relevant features . In this case , one cannot derive a single simple equation such as Equation 3 . Instead , we derive relationships between the characteristics of f ( I0 , σ ) curves and quantities calculated using white noise analysis . Fixed multidimensional models can display far more complex response patterns to different stimulus statistics than one-dimensional models , because linear components in the model can now interact nonlinearly [29] . For example , in white noise analysis , as the stimulus variance increases , the distribution of the filtered stimuli also expands and probes different regions of the nonlinear threshold structure of the model . This induces a variance-dependent rotation among the filters recovered through sampling by white noise analysis , and the corresponding changes in the spike-triggered average , spike-triggered covariance , and the sampled nonlinearity [19] . Here , we relate parameters of the changing spike-triggered average and spike-triggered covariance description to the form of the f-I curves . The relationships are derived by taking derivatives of each side of Equation 1 with respect to I0 and σ2 ( see Materials and Methods section ) . The first order in I0 establishes the relationship between the STA and the gain of the f-I curve with respect to the mean ( 4 ) The second order leads to a relationship between the second derivative of the f-I curve and the covariance matrix ( 5 ) The gain with respect to the variance is ( 6 ) where Equations 4–6 show how the nonlinear gain of an f-I curve with respect to input mean and variance is related to intrinsic adaptation as observed through changes in the STA and STC . Note that Equations 4–6 apply to one-dimensional LN models as well . In that case , the STA has the same shape as the feature in the model , and only its magnitude varies according to the overlap integral , Equation 1 , between the nonlinearity of the model and the prior stimulus . This is the same for the STC , and thus Equations 4–6 are not independent . This leads to a single form of variance gain modulation , given by Equation 3 . However , in a multidimensional model , changing the stimulus mean shifts the nonlinearity in a single direction , , while increasing the variance expands the prior in every direction in the stimulus space . Therefore , the overlap integral can show more diverse behaviors . We now examine whether the gain modulation behaviors we have described can be captured by a multi-dimensional LN model . We tested this by computing f-I curves , spike-triggered averages and the spike-triggered covariance matrices for the noise-driven HH and HHLS models for a range of input statistics . Figure 3A , B , and C show the result of fitting simulation data from the HH ( left ) and HHLS ( right ) model to Equations 4 , 5 , and 6 , respectively . The linear relationships are quite clear in Figure 3A and 3C which show the gains with respect to mean and variance . Figure 3B involves the curvature of f-I curves , which is more difficult to calculate accurately , and shows larger errors . In every case , goodness of fit is p<1 . 3×10−6 and p<5 . 8×10−6 for the HH and HHLS where the upper bounds of p-values are given by the case of Equation 5 , corresponding to Figure 3B . These results show that intrinsic adaptation of the LN model predicts the form of noise-induced gain modulation for these models . Here we discuss a consequence of intrinsic adaptation for neuronal encoding of mean and variance information for a one-dimensional model . In this case , Equation 3 completely specifies intrinsic adaptation , and therefore we will focus on this case . Our first observation is that Equation 3 is invariant under the simultaneous rescaling of the mean and standard deviation , I0→αI0 , σ→ασ , where α is an arbitrary positive number . This invariance is preserved if the solution is also a function of only a dimensionless variable I0/σ , which would represent a signal-to-noise ratio if we describe the neuron's input/output function in terms of an f-I curve at a fixed noise level σ . Note that this situation is analogous to the Weber–Fechner [39] , [40] and Fitts' law [41] , which states that perception tends to depend on only dimensionless variables that are invariant under scaling of the absolute magnitude of stimulus [42] . However , the invariance of Equation 3 under the scaling of a stimulus does not necessarily lead to the invariance of a firing rate solution . By rewriting Equation 2 in terms of the “rescaled” variables , y = x/σ and μ = I0/σ , we get ( 7 ) where f0 ( I ) = P ( spike|Iε̅ ) is an f-I curve with no noise . Thus , the scaling of f ( I0 , σ2 ) with standard deviation depends on the boundary condition , f0 ( I ) , which in principle can be any arbitrary function . Nevertheless , in practice , the f-I curves of many dynamical neurons are not completely arbitrary but can share a simple scaling property , at least asymptotically . For example , in the QIF and many other neuron models , the f-I curve with no noise asymptotically follows a power law f0∼ ( I0−Ic ) 1/2 around the rheobase Ic [34]–[36] . In general , if f0 ( I ) ∝Iα asymptotically in such a regime , from Equation 7 , the firing rate is asymptotically factorized into a σ dependent and μ = I0/σ dependent part as ( 8 ) In other words , I0/σ becomes an intermediate asymptotic of the f-I curves [43] . To test to what extent this scaling relationship holds in the models we have considered , we calculated the rescaled relative gain of the f-I curves , which we define as ( σ/f ) ∂f/∂I0 = σ ∂ log f/∂I0; the rescaled relative gain of Equation 8 depends only on μ = I0/σ , not on σ . Thus , if the rescaling strictly holds , this becomes a single-valued function of the signal-to-noise ratio , I0/σ , regardless of the noise level σ . We find evidence for this form of variance rescaling in the QIF , LIF , and HH models . Figure 4 shows the rescaled gains evaluated from the simulated data . The QIF and HH case , Figure 4B and 4D , match well with the solution of Equation 3 , Figure 4A . In the LIF case , Figure 4C , the relative gain shows deviations at low variance , but it approaches a variance-independent limit at large σ . We also present an analytic account in Text S1 . On the other hand , in Figure 4E , the HHLS model does not exhibit this form of asymptotic scaling at all . The role of the signal-to-noise ratio , I0/σ , in the HHLS model appears to be quite distinct from the other models . In summary , Equation 3 predicts that one-dimensional LN models will have the tendency to decrease gain with increasing noise level . However , if the f-I curve of a neuron is power-law-like , the resulting gain modulation will be such that the neuron's sensitivity to mean stimulus change at various noise levels is governed only by the signal-to-noise ratio .
In this paper , we have obtained analytical relationships between noise-dependent gain modulation of f-I curves and properties of the sampled linear/nonlinear model . We have shown that gain control arises as a simple consequence of the nonlinearity of the LN model , even with no changes in any underlying parameters . For a system described by an LN model with only one relevant feature , a simple single-parameter diffusion relationship relates the f-I curves at different variances , where the role of the diffusion coefficient is taken by the integral of the STA . This form strictly limits the possible forms of gain modulation that may be manifested by such a system . The result qualitatively describes the variance dependent gain modulation of different neuron models such as the LIF , QIF , and standard HH neuron models . Models based on dynamical spike generation , such as QIF , showed better agreement with this result than the LIF model . The QIF model case is a good example of how a nonlinear dynamical system can be mapped onto an LN model description [19] , [44] . The QIF model has a single dynamical equation whose subthreshold dynamics are captured approximately by a linear kernel , which takes the role of the feature; one can then determine a threshold which acts as a binary decision boundary for spiking . Thus , it is reasonable that the QIF model and the one-dimensional LN model show a similar response pattern to a noisy input . When the system has multiple relevant features , we obtain equations relating the gain with respect to the input mean and the input variance to parameters of the STA and STC . We verified these results using HH neurons displaying two different forms of noise-induced gain control . Previous work has related different gain control behaviors to a neuron's function as an integrator or a differentiator [3] , [7] . From an LN model perspective , the neuron's function is defined by specific properties of the filter or filters ε ( t ) . An integrating filter would consist of entirely positive weights; for leaky integrators these weights will decay at large negative times . A differentiating filter implements a local subtraction of the stimulus , and so should consist of a bimodal form where the positive weights approximately cancel the negative weights . In general , characterizations of neural function by LN model and by f-I curves are quite distinct . The f-I approach we have discussed here describes the encoding of stationary statistical properties of the stimulus by time-averaged firing rate , while the LN model describes the encoding of specific input fluctuations by single spikes , generally under a particular choice of stimulus statistics . Indeed , the LN characterization can change with the driving stimulus distribution , even , in principle , from an integrator to a differentiator . Thus , a model may , for example , act as a differentiator on short timescales but as an integrator on longer timescales . For systems whose LN approximation varies with mean and variance , the neuron's effective computation changes with stimulus statistics , and so does the information that is represented . One might then ask how the system can decode the represented information . It has been proposed that statistics of the spike train might provide the information required to decode slower-varying stimulus parameters [22] , [45] . The possibility of distinguishing between responses to different stimulus statistics using the firing rate alone depends on the properties of the f-I curves . The primary focus of this work is the restricted problem of single neurons responding to driving currents , where the integrated synaptic current in an in vivo-like condition is approximated to be a ( filtered ) Gaussian white noise [46]–[50] . However , our derivations can apply to arbitrary neural systems driven by white noise inputs , if f-I curves are interpreted as tuning functions with respect to the mean stimulus parameter . Given the generality of our results for neural systems , it would be interesting to test our results in cases where firing is driven by an external stimulus . A good candidate would be retinal ganglion cells , which are well-described by LN-type models [9] , [14] , [51]–[53] , show adaptation to stimulus statistics on multiple timescales [23] , [54] and display a variety of dimensionalities in their feature space [14] . A limitation of the tests we have performed here is a restriction to the low firing rate regime where spike-triggered reverse correlation captures most of the dependence of firing probability on the stimulus . The effects of interspike interaction can be significant [16] , [17] , [55] and models with spike history feedback have been developed to account for this [44] , [51] , [56] , [57] . We have not investigated how spike history effects would impact our results . Although evidence suggests that gain modulation by noise may be enhanced by slow afterhyperpolarization currents underlying spike frequency adaptation [3] , these slow currents are not required to generate gain enhancement in simple neuron models [7] , [19] , [25]–[29] . While one may generate diverse types of noise-induced gain modulation only by modifying the mechanism of generating a spike independent of spike history [7] , in realistic situations , slow adaptation currents are present and will affect neural responses over many timescales [58]–[60] . In principle , it is possible to extend our result to include these effects: f-I curves under conditions of spike frequency adaptation have been already discussed [61]–[63] and can be compared to LN models with spike history feedback . However , our goal here was to demonstrate the effects that can occur independent of slow adaptation currents and before such currents have acted to shift neuronal coding properties . The suggestive form of our result for one-dimensional LN models led us to look for a representation of neuronal output that is invariant under change in the input noise level . Our motivation is based on a simple principle of dimensional analysis: the gains of the f-I curves with noise may be asymptotically described by a signal-to-noise ratio , a dimensionless variable depending only on the stimulus itself . We showed that this may occur if the f-I curve with no noise obeys asymptotic power-law properties . Such a property has been determined to arise both from the bifurcation patterns of spike generation [31] , [34] , [35] and due to spike rate adaptation [61] . This relationship implies that the gain of the firing rate as a function of the mean should scale inversely with the standard deviation . Scaling of the gain of the nonlinear decision function with the stimulus standard deviation has been observed to some degree in a number of neural systems [10] , [15] , [22]–[25] , [29] , [64]–[67] . Such scaling guarantees maximal transmission of information [10] , [22] . As we and others have proposed , a static model might suffice to explain this phenomenon [25] , [27] , although in some cases slow adaptation currents are known to contribute [65] , [66] . In summary , we have presented theoretically derived relationships between the variance-dependent gain modulation of f-I curves and intrinsic adaptation in neural coding . In real neural systems , any type of gain modulation likely results from many different mechanisms , possibly involving long-time scale dynamics . Our results show that observed forms of gain modulation may be a result of a pre-existing static nonlinearity that reacts to changes in the stimulus statistics robustly and almost instantaneously .
We used two single compartmental models with Hodgkin–Huxley ( HH ) active currents . The first one is an HH model with standard parameters while the second model ( HHLS ) has a lower Na+ and higher K+ maximal conductance . The voltage changes are described by [32]and the activation variables m , n , and h behave according towhereThe voltage V is in millivolts ( mV ) . For the HH model , the conductance parameters are g̅K = 36 mS/cm2 and g̅Na = 120 mS/cm2 . The HHLS model has g̅K = 41 mS/cm2 and g̅Na = 79 mS/cm2 . All other parameters are common to both models . The leak conductance is g̅L = 0 . 3 mS/cm2 and the membrane capacitance per area C is 1 μF/cm2 . The reversal potentials are EL = −54 . 3 mV , ENa = 50 mV , and EK = −77 mV . The membrane area is 10−3 cm2 , so that a current density of 1 μA/cm2 corresponds to a current of 1 nA . All simulations of these models were done with the NEURON simulation environment [68] . Gaussian white noise currents with various means and variances are generated with an update rate of 5 kHz ( dt = 0 . 2 ms ) and delivered into the model via current clamp . For the f-I curves , we simulated 4 min of input for each mean and variance pair . The whole procedure was repeated five times to estimate the variance of the f-I relationship , σrepeat . We ran another set of simulations for reverse correlation analysis and collected about 100 , 000 spikes for each stimulus condition . The means and variances of the Gaussian noisy stimuli were chosen such that the mean firing rate did not exceed 10 Hz , and we selected eight means and seven variances for the HH model , and nine means and four variances for the HHLS model . In addition to the conductance-based model , we investigated the behavior of two heuristic model neurons driven by a noisy current input . Each model consists of a single dynamical equation describing voltage fluctuations of the form The first model is a leaky integrate-and-fire ( LIF ) model [69] , [70] , for which L ( V ) = −gL ( V−EL ) . We used the parameters gL = 2 , EL = 0 , and C = 1 . Since this choice of L ( V ) cannot generate a spike , we additionally imposed a spiking threshold Vth = 1 and reset voltage Vreset = −3 . The second is a quadratic integrate-and-fire ( QIF ) model [31] , [37] , [38] , for which L ( V ) = gL ( V−EL ) ( V−Vth ) /ΔV where ΔV = Vth−EL>0 . We used gL = 0 . 5 , EL = 0 , Vth = 0 . 1 , and C = 1 . In this model , the voltage V can increase without bound; such a trajectory is defined to be a spike if it crosses Vspike = 5 . After spiking , the system is reset to Vreset = 0 . These two models are simulated using a fourth-order Runge–Kutta integration method with an integration time step of dt = 0 . 01 . The input current I ( t ) was Gaussian white noise , updated at each time step , with a range of means and variances . The f-I curves were obtained from 1 , 000 s of stimulation for each ( mean , variance ) condition . We then compared the f-I curves from these models with the relationship derived in the Results section , Equation 5 . A numerical solution of the partial differential equation was obtained using a PDE solver in Mathematica ( Wolfram Research , Inc . ) . We use the linear/nonlinear ( LN ) cascade model framework to describe a neuron's input/output relation . We will focus on the dependence of the firing rate of a fixed LN model on the mean and variance of a Gaussian white noise input . We will take the driving input to be I ( t ) = I0+ξ ( t ) where I0 is the mean and ξ ( t ) is a Gaussian white noise with variance σ2 and zero mean . The linear part of the model selects , by linear filtering , a subset of the possible stimuli probed by I ( t ) . That subset is expressed as n relevant features {εμ ( t ) } , ( μ = 1 , 2 , … , n ) . Interpreted as vectors , the components of any stimulus that are relevant to changing the firing rate can be expressed in terms of projections onto these features . The firing rate of the model for a given temporal sequence I ( t ) depends only on s , the input filtered by the n relevant features . Thus the firing rate from the given stimulus depends on the convolution of the input with all n features and can be written as P ( spike|s = I0ε̅+x ) whereSince I ( t ) is white noise with stationary statistics , the projections xμ can be taken to be stationary random variables chosen from a Gaussian distribution at each t . Given the filtered stimulus , a nonlinear decision function P ( spike|I0ε̅+x ) generates the instantaneous time-varying firing rate . For a specified model and stimulus statistics , the mean firing rate f ( I0 , σ2 ) = P ( spike ) is simply ( 9 ) where Equation 9 describes an f-I curve of the model in the presence of added noise with variance σ2 . The slope or gain of the firing rate with respect to mean or variance can be computed if P ( spike|I0ε̅+x ) is known . However , the gains can be also obtained in terms of the first and second moments of P ( spike|I0ε̅+x ) , which can be measured directly by reverse correlation analysis . We used spike-triggered reverse correlation to probe the computation of the model neurons through an LN model . We collected about 100 , 000 spikes and corresponding ensembles of spike triggered stimulus histories in a 30 ms long time window preceding each spike . From the spike-triggered input ensembles , we calculated spike-triggered averages ( STAs ) and spike-triggered covariances ( STCs ) . The STA is simply the average of the set of stimuli that led to spikes subtracted from the mean of the “prior” stimulus distribution , the distribution of all stimuli independent of spiking output ( 10 ) Therefore , one may consider only the noise part of the zero mean stimulus . When computing the STC , the prior's covariance is subtracted ( 11 ) In calculating the slope and curvature of the f-I curves , we used 6–10 degree polynomial fitting of the f-I curves , where in any single case , the lowest degree was used which provided both a good fit and smoothness . From the fitting procedure , we obtained the standard deviation of the residuals , σfit . This was repeated five times for f-I curves computed using different noise samples , and from this we computed σrepeat , the standard deviation of each computed slope and curvature . We estimated the total error of our calculation as σtotal = ( σrepeat2+σfit2 ) 1/2 . In practice , σrepeat was always greater than σfit by an order of magnitude . This σtotal was used for the error bars in Figure 3 . To evaluate the goodness of fit in Figure 3 , we used the Pearson χ2 test by using the reduced χ2 statisticwhere O and E represent the right and left hand sides of Equations 4–6 , respectively . From this , the p-values are estimated from the cumulative density function of the χ2 distribution , Q ( χ2/k , k ) . The degree of freedom is k = 54 and k = 34 for the HH and HHLS , respectively . We first present two key identities: the first one , which depends on the form of s having additive mean and noise components , is a change of variables for the gradient of P ( spike|x+I0ε̅ ) ( 12 ) Secondly , when x is a Gaussian random variable with zero mean and variance σ2 , by using integration by parts in can be seen that any function F ( x ) satisfies ( 13 ) Then , we first take derivatives of both sides of Equation 9 ( or equivalently Equation 1 ) , by I0 and σ2 , and apply Equations 12 and 13 . The first order in I0 is ( 14 ) The second order is given by ( 15 ) where δμν is a Kronecker delta symbol . The gain with respect to variance is ( 16 ) Now , we show how the right hand sides of Equations 14–16 correspond to the STA and the STC . Given a Gaussian white noise signal ξ ( t ) , we can split it as ξ = ξ∥+ξ⊥ , where ξ∥ belongs to the space spanned by our basis features {εμ} , and therefore relevant to spiking . ξ⊥ is the orthogonal or irrelevant part . ξ∥ can be written asAgain , x is a Gaussian variable from a distribution Equation 9 . The STA issince ξ⊥ is irrelevant and does not make any contribution . Here we use Bayes theoremAs in Equation 9 , P ( s = x+I0ε̅ ) = p ( x ) , and therefore the STA becomesComparing this result with Equation 14 , we obtain Equation 4 . A similar calculation for the second order [19] showsThis result , combined with Equations 15 and 16 , leads to Equations 5 and 6 , respectively .
|
Many neurons are known to achieve a wide dynamic range by adaptively changing their computational input/output function according to the input statistics . These adaptive changes can be very rapid , and it has been suggested that a component of this adaptation could be purely input-driven: even a fixed neural system can show apparent adaptive behavior since inputs with different statistics interact with the nonlinearity of the system in different ways . In this paper , we show how a single neuron's intrinsic computational function can dictate such input-driven changes in its response to varying input statistics , which begets a relationship between two different characterizations of neural function—in terms of mean firing rate and in terms of generating precise spike timing . We then apply our results to two biophysically defined model neurons , which have significantly different response patterns to inputs with various statistics . Our model of intrinsic adaptation explains their behaviors well . Contrary to the picture that neurons carry out a stereotyped computation on their inputs , our results show that even in the simplest cases they have simple yet effective mechanisms by which they can adapt to their input . Adaptation to stimulus statistics , therefore , is built into the most basic single neuron computations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"neuroscience/theoretical",
"neuroscience",
"biophysics/theory",
"and",
"simulation",
"neuroscience/sensory",
"systems"
] |
2008
|
Intrinsic Gain Modulation and Adaptive Neural Coding
|
Yellow skin is an abundant phenotype among domestic chickens and is caused by a recessive allele ( W*Y ) that allows deposition of yellow carotenoids in the skin . Here we show that yellow skin is caused by one or more cis-acting and tissue-specific regulatory mutation ( s ) that inhibit expression of BCDO2 ( beta-carotene dioxygenase 2 ) in skin . Our data imply that carotenoids are taken up from the circulation in both genotypes but are degraded by BCDO2 in skin from animals carrying the white skin allele ( W*W ) . Surprisingly , our results demonstrate that yellow skin does not originate from the red junglefowl ( Gallus gallus ) , the presumed sole wild ancestor of the domestic chicken , but most likely from the closely related grey junglefowl ( Gallus sonneratii ) . This is the first conclusive evidence for a hybrid origin of the domestic chicken , and it has important implications for our views of the domestication process .
The origin of the domestic chicken has been under debate for centuries [1] . Not only has the geographical center of the first ( and possible additional ) domestication event remained contentious [1]–[3] , but because several closely related species of junglefowls exist in South Asia ( Figure 1 ) , the possibility that chickens originate from multiple wild ancestors has yet to be eliminated . On the basis of observed character differences and cross-breeding experiments , Darwin concluded that domestic chickens were derived solely from the red junglefowl [4] , though this was later challenged by Hutt [1] , who stated that as many as four different species of junglefowls may have contributed to chicken domestication . Molecular studies of mtDNA [5] and retroviral insertions [6] have supported Darwin's view . A study that analyzed both repeat nuclear elements and mitochondrial sequences found evidence that grey and Ceylon junglefowls may hybridize with domestic chickens , but did not provide evidence that these two species have contributed to chicken domestication [7] . To date , no studies have compared gene sequences associated with a specific phenotype found in domestic chickens across numerous wild junglefowls and domestic breeds . The majority of chickens used for commercial egg and meat production in the Western world are homozygous for the yellow skin allele . In live birds , the phenotype is most easily recognized by the presence of yellow legs . The expression of yellow skin is influenced by the amount of carotenoids , primarily xanthophylls , in the feed [1] . More carotenoids produce a more intense yellow color . There is a strong consumer preference for the yellow skin phenotype in certain geographic markets such as USA , Mexico , and China where synthetic pigment may be added to enhance the yellow color [8] , [9] . Carotenoids also play a crucial role for feather or skin pigmentation in some wild birds , a well-known example of which is the flamingo's pink feathers . Carotenoid-based ornaments ( skin or feathers ) in wild birds are considered to be an honest signal of an individual's nutritional status or health , reflecting its foraging efficiency or immune status and are therefore implied to affect sexual attractiveness [10]–[12] . A better understanding of the molecular mechanisms regulating the distribution of carotenoid pigmentation is therefore of considerable interest for evolutionary genetics .
The gene underlying yellow skin was identified by combining linkage analysis and Identical-by-Descent ( IBD ) mapping across breeds with the yellow skin phenotype; IBD mapping was carried out with the assumption that the yellow skin mutation has for most breeds , if not all , been inherited from a common ancestor . yellow skin was previously assigned to chromosome 24 [13] . A Y/W×Y/Y back-cross pedigree , comprising 91 informative meioses , was used to refine the map position of the locus . Close linkage was detected to a single nucleotide polymorphism ( SNP ) located within APOA1 at nucleotide position 5 , 237 , 523 bp at the distal end of chromosome 24 ( lod score = 16 . 4; recombination fraction = 6 . 9% ) . An examination of this chromosomal region revealed an obvious candidate gene for yellow skin , BCDO2 located at position 6 . 26–6 . 29 Mbp . BCDO2 encodes beta-carotene dioxygenase 2 , an enzyme that cleaves colorful carotenoids to colorless apocarotenoids by an asymmetric cleavage reaction [14] . Partial sequence analysis of BCDO2 immediately revealed a SNP in complete linkage disequilibrium with yellow skin across a divergent set of breeds ( Table 1 ) . This highly significant association across breeds and complete fixation within breeds homozygous for yellow skin confirmed our assumption that this allele has been inherited from a single ancestor . Thus , the causal mutation should be located within the minimum shared haplotype present in these breeds . Further sequence analysis revealed that this minimum haplotype spans 23 . 8 kb between nucleotide positions 6 , 264 , 083 to 6 , 287 , 900 . In addition to BCDO2 , this region only contains one other putative gene corresponding to a single chicken cDNA clone ( BX935617; Figure 2A ) . RT-PCR analysis revealed only weak expression of the transcript corresponding to BX935617 and no significant difference between genotypes was documented ( data not shown ) . BCDO2 showed fairly strong expression in both liver and skin . RT-PCR analysis followed by pyrosequencing of six heterozygous birds demonstrated that more than 90% of the transcripts expressed in skin originated from the white skin allele whereas yellow skin and white skin was expressed at about the same level in liver ( Figures 2B and 2C ) . We postulate that yellow skin is caused by tissue-specific regulatory mutation ( s ) that alter BCDO2 expression in skin . Yellow carotenoids are assumed to be taken up to skin in both genotypes but in white skin birds the carotenoids are degraded to colorless apocarotenoids by the action of BCDO2 . We searched for the causal mutation ( s ) by resequencing the entire 23 . 8 kb region from domestic chickens homozygous for yellow skin together with a set of domestic chickens and red junglefowls homozygous for white skin . This analysis revealed a surprisingly high sequence diversity between the two groups ( 0 . 81% ) , well above the genome average for chicken ( ∼0 . 5% ) [15] and approaching the sequence divergence between chimpanzee and human ( 1 . 2% ) [16] . We therefore included three other species of junglefowls in the sequence comparison: grey ( G . sonneratii ) , Ceylon ( G . lafayetii ) , and green ( G . varius ) junglefowls . This step was also motivated by the fact that grey and Ceylon junglefowls have red or yellowish legs which implies deposition of carotenoids and a Y/Y genotype [17] . This had previously prompted Hutt [1] to propose that yellow skin may have been derived from the grey junglefowl . The white skin allele from domestic chicken showed a high sequence identity to red junglefowl sequences whereas the yellow skin sequences clearly clustered with sequences from grey and Ceylon junglefowls ( Figure 3 ) ; Y showed only 13 nucleotide differences ( 0 . 07% ) and three insertions/deletions compared with one of the grey junglefowl sequences . In contrast , mtDNA sequences from the same samples showed the expected pattern in which domestic chickens cluster with red junglefowl within a clade well separated from other junglefowls ( Figure S1 ) . We reanalyzed previously published sequences [7] of chicken repeat 1 elements spread across the genome and demonstrated that trees constructed using three separate regions on chromosome 1 , and another on chromosome 5 , possessed the same approximate topology as revealed by an analysis of mitochondrial sequences ( Figures S2 , S3 , S4 , and S5 ) . On all of these trees , domestic chicken haplotypes cluster exclusively with those of red junglefowl . In contrast , a tree drawn using a region located about ∼650 kb proximal to BCDO2 gave an inconclusive picture ( Figure S6 ) consistent with our finding that introgression from other junglefowl species has affected this region of chromosome 24 . Partial resequencing of the 23 . 8 kb region revealed that all Western breeds fixed for yellow skin carried the same haplotype . We also resequenced haplotypes present in Chinese Shek-ki birds , all of which express the yellow skin phenotype , and found that they carried a distinct haplotype . This haplotype clustered with sequences from grey and Ceylon junglefowls , and could therefore represent a distinct introgression event ( Figure 3 ) . We analyzed five birds from this breed and four were homozygous for the alternate yellow skin haplotype whereas the fifth was a composite heterozygote between the alternate and the haplotype found in European breeds demonstrating that yellow skin is controlled by the same locus in Asian and European breeds . The fact that yellow skin is present among local breeds of domestic chicken across the world suggests that introgression of yellow skin to domestic chickens happened thousands of years ago rather than hundreds of years ago . We also resequenced ∼3 kb of the BCDO2 region from six additional grey junglefowls and all sequences belonged to the yellow skin cluster; three of the sequences clustered with the grey junglefowl ( Delhi ) sequence and the other three clustered with grey junglefowl ( GryJF_04-07 ) ( data not shown ) . In conclusion , all eight tested grey junglefowls were homozygous for alleles that were closely related to the yellow skin allele . In contrast , our SNP screen showed that all 24 tested red junglefowls carried alleles at this locus that are closely related to white skin alleles in domestic chicken ( Table 1 ) . Billions of chickens used for producing meat ( broilers ) and eggs ( layers ) are homozygous Y/Y , though it is unclear why yellow skin has almost replaced the white skin allele in commercial populations . We tested whether yellow skin may have pleiotropic effects on other traits using our large intercross between the red junglefowl ( W/W ) and White Leghorn chickens ( Y/Y ) [18] . The pedigree comprises about 800 F2 progeny which have been scored for a number of phenotypic traits including growth , body composition , egg production , bone density , and behavior . The results of the QTL analysis using the BCDO2 marker are compiled in Table S3 . We tested a total of 81 traits and only four reached nominal significance , which is not more than expected by chance alone . Thus , no highly significant trait association was detected for the BCDO2 locus taking into account the number of tests performed . The most interesting association in relation to the selection for yellow skin in domestic chickens was the slightly higher egg production in birds carrying this allele . However , the statistical support for this association was weak and requires additional investigations .
This study convincingly demonstrates that while domestic chickens inherited the mitochondrial , and most of their nuclear genome from red junglefowl , the yellow skin allele originates from a species of junglefowl other than the red junglefowl , most likely from the grey junglefowl . The alternative explanation that W and Y haplotypes have been segregating within red junglefowl populations for a sufficient period of time to have accumulated the observed sequence divergence can be ruled out because the yellow skin sequence is too similar to the grey junglefowl sequence . As shown in Figures S1 , S2 , S3 , S4 , and S5 , mtDNA and nuclear sequences from the grey junglefowl are clearly distinct from those found in red junglefowl and domestic chicken . The only exception to this rule detected so far is the yellow skin locus . For instance , the divergences between sequences from the grey and red junglefowl are generally similar to the sequence divergence between red and green junglefowl ( Figure 1 and Figures S1 , S2 . S3 , S4 , S5 , and S6 ) . In contrast , the minimal sequence divergence between the grey junglefowl BCDO2 sequence and the domestic yellow skin allele makes it highly unlikely that the divergence between the white and yellow skin alleles predates the speciation of the red and grey junglefowl; the Y sequence would have accumulated numerous sequence differences since the split between the red and grey junglefowl . We cannot exclude the possibility that yellow skin was introgressed to the red junglefowl by hybridization with grey junglefowl prior to domestication , but it is much more plausible that introgression was facilitated by human activities . The red and grey junglefowls are full species as demonstrated by the fact that hybridization does not occur in the wild [17] and when attempted in captivity , only a cross between grey cocks and red hens produced mostly sterile offspring [19] . Hybridization between grey junglefowl and domesticated fowl , however , have been reported in the vicinity of villages within the area of contact between the two wild species [17] , suggesting that the introgression of yellow skin into domestic birds took place after chickens were initially domesticated . A QTL analysis did not reveal any convincing QTL effects associated with the segregation at the yellow skin locus in an intercross between the red junglefowl and White Leghorn chickens . This result is consistent with a previous back-cross experiment which did not reveal any significant difference in body weight or egg production between W/Y and Y/Y birds [20] . However , studies in other species have indicated that access to carotenoids is a limiting factor for egg-laying capacity [21] . During lay , carotenoids are mobilized and deposited in the yolk of the egg . It is therefore worth speculating that the bright yellow skin color , expressed by well-fed yellow skin homozygotes but not by well-fed white skin birds , has been associated with high production and good health at some point during domestication and was therefore favored by early farmers . Of course , yellow skin may also have been selected purely for cosmetic reasons . This study also contributes to the accumulating data supporting King's and Wilson's [22] conjecture of the importance of regulatory mutations as a source for phenotypic variation . Because BCDO2 is expected to have an essential role for the Vitamin A metabolism in vertebrates [14] , loss-of-function mutations may cause severe defects or lethality , whereas a tissue–specific regulatory mutation , like the one presented here , can be tolerated more readily . Other examples of regulatory mutations with important phenotypic effects include a substitution in IGF2 leading to higher muscle-specific expression in pigs [23] and a Pitx1 mutation leading to reduction in pelvic size in sticklebacks [24] . The mutation ( s ) causing the yellow skin phenotype must be located within the 23 . 8 kb region which shows complete association with the yellow skin phenotype across breeds ( Figure 2A ) . The identification of the mutation ( s ) for yellow skin is hampered by two facts: 1 ) this phenotype is not caused by a recent mutation event but instead represents a species difference that may involve multiple substitutions with phenotypic effects , and 2 ) it is not clear whether the red junglefowl allele represents the ancestral or derived state . At present there are 115 fixed nucleotide substitutions between the clusters of yellow skin and white skin sequences; one of these is a missense mutation ( K416N ) but it affects a residue that is not well conserved between species . Sequence data from more distantly related bird species like the zebra finch are required to identify evolutionary conserved regulatory elements where the causal mutation may reside . An obvious topic for future research is to study the role of BCDO2 in carotenoid pigmentation variation in birds , fish , and other vertebrates , including humans . In fact , the presence of yellow fat has been shown to be inherited as a recessive trait in both rabbits [25] and sheep [26] . BCDO2 is now an obvious candidate gene for these traits . This study contradicts the assumption that the red junglefowl is the sole wild ancestor of the domestic chicken [5] and provides the first conclusive evidence that other species have contributed to the domestic chicken genome . We therefore propose that the taxonomy of the domestic chicken should be changed from Gallus gallus domesticus to Gallus domesticus to reflect the polyphyletic origin of chicken [27] . The emerging technologies for total genome resequencing can be readily employed to determine if other parts of the chicken genome also originate from other species of junglefowls . Such regions are expected to be enriched for functionally important variants , like yellow skin , because neutral sequences should have been diluted out during the extensive back-crossing that must have taken place after introgression . It is possible that the introgression of yellow skin was facilitated by the fact that it resides on a microchromosome ( only 6 . 4 Mb in size ) with a high recombination rate , which reduces the amount of genetic material affected by linkage drag .
DNA samples from a pedigree comprising 91 informative meiosis from a W*W/W*Y×W*Y/W*Y backcross , collected by Hendrix Genetics B . V . ( Holland ) , were used for the linkage analysis . DNA samples from various domestic breeds collected by the AvianDiv project [28] were used for IBD mapping together with samples from experimental populations used by the Uppsala group . The origin of samples from different species of junglefowl is shown in Table S1 together with information on the domestic chicken included in this study . Tissues from breast skin and liver used in the expression analysis were sampled from an experimental cross at the INRA experimental station ( Tours , France ) segregating for yellow skin . Tissues were kept in −70°C until the expression analysis was performed . A single nucleotide polymorphism ( SNP ) at position chr24:5 , 237 , 523 ( A→G ) , was genotyped in 91 individuals from the pedigree material provided by Hendrix Genetics B . V by single base extension; primer sequences are given in Table S2 . All other SNP typings were done by pyrosequencing using the Pyro Gold chemistry ( Biotage , Uppsala , Sweden ) and the PCR and sequencing primers are listed in Table S2 . The 23 . 8 kb region was resequenced in eleven chickens . Three of these were expected to be homozygous for yellow skin ( White Leghorn Line 13 from Uppsala , White Leghorn OS strain , and Chinese She-ki ) while four were expected to be homozygous white skin ( a Friesian Fowl and samples from three different subspecies of red junglefowl , Gallus gallus gallus , Gallus gallus jabouillei and Gallus gallus bankiva ) . Furthermore , samples from two grey junglefowls ( Gallus sonneratii ) and one Ceylon junglefowl ( Gallus lafayetii ) , both expected to be homozygous for yellow skin on the basis of leg color , were included . A sample of green junglefowl ( Gallus varius ) was also included in the sequence comparison though we have no reliable information of the leg color of this species . All primers pairs used to generate overlapping PCR amplicons ranging between 700–1000 bps in size are shown in Table S2 and they were designed using the Primer3 software [29] . The same primers were also utilized for sequencing . The sequences were analyzed and edited with Codon Code Aligner ( CodonCode , Dedham , MA ) . The sequence from the red junglefowl used to generate the chicken genome sequence was downloaded from GenBank and used as a reference in the alignment . The D-loop of mitochondrial DNA was PCR-amplified and sequenced from a number of domestic and wild chickens . The PCR primers GalCR_L16750 and GalCR_rev [5] were used to amplify a 1325 bp fragment . Sequencing primers are listed in Table S2 . Tissues ( skin and liver ) were collected from animals being yellow skin heterozygotes ( confirmed by genotyping of SNP chr24:6 , 268 , 434 ) . Total RNA was extracted from skin and liver with TRIzol ( Invitrogen , Frederick , MO , USA ) and then treated with DNA-free™ ( Applied Biosystems , Foster City , CA , USA ) to remove DNA contamination . The RNA quality was controlled using an Agilent 2100 Bioanalyzer ( Agilent Technologies Inc . , Santa Clara , CA , USA ) . The First-Strand cDNA synthesis kit ( GE Healthcare Bio-Sciences ) was used for cDNA-synthesis with the pd ( N ) 6 random hexamers . PCR amplifications were done over intron/exon borders with the ex6pf_m13 , ex6pr primers , and a 5′ biotinylated M13 primer ( Table S2 ) . The relative expression of the W*W and W*Y transcripts was scored by analyzing the SNP at position chr24:6 , 268 , 434 by pyrosequencing . Primer ex6p_seq was used as the sequencing primer and all steps were performed according to manufacturer's protocol ( Biotage AB ) . All samples except skin sample 3 were analyzed in triplicates . QTL analysis was performed using a red junglefowl×White Leghorn intercross on a series of traits including , growth , egg production , skeletal traits , and behavioral traits . Full descriptions of traits are given elsewhere [18] , [30] , [31] . Single marker analysis was performed using a fully informative SNP at the BCDO2 locus ( chr24:6 , 273 , 428 ) . A general linear model was used to test for significant genotypic differences , with the fixed factors of batch and sex being included for all traits , whilst in the case of morphological traits , body mass was also included as a covariate . Multiple testing correction due to multiple marker intervals was not needed , due to only one marker being tested , though multiple testing of many traits remains an issue . Neighbor-Joining trees were constructed from a total of seven independent loci , including BCDO2 , from sequences aligned by eye using Se-Al [32] . In the case of the control region , an additional phylogenetic analysis was performed using MrBayes 3 [33] . Parameter estimates ( including posterior probabilities ) and consensus trees resulting from several independent MrBayes runs of at least 10 million generations each were recorded and contrasted . The posterior probabilities listed on the tree in Figure S1 represent the lowest recorded values amongst all the runs . The MrBayes analysis was run firstly using Japanese Quail ( AP003195 ) as an outgroup , and then without an outgroup . The resulting topologies of the trees were identical . Previously published sequence data [7] for five nuclear markers ( four distinct CR1 repeat regions and OTC intron 9 ) were harvested from GenBank ( Figures S2 , S3 , S4 , S5 , and S6 ) . The analysis of the mitochondrial control region was performed using 20 sequences generated as part of this study combined with 61 previously published sequences representative of the variation found in modern domestic stocks and in the four species of Gallus . Lastly , the analysis of the 23 . 8 kb region encompassing the BCDO2 gene consisted entirely of newly generated sequences and the publicly available genome sequence . All samples that carried the W*Y allele possessed a 598 bp insertion absent in the published RJF sequence ( UCDI ) and in most of our red junglefowl sequences . The samples GGA15 ( Gallus gallus jaboiuellei ) and GJF ( green junglefowl ) carried this fragment located at nucleotide position 6 , 283 , 696 on chromosome 24 suggesting that it represent a deletion that happened in the red junglefowl lineage . The sequences OS and L13 were identical . Partial resequencing of the 23 . 8 kb region was also conducted using a divergent set of domestic breeds homozygous for W*Y which revealed SNPs in complete linkage disequilibrium ( data not shown ) . A comparison of the topologies derived from the control region of the mitochondrial genome and from the nuclear markers not on chromosome 24 , revealed that while Ceylon and grey junglefowl always clustered together , sequences derived from green junglefowl sometimes clustered with Ceylon and grey junglefowl , and sometimes clustered with red junglefowl . This inconsistency is probably the result of the relatively few number of basepairs used in the alignments of each of the markers . Seven of the nine grey junglefowls used in this study ( three from GenBank and six novel sequences ) possessed an identical mtDNA control region haplotype that matched one of the most common and globally distributed domestic haplotypes ( E1 ) as identified by Liu et al . [3] , and a single other grey junglefowl sequence differed by only a single base pair ( Figure S1 ) . This evidence suggests that many , if not most of grey junglefowl populations in zoos and in captivity outside of India are descended from ancestors who were mated with domestic hens . Only a single grey junglefowl sequence obtained from a Delhi National Park specimen possessed an mtDNA sequence that was phylogenetically distinct from red junglefowl sequences; the sequence of this bird was replicated in the Uppsala lab . This sample is more likely to be representative of the grey junglefowl not only because of its position on the phylogenetic tree ( more closely related to the Ceylon junglefowl sequences ) , but also because it possesses a 62 bp insert within the control region , a trait shared only by Ceylon junglefowl and not found in any domestic chicken or other species of Gallus sequenced thus far . 23 . 8 kb of the BCDO2 region was generated from this sample which was found to be heterozygous , one allele of which most closely matched GryJF_04-7 and another which more closely matched TY_216372 . The differing nucleotide positions across this region are shown in Figure S7 . Nishibori et al . [7] [7] concatenated all the CR1 sequences into one alignment and presented a single tree . Because these sequences rested on different locations within the chicken genome ( CR1a , b , c , d , and e on Chr . 24 , 1 , 1 , 5 , and 1 , respectively ) , each CR1 sequence was used here to produce a single tree ( see Figures S2 , S3 , S4 , S5 , and S6 ) in order to identify possible regions of introgression into the G . gallus genome by other species of the genus Gallus . Information on the chicken genome sequence is available at http://www . genome . ucsc . edu . The sequence data presented in this paper have been submitted to GenBank with the following accession numbers EU329393–EU329413 and EU334146–334166 .
|
Many bird species possess yellow skin and legs whereas other species have white or black skin color . Yellow or white skin is due to the presence or absence of carotenoids . The genetic basis underlying this diversity is unknown . Domestic chickens with yellow skin are homozygous for a recessive allele , and white skinned chickens carry the dominant allele . As a result , chickens represent an ideal model for analyzing genetic mechanism responsible for skin color variation . In this study we demonstrate that yellow skin is caused by regulatory mutation ( s ) that inhibit expression of the beta-carotene dioxygenase 2 ( BCDO2 ) enzyme in skin , but not in other tissues . Because BCDO2 cleaves colorful carotenoids into colorless apocarotenoids , a reduction in expression of this gene produces yellow skin . This study also provides the first conclusive evidence of a hybrid origin of the domestic chicken . It has been generally assumed that the red junglefowl is the sole ancestor of the domestic chicken . A phylogenetic analysis , however , demonstrates that though the white skin allele originates from the red junglefowl , the yellow skin allele originates from a different species , most likely the grey junglefowl . This result significantly advances our understanding of chicken domestication .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/animal",
"genetics",
"genetics",
"and",
"genomics/comparative",
"genomics",
"genetics",
"and",
"genomics/functional",
"genomics",
"genetics",
"and",
"genomics/gene",
"expression",
"evolutionary",
"biology/evolutionary",
"and",
"comparative",
"genetics",
"genetics",
"and",
"genomics/gene",
"function"
] |
2008
|
Identification of the Yellow Skin Gene Reveals a Hybrid Origin of the Domestic Chicken
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.