MBE Advance Access originally published online on October 20, 2006
Molecular Biology and Evolution 2007 24(1):182-191; doi:10.1093/molbev/msl141
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Articles |
Recurrent Positive Selection at Bgcn, a Key Determinant of Germ Line Differentiation, Does Not Appear to be Driven by Simple Coevolution with Its Partner Protein Bam
Department of Molecular Biology and Genetics, Cornell University
E-mail: cfa1{at}cornell.edu.
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Surveys of nucleotide sequence polymorphism in Drosophila melanogaster and Drosophila simulans were performed at 2 interacting loci crucial for gametogenesis: bag-of-marbles (bam) and benign gonial cell neoplasm (bgcn). At the polymorphism level, both loci appear to be evolving under the expectations of the neutral theory. However, ratios of polymorphism and divergence for synonymous and nonsynonymous mutations depart significantly from neutral expectations for both loci consistent with a previous observation of positive selection at bam. The deviations suggest either an excess of synonymous polymorphisms or an excess of nonsynonymous fixations at both loci. Synonymous evolution appears to conform to neutrality at bam. At bgcn, there is evidence of positive selection affecting preferred synonymous mutations along the D. simulans lineage. However, there is also a significantly higher rate of nonsynonymous fixations at bgcn within D. simulans. Thus, the deviation from neutrality detected by the McDonald–Kreitman test at these 2 loci is likely due to the selective acceleration of nonsynonymous fixations. Differences in the pattern of amino acid fixations between these 2 interacting proteins suggest that the detected positive selection is not due to a simple model of coevolution.
Key Words: Drosophila • bam • bgcn • positive selection • nonsynonymous
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
A growing number of studies have revealed that proteins involved in reproduction and gametogenesis tend to evolve rapidly at the amino acid level (e.g., Swanson and Vacquier 2002
; Torgerson et al. 2002; Castillo-Davis et al. 2004; Glassey and Civetta 2004; Swanson et al. 2004; Jagadeeshan and Singh 2005; Nielsen et al. 2005; Civetta et al. 2006). Numerous hypotheses have been proposed to explain the rapid evolution of reproductive proteins, including sperm competition, an arms race between the sexes, maternal–fetal conflicts, immune response in reproductive tracts, and suppression of segregation distortion during gamete formation. In addition, Castillo-Davis et al. (2004) have shown in mammals and nematodes that proteins with functions requiring specific molecular interactions (i.e., protein–protein interactions) are overrepresented in scans for rapidly evolving loci. They believe this observation is due to active or passive coevolution between proteins. In this study, we explore the molecular evolution at 2 genetically interacting loci, bag-of-marbles (bam) and benign gonial cell neoplasm (bgcn), which have specific key functions in regulating and promoting gametogenesis in both sexes (Lavoie et al. 1999; Ohlstein et al. 2000).
In some dipteran species, females have retained gonad stem cell activity into adulthood similar to males (Spradling et al. 2001). When the female stem cell divides, 1 daughter cell remains a stem cell, whereas the other differentiates into a cystoblast. The cystoblast goes through 4 synchronized mitotic divisions with incomplete cytokinesis. Of the resulting 16-cell cyst, only 1 cell will become the egg, whereas the others serve as nurse cells. Proper function of bam and bgcn is required for the initiation of cystoblast differentiation. Although there is no biochemical evidence of physical interaction between Bam and Bgcn, the accumulating evidence for genetic interaction between the 2 proteins (i.e., Bgcn is required for Bam localization and bgcn acts as a dominant enhancer of the bam phenotype) has led researchers to propose an interaction (Lavoie et al. 1999; Ohlstein et al. 2000). Current models of germ line stem cell differentiation suggests that the Bam/Bgcn complex releases the proteins Pumilio and Nanos from mRNAs required for differentiation (Chen and McKearin 2005; Szakmary et al. 2005).
In addition, Bam and Bgcn are known to participate in the assembly of the endoplasmic reticulum–like fusome. The fusome is an organelle required for synchronizing the 4 cystoblast mitotic divisions and connects the cells in the forming cysts. The organelle also aids in the proper transport of maternally inherited components into the oocyte from the nurse cells (León and McKearin 1999; Cox and Spradling 2003) and is necessary for the determination of which cell in the 16-cell cyst will ultimately become the egg (Cooley and Theurkauf 1994; de Cuevas et al. 1997).
In males, when the germ line stem cell divides, 1 daughter cell differentiates into a gonialblast. The gonialblast also undergoes 4 synchronized cell divisions with incomplete cytokinesis producing 16 spermatogonials, all of which differentiate into spermatocytes and eventually sperm. In males, Bam and Bgcn are required for the switch from the spermatogonial program of mitotic divisions to the spermatocyte differentiation and meiotic cell cycle (Gönczy et al. 1997; Fuller 1998; Schulz et al. 2004). The fusome is also a crucial organelle in spermatogenesis, but whereas in oogenesis it is present until the onset of meiosis, the fusome is present throughout the entire process of sperm production and is crucial for the synchronization of the mitotic and meiotic divisions. In addition, it allows for the transfer of gene products between the individual developing spermatids, allowing for the ultimately haploid cells to be phenotypically diploid (Braun et al. 1989).
Given that bam and bgcn are necessary for fundamental developmental switches during gametogenesis, one might expect them to be evolving under a high level of selective constraint. However, evidence for rapid amino acid evolution at bam has recently been documented (Civetta et al. 2006). In this study, we confirm and elaborate on the findings of Civetta et al. (2006) and present evidence of rapid amino acid divergence at bgcn, a genetic interactor of bam. This pattern of molecular evolution appears to have occurred along both the Drosophila melanogaster and Drosophila simulans lineages at bam. At bgcn, the rapid amino acid fixations are more pronounced in D. simulans. These contrasting patterns between bam and bgcn suggest that a simple coevolutionary hypothesis does not explain our results.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Samples
For the bam locus, population samples from the United States and Zimbabwe were surveyed in D. melanogaster, whereas a single US population was surveyed in D. simulans. For bgcn, a US population was surveyed for variation in each species. Collection data for these populations have been reported previously (Aquadro et al. 1988
; Begun and Aquadro 1994; Lazzaro and Clark 2001). For all D. melanogaster populations, stocks made homozygous (Aquadro et al. 1988; Lazzaro and Clark 2001) for the third (for bam) or second (for bgcn) chromosomes were used so heterozygous sites are not a concern. Only 1 population (Pennsylvania, United States) of D. melanogaster was sequenced for bgcn due to the limited availability of extracted second chromosomes lines. For D. simulans, inbred lines were used. Heterozygotes were not observed at bam. However, they were encountered at bgcn, and phase between heterozygous sites was inferred using the program Phase version 2.1 (Stephens et al. 2001; Stephens and Donnelly 2003). One of the inferred haplotypes was randomly chosen and used in subsequent analyses. The sample sizes at bam were 7, 12, and 10 chromosomes for D. melanogaster Zimbabwe, D. melanogaster United States, and D. simulans, respectively. The sample sizes at bgcn were 12 and 10 chromosomes in D. melanogaster United States and D. simulans, respectively. A single sequence from the following species was also obtained at bam: Drosophila yakuba, Drosophila santomea, Drosophila teissieri, Drosophila erecta, Drosophila biarmipes, and Drosophila prostipennis. A single sequence from the following species was obtained for bgcn: D. yakuba, D. santomea, D. teissieri, and D. erecta. All sequences have been deposited with the European Molecular Biology Laboratory/Genbank Data Libraries under accession nos. EF055903–EF055986
[GenBank]
specimens of D. yakuba, D. santomea, D. teissieri, and D. erecta were kindly provided by the lab of A. G. Clark. For D. biarmipes and D. prostipennis, flies were obtained from the Drosophila Species Stock Center (http://stockcenter.arl.arizona.edu/).
Analysis of Sequence Polymorphism Data
At bam, polymorphism data was obtained for an approximately 1.3-kb region, which includes most of the coding region of this locus. At bgcn, polymorphism data was obtained in 2 separate regions. One labeled bgcn 5' is 1.5 kb long and encompasses most of the first exon. The second region, bgcn 3', is 1.7 kb long and covers most of exons 3 through 5. Approximately, 1.8 kb separate bgcn 5' and bgcn 3'.
Sequencing was performed by the Biotechnology Resource Center DNA Sequencing Facility at Cornell University (http://cores.lifesciences.cornell.edu/brcinfo/) using ABI chemistry and separating products on an ABI 3730. Both strands were sequenced in D. simulans. In D. melanogaster, roughly half of the region was sequenced on a single strand for this resequencing effort because the use of homozygous lines meant that we did not have to call heterozygotes. Any region showing ambiguous base calls within an individual were sequenced on both strands. Importantly, for regions where high-quality sequence was obtained for both strands, the sequences were always identical. We are therefore confident in our resequencing data based on a single strand.
Sequences were aligned using MegAlign of the DNASTAR software package and analyzed using the DnaSP 4.0 program (Rozas et al. 2003). This program was used to calculate 
w and 
, estimates of 4Neµ. DnaSP 4.0 was also used to perform the following tests of the neutral theory: Tajima's D (Tajima 1989), Fu and Li's D (Fu and Li 1993), Fay and Wu's H (Fay and Wu 2000), and the McDonald–Kreitman (MK) test (McDonald and Kreitman 1991). P values for the Tajima's D, Fu and Li's D, and Fay and Wu's H test statistics were obtained using the coalescent simulator of DnaSP 4.0 either with or without recombination. For this analysis, we used R = 4Nerm = 113 for both loci, where r is the rate of recombination per basepair per generation and m is the length in basepairs of the region analyzed. These recombination rates roughly corresponds to regional sex-averaged, genetic map–based recombination rates of r = 2.0 x 10–8 recombinants/generation/bp and r = 1.7 x 10–8 recombinants/generation/bp for bam and bgcn, respectively (Hey and Kliman 2002), and a conservative estimate of Ne of 1 x 106 (Kreitman 1983) for D. melanogaster.
Analysis of Sequence Divergence Data
Sequencing was performed as above. For both loci, the majority of the coding region was analyzed. For bam, the region considered in the phylogenetic analysis directly corresponds to the regions surveyed for polymorphism. For bgcn, the phylogenetic analysis covers more of the coding region than surveyed for polymorphisms resulting in only the most 5' and 3' ends of the coding region not being analyzed. All additional species were sequenced on both strands. For bam, polymerase chain reaction amplification was performed using genomic DNA, whereas bgcn was amplified from cDNA. At both loci, heterozygous sites were encountered and were not included in further analyzes (treated as missing data). This treatment of heterozygous sites was only done for the species sequences other than D. melanogaster and D. simulans resulting in this treatment only affecting the phylogenetic analyzes. Protein-based alignments were obtained using the alignment program Tcoffee (Poirot et al. 2004; http://igs-server.cnrs-mrs.fr/Tcoffee/tcoffee_cgi/index.cgi).
We applied the standard implementation of PAML version 3.14 (Yang 1997; Yang and Nielsen 2002). We compared models: M0 versus M3, M7 versus M8, and M8 versus M8a. We also utilized the Genetic Algorithm (GA) branch method of Kosakovsky Pond and Frost (2005; http://www.hyphy.org/gabranch/) to determine if branches across the multiple species phylogenies for these loci have different 
values. The appeal of this method is that branches are not a priori classified into different groups. To determine the relative significance of the GA-branch results, we used the AnalyzeCodonData and BranchClassDNDS analyses of the HYPHY package (Kosakovsky Pond et al. 2005; http://www.hyphy.org/). The HYPHY program was also used to construct trees based on synonymous and nonsynonymous sites.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Table 1 reports the level of variation observed in both species at bam and bgcn. Polymorphic site tables can be found in Supplementary Material online (figs. S1–S4). In general, bgcn harbors more variation than bam. As is typically observed (e.g., Moriyama and Powell 1996
), except for nonsynonymous polymorphisms at bgcn, D. simulans harbors more variation than D. melanogaster. At bam, within D. melanogaster, the Zimbabwe population contains more variation than the US population at synonymous and total, but not nonsynonymous, variation, which was also observed by Civetta et al. (2006) when using a Canadian population and a different chromosomal sample from Zimbabwe.
|
Table 2 shows the results of tests of neutrality that focus on the site frequency spectrum within a population sample. For D. melanogaster, none of the tests depart significantly from neutral expectations at either locus. This is also true for D. simulans except for a significantly positive Fu and Li's D statistic at bgcn 5'. This departure appears to be due to the strong haplotype structure that is often observed in non-African populations in D. simulans (e.g., Begun and Aquadro 1994
; Eanes et al. 1996; Hamblin and Aquadro 1996; Hamblin and Veuille 1999; Labate et al. 1999), resulting in many intermediate frequency variants. This structure is more evident at bgcn 5' than at bgcn 3'. Given that this structure is frequently observed in D. simulans, our significant Fu and Li's test statistic is probably due to demographic forces rather than locus-specific selection. Thus, polymorphism and allele frequency data within D. melanogaster and D. simulans do not reveal departures from neutral evolution. These results at bam are in agreement with the pattern observed by Civetta et al. (2006).
|
However, at both of these putatively interacting proteins, the MK test detects significant deviations from the null hypothesis that the ratio of synonymous polymorphisms to fixed synonymous differences between species should equal the ratio of nonsynonymous polymorphisms to fixed nonsynonymous differences (tables 3 and 4). At bam, these ratios are not equal and the difference is statistically significant. A similar observation has recently been noted for this locus (Civetta et al. 2006
). Using D. yakuba as an outgroup, we inferred the number of synonymous and nonsynonymous fixed differences that occurred on the D. melangaster and D. simulans lineages. Roughly, equal numbers of synonymous and nonsynonymous changes were observed along the 2 lineages. Lineage-specific MK test results remain strongly significant for D. melanogaster and is marginally so in D. simulans, suggesting that the potential positive selection detected by this method has occurred along both species' lineages. This finding is in contrast to Civetta et al. (2006). They did not perform lineage-specific MK analyzes and so interpreted their observation of a slight excess of amino acid fixations along the D. simulans lineage as suggestive of selection specific to this species.
|
|
The MK test is highly significant when the 2 regions of bgcn are tested individually in D. simulans. This is the case with total or lineage-specific divergence, as inferred using D. yakuba as the outgroup. For D. melanogaster, the MK test is significant at bgcn 5' only when total divergence to D. simulans is used, although there remains the trend of an excess of nonsynonymous fixations using lineage-specific divergence (the individual MK test P value is 0.026, but this is not significant with a Bonferroni correction). Pooling of the data for the 2 regions of bgcn and considering total divergence produces a significant test result in both species. The test remains strongly significant for D. simulans, when considering lineage-specific divergence only but is not significant for D. melanogaster.
The deviation detected with the MK test in both species at bam and bgcn suggests either an excess of synonymous polymorphism (due to strong balancing selection or weak deleterious selection on synonymous mutations) or an excess of amino acid fixations (due to positive selection on nonsynonymous mutations). As noted previously, tests of neutrality based on polymorphism failed to reject neutrality. Furthermore, the Bauer DuMont et al. (2004) method fails to detect selection affecting synonymous sites at bam in either species. Pairwise synonymous site divergence at bam is 0.125, which is average for such divergence between D. melanogaster and D. simulans (Betancourt and Presgraves 2002; mean 0.125 ± 0.02, excluding accessory gland proteins). In contrast, pairwise nonsynonymous divergence of 0.068 is outside the 95% confidence interval of such divergence (Betancourt and Presgraves 2002; mean 0.018 ± 0.004, not including accessory gland proteins). In light of these observations, the MK test result at bam appears to be due to an excess of amino acid fixations.
Synonymous sites at bgcn also appear to be evolving neutrally in D. melanogaster based on polymorphism-based tests of neutrality and the method of Bauer DuMont et al. (2004). However, in D. simulans, selection appears to have accelerated preferred synonymous changes (mutations from unpreferred to preferred codons) compared with unpreferred (mutations from preferred to unpreferred codons, P value = 0.008). The corresponding selection against unpreferred synonymous mutations could have led to a MK test departure in D. simulans due to an accumulation of low frequency, weakly deleterious unpreferred synonymous mutations. We therefore removed synonymous and nonsynonymous polymorphisms segregating as singletons in our sample and reapplied the MK test (as suggested by Fay et al. 2001). This revised MK test still shows a significant departure from neutrality (P value < 0.01, both regions combined using species-specific divergence). We also note that nonsynonymous fixations at bgcn have not occurred equally between these species—the majority occur on the D. simulans lineage, resulting in a significant relative rate test (Tajima 1993; 0.001 < P value < 0.005). There is no significant difference between these species in synonymous fixations (0.75 > P value > 0.50). These observations suggest that an excess of nonsynonymous fixations is causing the significant MK test result at bgcn.
Thus, we observe evidence for accelerated amino acid fixations at both genetically interacting loci. However, the MK and relative rate test results suggest that the details of this acceleration are not the same for bam and bgcn. At bam, both D. melanogaster and D. simulans appear to have experienced repeated nonsynonymous fixations, though the D. melanogaster lineage appears to have been affected to a greater extent. In contrast, the acceleration is only significant for bgcn along the D. simulans lineage.
We sequenced bam and bgcn from several additional Drosophila species to investigate the phylogenetic pattern of nonsynonymous fixations at these loci. For bam, we sequenced D. melanogaster, D. simulans, D. yakuba, D. santomea, Drosophila teissieri, D. erecta, D. biarmipes, and D. prostipennis, and for bgcn, we sequenced D. melanogaster, D. simulans, D. yakuba, D. santomea, D. teissieri, and D. erecta. Although we were able to retrieve sequences from Drosophila ananassae and Drosophila pseudoobscura from the database for both loci, due to alignment ambiguities (associated with extreme divergence in the latter species), we only used the D. ananassae bgcn sequence.
The PAML (Yang 1997) program uses maximum likelihood and Bayesian inference to detect positive selection acting on nonsynonymous substitutions in the context of a phylogenetic tree. In its standard form, it assumes a single ratio () of the number of nonsynonymous differences per site to the number of synonymous differences per site (e.g., dN/dS) distribution across a phylogenetic tree. Assuming that synonymous mutations are selectively neutral, an value less than 1.0 indicates selective constraint acting on nonsynonymous changes, and an of 1.0 is indistinguishable from strict neutrality, whereas an greater than 1.0 signals the action of positive selection.
When the standard implementation of PAML is applied to bam and bgcn, there is no indication that positive selection has affected amino acid replacements consistently across the sampled species. In light of the significant relative rate test for nonsynonymous changes between D. melanogaster and D. simulans at bgcn, this negative result should perhaps not be surprising because, as used, PAML cannot detect lineage-specific selection. Further evidence of lineage specificity is shown in figure 1, where the clade containing D. melanogaster and D. simulans has longer branches relative to other species in the nonsynonymous tree but not in the synonymous tree, for both bam and bgcn. In addition, at bgcn, there is also a dramatic difference between D. melanogaster and D. simulans in branch lengths in the nonsynonymous tree.
|
We thus applied the GA-branch method described by Kosakovsky Pond and Frost (2005)
to determine if the ratio significantly varies across branches between these species at bam and bgcn (table 5). The appeal of the GA-branch method is that branches with potentially different ratios are not designated a priori. We first determined whether a model with 1, 2, 3, 4, or a free number of branch rate classes best fit the data by comparing their Akaike's Information Criterion (AIC) values (and, when applicable, log likelihoods). We found that a model with 4 rate branch classes (the highest discrete number the method allows) fit the data best for either locus. These results suggest that the rate of nonsynonymous evolution as compared with synonymous has not been consistent across these species at bam and bgcn. We next included site-to-site variation in either the strength of selection or the mutation pressure in the 4 branch class model. We found at both loci that the best-fitting model in terms of both log likelihood and AIC values is one with 4 branch class ratios and allowing variation in the strength of selection across sites.
|
Table 5 lists the
values predicted for each branch class (along with the lineages placed within each class) for these loci. The best-fitting model for bam (AIC = 10,195.5) has 1 (of 4) branch class with an ratio of 1.49 (indicative of positive selection). In contrast, the ratios are lower at bgcn, with none greater than 1. The D. melanogaster lineage was placed within the highest class at bam but is within one of the lowest classes at bgcn, further illustrating that a simple pattern of coevolution is not evident between these 2 genes.
We next evaluated the significance of the differences between ratios for individual branches at bam and bgcn (figs. 2 and 3, respectively). For bam, the terminal branches leading to D. melanogaster and to D. simulans and the internal branch leading to these species (labeled IB5 in fig. 2) tend to have significantly higher ratios than other branches. On the other hand, the branches leading to the D. santomea to D. erecta clade (IB2) and to the D. yakuba/D. santomea/D. teissieri clade (IB3), have a significantly lower ratio as compared with many other branches in the tree. For bgcn, D. simulans, D. santomea, and IB5 tend to have significantly higher ratios than other branches in the tree. The D. erecta and D. teissieri branches along with IB3 appear to be slow lineages.
|
|
The GA-branch method results suggest that the
ratio has not remained constant across the bam and bgcn phylogenetic trees, with some branches having significantly higher ratios than others. However, this method does not distinguish between relaxation of constraint or positive selection as being the cause of the higher ratios. The significant MK test results for some branches with higher estimated ratios at bam and bgcn support the positive selection hypothesis.
Our results for bam are in agreement with those of Civetta et al. (2006) in suggesting that positive selection has accelerated nonsynonymous fixations at bam along the D. simulans lineage. However, our broader species coverage and application of the GA-branch method to our data revealed that both D. melanogaster and the internal branch leading to these species (IB5) also show evidence of rapid nonsynonymous fixations at bam. At bgcn, the lineage leading to the D. melanogaster/D. simulans clade also appears to show accelerated amino acid sequence evolution. However, the rate has apparently slowed down (relative to this internal branch) along the D. melanogaster lineage, resulting in a striking difference at bgcn in the rate of nonsynonymous evolution between D. melanogaster and D. simulans. The D. santomea branch appears to have an elevated at bgcn.
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
The MK test results suggest that positive selection has accelerated nonsynonymous fixations at both bam and bgcn in the D. melanogaster/D. simulans clade. However, phylogeny-based analyses do not provide statistically significant evidence for positive selection. Several reasons could account for these seemingly conflicting conclusions. First, the significant departure for the MK test could be the result of elevated synonymous polymorphisms (rather than selectively elevated nonsynonymous divergence). However, at bam, synonymous site divergence appears "normal," whereas nonsynonymous divergence is elevated. At bgcn, there is a significant relative rate test for nonsynonymous fixations (but not synonymous fixations). Second, perhaps nonequilibrium demography has led to a false rejection of neutrality in the MK tests. Such effects are not, however, thought to strongly affect the MK test when implemented as simply a test of the neutral theory (Wakeley 2003
; but see also Eyre-Walker 2002). Further, we observe significant departures with the MK test at bam for D. simulans as well as for both the US and Zimbabwe populations of D. melanogaster individually (data not shown). At bgcn, we see the same pattern in both species (although Bonferonni correction makes the D. melanogaster–specific test not significant). There is no reason to assume that each of our individual populations and species has the same demographic history. Thus, the contrast between the MK test results and those from the phylogeny-based likelihood analyses appear to be due to the consideration of different types of data in the tests coupled with lineage-specific (as also noted by Civetta et al. [2006] at bam) and potentially sporadic positive selection.
What selective force(s) could explain the bursts of amino acid fixations at bam and bgcn? These proteins participate in 2 separate functions, gamete differentiation and fusome formation, in which selection could act. The selective force(s) could be related to the direct one-to-one functional interactions among these proteins. Each protein could also be independently responding to external selective pressures. The signal we observe could be the result of a cascade of changes due to the dependence of gametogenesis on specific molecular interactions (i.e., passive coevolution; Castillo-Davis et al. 2004; DePristo et al. 2005; Haag and Molla 2005). Finally, Bam contains a PEST motif, which is associated with proteins that are rapidly degraded (Rogers et al. 1986). Civetta et al. (2006) proposed that positive selection at bam is associated with this degradation as such proteins have been shown to evolve rapidly (Cutter and Ward 2005). However, despite its rapid evolution, bgcn does not appear to have a PEST sequence (evaluated using PestFind; http://emb1.bcc.univie.ac.at/toolbox/pestfind/pestfind-analysis-webtool.htm).
The differences between these 2 loci in their MK test results and in the allocation of lineages into fast- or slow-branch classes by the GA method suggest that the selective pressure is not due to a simple one-to-one coevolutionary interaction between these proteins. Coevolutionary forces (either active between bam and bgcn or passive between these and other gametogenesis proteins) or PEST-associated rapid protein degradation would likely affect all species studied in a roughly similar manner. This pattern is not seen, suggesting that these explanations are not sufficient to explain our observations.
Finally, there could be direct species-specific selective pressure acting on the rate of gametogenesis, which could cause lineage-specific rapid evolution of proteins in this pathway. It is also possible that positive selection has acted through these proteins' involvement in fusome assembly. During spermatogenesis, the fusome, along with the intercellular connections it helps to form, allows for connectivity between the developing spermatids. This connectivity may play a role in restricting segregation distortion (as suggested by de Cuevas et al. 1997). Also, Wolbachia, a cytoplasmic parasite, can cause postfertilization incompatibility and a decrease in sperm production in Drosophila (Snook et al. 2000). Wolbachia has been found to localize in a similar manner as the fusome during sperm production (Clark et al. 2002). One could then hypothesize a selective pressure toward decreasing the detrimental affects of this parasite. In the female, the fusome plays a role in the determination of which cell within the 16-cell cyst will become the oocyte and in early trafficking of cytoplasmic components into it. As such, the fusome may play a role in inhibiting the transmission of maternally inherited cytoplasmic parasites.
Another possible mode of selection involving the fusome in the female is the phenomenon of mitochondrial bottlenecks. Cox and Spradling (2003) show that the first round of entry of mitochondria into the oocyte is fusome dependent. These mitochondria will associate with the germplasm and are thus overrepresented in further generations. Cox and Spradling (2003) have shown that a selective process determines which mitochondria associate with the fusome. At present, we do not have sufficient data to test any of these fusome function–related hypotheses. Analyses of the pattern of molecular evolution at other gametogenesis loci should contribute to a better understanding of the nature of selective pressures acting on this important biological pathway.
|
Supplementary Material |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Supplementary figures S1–S4 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).
|
Acknowledgements |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
We thank Alex Wong for RNA and cDNA preps from multiple Drosophila species. We also thank Sergei Kosakovsky Pond for his assistance in performing the GA-branch and HYPHY analyses. Particularly helpful comments on the manuscript were provided by John Pool, Alex Wong, and Martha Hamblin, and by 3 anonymous reviewers. This research was supported by National Institutes of Health grant number GM36431 to C.F.A.
|
Footnotes |
|---|
John H. McDonald, Associate Editor
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Aquadro CF, Lado KM, Noon WA. (1988) The rosy region of Drosophila melanogaster and Drosophila simulans. I. Contrasting levels of naturally occurring DNA restriction map variation and divergence. Genetics 119:875–888.
Bauer DuMont V, Fay JC, Calabrese PP, Aquadro CF. (2004) DNA variability and divergence at the Notch locus in Drosophila melanogaster and D. simulans: a case of accelerated synonymous site divergence. Genetics 167:171–185.
Begun DJ and Aquadro CF. (1994) Evolutionary inferences from DNA variation at the 6-phosphogluconate dehydrogenase locus in natural populations of Drosophila: selection and geographic differentiation. Genetics 136:155–171.[Abstract]
Benjamini Y and Hochberg Y. (1995) Controlling the false discovery rate—a practical and powerful approach to multiple testing. J R Stat Soc Ser B (Methodol) 57:289–300.
Betancourt AJ and Presgraves DC. (2002) Linkage limits the power of natural selection in Drosophila. Proc Natl Acad Sci 99:13616–13620.
Braun RE, Behringer RR, Peschon JJ, Brinster RL, Palmiter RD. (1989) Genetically haploid spermatids are phenotypically diploid. Nature 337:373–376.[CrossRef][Medline]
Castillo-Davis CI, Kondrashov FA, Hartl DL, Kulathinal RJ. (2004) The functional genomic distribution of protein divergence in two animal phyla: coevolution, genomic conflict, and constraint. Genome Res 14:802–811.
Chen D and McKearin D. (2005) Gene circuitry controlling a stem cell niche. Curr Biol 15:179–184.[CrossRef][ISI][Medline]
Civetta A, Rajakumar SA, Brouwers B, Bacik JP. (2006) Rapid evolution and gene-specific patterns of selection for three genes of spermatogenesis in Drosophila. Mol Biol Evol 23:655–662.
Clark ME, Veneti Z, Bourtzis K, Karr TL. (2002) The distribution and proliferation of the intracellular bacteria Wolbachia during spermatogenesis in Drosophila. Mech Dev 111:3–15.[CrossRef][ISI][Medline]
Cooley L and Theurkauf WE. (1994) Cytoskeletal functions during Drosophila oogenesis. Science 266:590–596.
Cox RT and Spradling AC. (2003) A Balbiani body and the fusome mediate mitochondrial inheritance during Drosophila oogenesis. Development 130:1579–1590.
Cutter AD and Ward S. (2005) Sexual and temporal dynamics of molecular evolution in C. elegans development. Mol Biol Evol 22:178–188.
de Cuevas M, Lilly MA, Spradling AC. (1997) Germline cyst formation in Drosophila. Annu Rev Genet 31:405–428.[CrossRef][ISI][Medline]
DePristo MA, Weinreich DM, Hartl DL. (2005) Missense meanderings in sequence space: a biophysical view of protein evolution. Nat Rev Genet 6:678–687.[Medline]
Eanes WF, Kirchner M, Yoon J, Biermann CH, Wang I-N, McCartney MF, Verrelli BC. (1996) Historical selection, amino acid polmorphism and lineage-specific divergence at the G6pd locus in Drosophila melanogaster and D. simulans. Genetics 144:1027–1041.[Abstract]
Eyre-Walker A. (2002) Changing effective population size and the McDonald-Kreitman test. Genetics 162:2017–2024.
Fay JC and Wu C-I. (2000) Hitchhiking under positive Darwinian selection. Genetics 155:1405–1413.
Fay JC, Wyckoff GJ, Wu C-I. (2001) Positive and negative selection on the human genome. Genetics 158:1227–1234.
Fu Y-X and Li W-H. (1993) Statistical tests of neutrality of mutations. Genetics 133:693–709.[Abstract]
Fuller MT. (1998) Genetic control of cell proliferation and differentiation in Drosophila spermatogenesis. Semin Cell Dev Biol 9:433–444.[CrossRef][ISI][Medline]
Glassey B and Civetta A. (2004) Positive selection at reproductive ADAM genes with potential intercellular binding activity. Mol Biol Evol 21:851–859.
Gönczy P, Matunis E, DiNardo S. (1997) bag-of-marbles and benign gonial cell neoplasm act in the germline to restrict proliferation during Drosophila spermatogenesis. Development 124:4361–4371.[Abstract]
Haag ES and Molla MN. (2005) Compensatory evolution of interacting gene products through multifunctional intermediates. Evolution 59:1620–1632.[CrossRef][ISI][Medline]
Hamblin MT and Aquadro CF. (1996) High nucleotide sequence variation in a region of low recombination in Drosophila simulans is consistent with the background selection model. Mol Biol Evol 13:1133–1140.[Abstract]
Hamblin MT and Veuille M. (1999) Population structure among African and derived populations of Drosophila simulans: evidence for ancient subdivision and recent admixture. Genetics 153:305–317.
Hey J and Kliman RM. (2002) Interactions between natural selection, recombination and gene density in the genes of Drosophila. Genetics 160:595–608.
Jagadeeshan S and Singh RS. (2005) Rapidly evolving genes of Drosophila: differing levels of selective pressure in testis, ovary, and head tissues between sibling species. Mol Biol Evol 22:1793–1801.
Kosakovsky Pond SL and Frost SDW. (2005) A genetic algorithm approach to detecting lineage-specific variation in selection pressure. Mol Biol Evol 22:478–485.
Kosakovsky Pond SL, Frost SDW, Muse SV. (2005) HyPhy: hypothesis testing using phylogenies. Bioinformatics 21:676–679.
Kreitman M. (1983) Nucleotide polymorphism at the alcohol dehydrogenase locus of Drosophila melanogaster. Nature 304:412–417.[CrossRef][Medline]
Labate JA, Bierman CH, Eanes WF. (1999) Nucleotide variation at the runt locus in Drosophila melanogaster and Drosophila simulans. Mol Biol Evol 16:724–731.[Abstract]
Lavoie CA, Ohlstein B, McKearin DM. (1999) Localization and function of Bam protein require the benign gonial cell neoplasm gene product. Dev Biol 212:405–413.[CrossRef][ISI][Medline]
Lazzaro BP and Clark AG. (2001) Evidence for recurrent paralogous gene conversion and exceptional allelic divergence in the Attacin genes of Drosophila melanogaster. Genetics 159:659–671.
León A and McKearin D. (1999) Identification of TER94, an AAA ATPase protein, as a Bam-dependent component of the Drosophila fusome. Mol Biol Cell 10:3825–3834.
McDonald JH and Kreitman M. (1991) Adaptive protein evolution at the Adh locus in Drosophila. Nature 20:652–654.
Moriyama EN and Powell JR. (1996) Intraspecific nuclear DNA variation in Drosophila. Mol Biol Evol 13:261–277.[Abstract]
Nielsen R, Bustamante C, Clark AG, et al. (13 co-authors). (2005) A scan for positively selected genes in the genomes of humans and chimpanzees. Plos Biol. 6:976–985.
Ohlstein B, Lavoie CA, Vef O, Gateff E, McKearin DM. (2000) The Drosophila cystoblast differentiation factor, benign gonial cell neoplasm, is related to DExH-box proteins and interacts genetically with bag-of-marbles. Genetics 155:1809–1819.
Poirot O, Suhre K, Abergel C, O'Toole E, Notredame C. (2004) 3DCoffee: a web server for mixing sequences and structure into multiple sequence alignments. Nucleic Acids Res. 32:W37–W40.
Rodgers S, Wells R, Rechsteiner M. (1986) Amino acid sequences common to rapidly degraded proteins: the PEST hypothesis. Science 234:364–368.
Rozas J, Sánchez-DelBarrio JC, Messeguer X, Rozas R. (2003) DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19:2496–2497.
Schulz C, Kiger AA, Tazuke SI, Yamashita YM, Pantalena-Filho LC, Jones DL, Wood CG, Fuller MT. (2004) A misexpression screen reveals effects of bag-of-marbles and TGFß class signaling on the Drosophila male germ-line stem cell lineage. Genetics 167:707–723.
Snook RR, Cleland SY, Wolfner MF, Karr TL. (2000) Offsetting effects of Wolbachia infection and heat shock on sperm production in Drosophila simulans: comparative analyses of fecundity, fertility and accessory gland proteins. Genetics 155:167–178.
Spradling A, Drummond-Barbosa D, Kai T. (2001) Stem cells find their niche. Nature 414:98–104.[CrossRef][Medline]
Stephens M, Smith NJ, Donnelly P. (2001) A new statistical method for haplotype reconstruction from population data. Am J Hum Genet 68:978–989.[CrossRef][ISI][Medline]
Stephens M and Donnelly P. (2003) Inference in molecular population genetics. J R Stat Soc Ser B 62:605–655.[CrossRef]
Swanson WJ and Vacquier VD. (2002) The rapid evolution of reproductive proteins. Nat Rev Genet 3:137–144.[ISI][Medline]
Swanson WJ, Wong A, Wolfner MF, Aquadro CF. (2004) Evolutionary expressed sequence tag analysis of Drosophila female reproductive tracts identifies genes subjected to positive selection. Genetics 168:1457–1465.
Szakmary A, Cox DN, Wang Z, Lin H. (2005) Regulatory relationship among piwi, pumilio, and bag-of-marbles in Drosophila germline stem cell self-renewal and differentiation. Curr Biol 15:171–178.[CrossRef][ISI][Medline]
Tajima F. (1989) Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123:585–595.
Tajima F. (1993) Simple methods for testing the molecular evolutionary clock hypothesis. Genetics 135:599–607.[Abstract]
Torgerson DG, Kulathinal RJ, Singh RS. (2002) Mammalian sperm proteins are rapidly evolving: evidence of positive selection in functionally diverse genes. Mol Biol Evol 19:1973–1980.
Wakeley J. (2003) Polymorphism and divergence for island-model species. Genetics 163:411–420.
Yang Z. (1997) PAML: a program package for phylogenetic analysis by maximum likelihood. Comput Appl Biosci 13:555–556.
Yang Z and Nielsen R. (2002) Codon-substitution models for detecting molecular adaptation at individual sites along specific lineages. Mol Biol Evol 19:908–917.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Llopart and J. M. Comeron Recurrent Events of Positive Selection in Independent Drosophila Lineages at the Spermatogenesis Gene roughex Genetics, June 1, 2008; 179(2): 1009 - 1020. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Haerty, S. Jagadeeshan, R. J. Kulathinal, A. Wong, K. Ravi Ram, L. K. Sirot, L. Levesque, C. G. Artieri, M. F. Wolfner, A. Civetta, et al. Evolution in the Fast Lane: Rapidly Evolving Sex-Related Genes in Drosophila Genetics, November 1, 2007; 177(3): 1321 - 1335. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||