Positive Correlation Between Evolutionary Rate and Recombination Rate in Drosophila Genes with Male-Biased Expression
http://www.100md.com
分子生物学进展 2005年第10期
Department of Biology II, Section of Evolutionary Biology, University of Munich, Munich, Germany
E-mail: parsch@zi.biologie.uni-muenchen.de.
Abstract
Previous studies have shown that genes that are expressed predominantly or exclusively in males tend to evolve rapidly in comparison to other genes. In most cases, however, it is unknown whether this rapid evolution is the result of increased positive (or sexual) selection on male-expressed traits or if it is due to a relaxation of selective constraints. To distinguish between these two possibilities, we analyzed the relationship between the nonsynonymous substitution rate (dN) and local recombination rate for 343 Drosophila genes that were classified as male, female, or nonsex biased in their expression. For the male-biased genes, a positive correlation between dN and recombination rate was observed. This can be explained by an increased rate of adaptive evolution in regions of higher recombination due to a reduction of Hill-Robertson interference. In contrast, the correlation between dN and recombination rate was negative for both female- and nonsex-biased genes, suggesting that these genes are primarily subject to purifying selection, which is expected to be less effective in regions of reduced recombination.
Key Words: adaptive evolution ? Hill-Robertson interference ? sex-biased genes
Sexual dimorphism is common among higher eukaryotes and is thought to result from the differential action of natural (or sexual) selection on individuals of the two sexes. Darwin (1871) proposed that sexual selection, either through direct male-male competition or female mate choice, was responsible for the extravagant secondary sexual characteristics present in the males of many species. The rapid evolution of male reproductive traits also may play a role in the frequent occurrence of hybrid male sterility (Wu and Davis 1993). Modern molecular evolutionary studies suggest that sexual selection may affect a broad spectrum of sex-related genes (Civetta and Singh 1999; Singh and Kulathinal 2000; Swanson and Vacquier 2002). A recent study that used microarray data to identify Drosophila genes with sex-biased expression and compared their rates of evolution between species found that genes with male-biased expression had significantly higher rates of nonsynonymous substitution (dN) than genes with female- or nonsex-biased expression (Zhang, Hambuch, and Parsch 2004). The accelerated evolutionary rate of male-biased genes could have two different explanations. One possibility is that they are subject to less selective constraint than female- or nonsex-biased genes, allowing them to accumulate more neutral (or slightly deleterious) amino acid changes. Alternatively, male-biased genes could be frequent targets of positive (or sexual) selection and thus accumulate more adaptive amino acid changes. An analysis of the available polymorphism data from Drosophila melanogaster supported the latter explanation: male-biased genes did not show an elevated level of nonsynonymous polymorphism as would be expected under the relaxed selective constraint hypothesis but instead showed evidence of being subject to increased positive selection. However, there were several limitations to this analysis that made this conclusion questionable. For example, the polymorphism comparison used a small number of sex-biased genes that were collected from a survey of the literature and thus included only a small fraction of the sex-biased genes in the genome. Furthermore, these genes likely were not a random sample because some were investigated with an a priori expectation of positive selection.
A further way to distinguish the evolutionary forces responsible for the increased dN in genes with male-biased expression is to examine the relationship between dN and local recombination rate. If most amino acid replacements are adaptive, then a positive correlation between dN and recombination rate is expected. This is because positive selection is more effective in regions of higher recombination due to a relaxation of Hill-Robertson interference among selected sites (Hill and Robertson 1966; Marais and Charlesworth 2003). In contrast, if most amino acid replacements are neutral (or slightly deleterious), then there should be no correlation (or a negative correlation) between dN and recombination rate. This is because purifying selection is less effective in regions of lower recombination for the same reason given above.
Betancourt and Presgraves (2002) investigated the relationship between dN and recombination rate for 255 orthologous genes of D. melanogaster and Drosophila simulans. They found that genes in regions of high recombination had significantly higher dN than those in regions of low recombination. Their study focused primarily on a subset of male-expressed genes, namely, those encoding accessory gland proteins (Acps), and suggested that most of the amino acid replacements in these genes were adaptive. When the Acp genes were removed from the analysis, there was no longer a significant difference in dN between genes in regions of high and low recombination. Recently, Marais et al. (2004) investigated the relationship between dN and recombination rate for 630 genes compared between D. melanogaster and Drosophila yakuba. For this data set, which was not enriched for male-expressed genes, a slightly negative correlation between dN and recombination rate was observed. The results of the two above studies suggest that the evolution of most genes is governed primarily by purifying selection but that positive selection is responsible for the rapid evolution seen in male Acp genes. To investigate whether such adaptive evolution occurs in male-biased genes, in general, we classified the genes used in Marais et al. (2004) as male, female, or nonsex biased in their expression and examined the relationship between dN and recombination rate separately for each group.
Our final data set consisted of 343 genes for which both the ratio of male to female expression (Parisi et al. 2003; Ranz et al. 2003) and an estimate of the local recombination rate were available (Marais et al. 2004). Because comparable estimates of gene expression and local recombination rates were not available for D. yakuba, all our estimates are for D. melanogaster. The average dN of the male-biased genes was about fourfold greater than that of the female-biased genes and about twofold greater than that of the nonsex-biased genes (table 1). Comparisons of dN for all pairwise combinations of the three groups were significant (Mann-Whitney test, P 0.01), and these results held whether a 1.5-fold, twofold, or threefold expression difference between the sexes was used to define genes as sex biased. Similarly, the synonymous substitution rate (dS) was significantly higher for male-biased genes than for female- and nonsex-biased genes over all cutoffs (P 0.01). These results cannot be explained solely by an increased mutation rate in male-biased genes because dN/dS for male-biased genes was also significantly higher than that of both female- and nonsex-biased genes (P 0.05; with the exception of male vs. nonsex biased at 1.5-fold where P = 0.18). The observed differences in evolutionary rate are consistent with those reported previously for a partially overlapping set of genes compared between D. melanogaster and D. yakuba (Zhang, Hambuch, and Parsch 2004). In total, 161 genes are common to the two studies.
Table 1 Evolutionary Rates of Genes with Male-, Female-, and Nonsex-Biased Expression
To examine the type of natural selection acting on genes of the three sex-biased expression classes, we looked at the relationship between dN and local recombination rate within each class of genes. For this, we used eight different estimators of local recombination rate (Marais et al. 2004). All eight estimators were positively correlated with dN for male-biased genes, and in most cases this correlation was significant (table 2). In contrast, all eight estimators were negatively correlated with dN for female- and nonsex-biased genes, and this correlation was significant for many of the estimators (table 2). Furthermore, the correlation coefficients of the male- and female-biased genes were significantly different (P < 0.05) for all recombination estimators, and those for the male- and nonsex-biased genes were significantly different (P < 0.05) for seven of the eight estimators. Differences between the correlation coefficients of female- and nonsex-biased genes were not significant for any recombination rate estimator. We also observed a positive correlation between dS and recombination rate for the male-biased genes (table 2). A possible explanation for this is that there is an elevated rate of mutation in male-biased genes in regions of high recombination. However, mutational effects alone cannot explain our results because dN/dS is also positively correlated with recombination rate (fig. 1 and table 2). Our results also cannot be explained by differences between X-linked and autosomal genes. Although male-biased genes are significantly underrepresented on the X chromosome, all the above results are unchanged when only autosomal genes are considered (not shown).
Table 2 Correlation Between Substitution Rate and Recombination Rate for Genes with Male-, Female-, and Nonsex-Biased Expression
FIG. 1.— Correlation between dN/dS and local recombination rate (estimated by HKw [table 2] and using the 1.5-fold cutoff [table 1]) for genes with (A) male-biased (R = 0.30, P = 0.02), (B) female-biased (R = –0.19, P = 0.02), and (C) nonsex-biased expression (R = –0.19, P = 0.04). Least-squares linear regression lines are shown.
Our analyses indicate that male-biased genes are subject to different selective forces than female- and nonsex-biased genes. The positive correlation between dN and recombination rate seen for male-biased genes suggests that they are often targets of positive selection, which is expected to be more effective in genomic regions with higher recombination rates due to a reduction of Hill-Robertson interference. In contrast, the negative correlation between dN and recombination rate seen for female- and nonsex-biased genes suggests that these genes are predominantly subject to purifying selection, which is expected to be less effective in regions of lower recombination, allowing fixation of more slightly deleterious mutations. The reduced efficacy of purifying selection in regions of reduced recombination is also expected to affect male-biased genes and would counteract the positive correlation between dN and recombination rate. Thus, a significantly positive correlation, as is seen in our data, is a conservative criterion for the inference of positive selection.
Marais et al. (2004) observed a slightly negative correlation between dN and recombination rate for 630 genes compared between D. melanogaster and D. yakuba. Their analysis, however, did not consider male-, female-, and nonsex-biased genes separately. Because the vast majority of genes are female or nonsex biased (table 1), the negative correlation in these genes would obscure any positive correlation present in the male genes.
Our results are similar to those seen for male-specific Acp genes. However, the genes analyzed here came from a random expressed sequence tag (EST) survey and were classified only by their degree of sex-biased expression. They are not enriched for genes of a particular functional class. In fact, none of the male-biased genes used in our analysis match annotated Acp genes or putative Acp genes identified in an accessory gland-specific EST screen (Swanson et al. 2001). Thus, it appears that rapid evolution due to positive selection is a general feature of male-biased genes and is not limited to a relatively small set of Acp genes.
References
Betancourt, A. J., and D. C. Presgraves. 2002. Linkage limits the power of natural selection in Drosophila. Proc. Natl. Acad. Sci. USA 99:13616–13620.
Civetta, A., and R. S. Singh. 1999. Broad-sense sexual selection, sex gene pool evolution, and speciation. Genome 42:1033–1041.
Darwin, C. 1871. The descent of man, and selection in relation to sex. John Murray, London.
Hill, W. G., and A. Robertson. 1966. The effect of linkage on limits to artificial selection. Genet. Res. 8:269–294.
Marais, G., and B. Charlesworth. 2003. Genome evolution: recombination speeds up adaptive evolution. Curr. Biol. 13:R68–R70.
Marais, G., T. Domazet-Loso, D. Tautz, and B. Charlesworth. 2004. Correlated evolution of synonymous and nonsynonymous sites in Drosophila. J. Mol. Evol. 59:771–779.
Parisi, M., R. Nuttal, D. Naiman, G. Bouffard, J. Malley, J. Andrews, S. Eastman, and B. Oliver. 2003. Paucity of genes on the Drosophila X chromosome showing male-biased expression. Science 299:697–700.
Ranz, J. M., C. I. Castillo-Davis, C. D. Meiklejohn, and D. L. Hartl. 2003. Sex-dependent gene expression and evolution of the Drosophila transcriptome. Science 300:1742–1745.
Singh, R. S., and R. J. Kulathinal. 2000. Sex gene pool evolution and speciation: a new paradigm. Genes Genet. Syst. 75:119–130.
Swanson, W. J., A. G. Clark, H. M. Waldrip-Dail, M. F. Wolfner, and C. F. Aquadro. 2001. Evolutionary EST analysis identifies rapidly evolving male reproductive proteins in Drosophila. Proc. Natl. Acad. Sci. USA 98:7375–7379.
Swanson, W. J., and V. D. Vacquier. 2002. The rapid evolution of reproductive proteins. Nat. Rev. Genet. 3:137–144.
Wu, C. I., and A. W. Davis. 1993. Evolution of postmating reproductive isolation: the composite nature of Haldane's rule and its genetic bases. Am. Nat. 142:187–212.
Zhang, Z., T. M. Hambuch, and J. Parsch. 2004. Molecular evolution of sex-biased genes in Drosophila. Mol. Biol. Evol. 21:2130–2139.(Zhi Zhang and John Parsch)
E-mail: parsch@zi.biologie.uni-muenchen.de.
Abstract
Previous studies have shown that genes that are expressed predominantly or exclusively in males tend to evolve rapidly in comparison to other genes. In most cases, however, it is unknown whether this rapid evolution is the result of increased positive (or sexual) selection on male-expressed traits or if it is due to a relaxation of selective constraints. To distinguish between these two possibilities, we analyzed the relationship between the nonsynonymous substitution rate (dN) and local recombination rate for 343 Drosophila genes that were classified as male, female, or nonsex biased in their expression. For the male-biased genes, a positive correlation between dN and recombination rate was observed. This can be explained by an increased rate of adaptive evolution in regions of higher recombination due to a reduction of Hill-Robertson interference. In contrast, the correlation between dN and recombination rate was negative for both female- and nonsex-biased genes, suggesting that these genes are primarily subject to purifying selection, which is expected to be less effective in regions of reduced recombination.
Key Words: adaptive evolution ? Hill-Robertson interference ? sex-biased genes
Sexual dimorphism is common among higher eukaryotes and is thought to result from the differential action of natural (or sexual) selection on individuals of the two sexes. Darwin (1871) proposed that sexual selection, either through direct male-male competition or female mate choice, was responsible for the extravagant secondary sexual characteristics present in the males of many species. The rapid evolution of male reproductive traits also may play a role in the frequent occurrence of hybrid male sterility (Wu and Davis 1993). Modern molecular evolutionary studies suggest that sexual selection may affect a broad spectrum of sex-related genes (Civetta and Singh 1999; Singh and Kulathinal 2000; Swanson and Vacquier 2002). A recent study that used microarray data to identify Drosophila genes with sex-biased expression and compared their rates of evolution between species found that genes with male-biased expression had significantly higher rates of nonsynonymous substitution (dN) than genes with female- or nonsex-biased expression (Zhang, Hambuch, and Parsch 2004). The accelerated evolutionary rate of male-biased genes could have two different explanations. One possibility is that they are subject to less selective constraint than female- or nonsex-biased genes, allowing them to accumulate more neutral (or slightly deleterious) amino acid changes. Alternatively, male-biased genes could be frequent targets of positive (or sexual) selection and thus accumulate more adaptive amino acid changes. An analysis of the available polymorphism data from Drosophila melanogaster supported the latter explanation: male-biased genes did not show an elevated level of nonsynonymous polymorphism as would be expected under the relaxed selective constraint hypothesis but instead showed evidence of being subject to increased positive selection. However, there were several limitations to this analysis that made this conclusion questionable. For example, the polymorphism comparison used a small number of sex-biased genes that were collected from a survey of the literature and thus included only a small fraction of the sex-biased genes in the genome. Furthermore, these genes likely were not a random sample because some were investigated with an a priori expectation of positive selection.
A further way to distinguish the evolutionary forces responsible for the increased dN in genes with male-biased expression is to examine the relationship between dN and local recombination rate. If most amino acid replacements are adaptive, then a positive correlation between dN and recombination rate is expected. This is because positive selection is more effective in regions of higher recombination due to a relaxation of Hill-Robertson interference among selected sites (Hill and Robertson 1966; Marais and Charlesworth 2003). In contrast, if most amino acid replacements are neutral (or slightly deleterious), then there should be no correlation (or a negative correlation) between dN and recombination rate. This is because purifying selection is less effective in regions of lower recombination for the same reason given above.
Betancourt and Presgraves (2002) investigated the relationship between dN and recombination rate for 255 orthologous genes of D. melanogaster and Drosophila simulans. They found that genes in regions of high recombination had significantly higher dN than those in regions of low recombination. Their study focused primarily on a subset of male-expressed genes, namely, those encoding accessory gland proteins (Acps), and suggested that most of the amino acid replacements in these genes were adaptive. When the Acp genes were removed from the analysis, there was no longer a significant difference in dN between genes in regions of high and low recombination. Recently, Marais et al. (2004) investigated the relationship between dN and recombination rate for 630 genes compared between D. melanogaster and Drosophila yakuba. For this data set, which was not enriched for male-expressed genes, a slightly negative correlation between dN and recombination rate was observed. The results of the two above studies suggest that the evolution of most genes is governed primarily by purifying selection but that positive selection is responsible for the rapid evolution seen in male Acp genes. To investigate whether such adaptive evolution occurs in male-biased genes, in general, we classified the genes used in Marais et al. (2004) as male, female, or nonsex biased in their expression and examined the relationship between dN and recombination rate separately for each group.
Our final data set consisted of 343 genes for which both the ratio of male to female expression (Parisi et al. 2003; Ranz et al. 2003) and an estimate of the local recombination rate were available (Marais et al. 2004). Because comparable estimates of gene expression and local recombination rates were not available for D. yakuba, all our estimates are for D. melanogaster. The average dN of the male-biased genes was about fourfold greater than that of the female-biased genes and about twofold greater than that of the nonsex-biased genes (table 1). Comparisons of dN for all pairwise combinations of the three groups were significant (Mann-Whitney test, P 0.01), and these results held whether a 1.5-fold, twofold, or threefold expression difference between the sexes was used to define genes as sex biased. Similarly, the synonymous substitution rate (dS) was significantly higher for male-biased genes than for female- and nonsex-biased genes over all cutoffs (P 0.01). These results cannot be explained solely by an increased mutation rate in male-biased genes because dN/dS for male-biased genes was also significantly higher than that of both female- and nonsex-biased genes (P 0.05; with the exception of male vs. nonsex biased at 1.5-fold where P = 0.18). The observed differences in evolutionary rate are consistent with those reported previously for a partially overlapping set of genes compared between D. melanogaster and D. yakuba (Zhang, Hambuch, and Parsch 2004). In total, 161 genes are common to the two studies.
Table 1 Evolutionary Rates of Genes with Male-, Female-, and Nonsex-Biased Expression
To examine the type of natural selection acting on genes of the three sex-biased expression classes, we looked at the relationship between dN and local recombination rate within each class of genes. For this, we used eight different estimators of local recombination rate (Marais et al. 2004). All eight estimators were positively correlated with dN for male-biased genes, and in most cases this correlation was significant (table 2). In contrast, all eight estimators were negatively correlated with dN for female- and nonsex-biased genes, and this correlation was significant for many of the estimators (table 2). Furthermore, the correlation coefficients of the male- and female-biased genes were significantly different (P < 0.05) for all recombination estimators, and those for the male- and nonsex-biased genes were significantly different (P < 0.05) for seven of the eight estimators. Differences between the correlation coefficients of female- and nonsex-biased genes were not significant for any recombination rate estimator. We also observed a positive correlation between dS and recombination rate for the male-biased genes (table 2). A possible explanation for this is that there is an elevated rate of mutation in male-biased genes in regions of high recombination. However, mutational effects alone cannot explain our results because dN/dS is also positively correlated with recombination rate (fig. 1 and table 2). Our results also cannot be explained by differences between X-linked and autosomal genes. Although male-biased genes are significantly underrepresented on the X chromosome, all the above results are unchanged when only autosomal genes are considered (not shown).
Table 2 Correlation Between Substitution Rate and Recombination Rate for Genes with Male-, Female-, and Nonsex-Biased Expression
FIG. 1.— Correlation between dN/dS and local recombination rate (estimated by HKw [table 2] and using the 1.5-fold cutoff [table 1]) for genes with (A) male-biased (R = 0.30, P = 0.02), (B) female-biased (R = –0.19, P = 0.02), and (C) nonsex-biased expression (R = –0.19, P = 0.04). Least-squares linear regression lines are shown.
Our analyses indicate that male-biased genes are subject to different selective forces than female- and nonsex-biased genes. The positive correlation between dN and recombination rate seen for male-biased genes suggests that they are often targets of positive selection, which is expected to be more effective in genomic regions with higher recombination rates due to a reduction of Hill-Robertson interference. In contrast, the negative correlation between dN and recombination rate seen for female- and nonsex-biased genes suggests that these genes are predominantly subject to purifying selection, which is expected to be less effective in regions of lower recombination, allowing fixation of more slightly deleterious mutations. The reduced efficacy of purifying selection in regions of reduced recombination is also expected to affect male-biased genes and would counteract the positive correlation between dN and recombination rate. Thus, a significantly positive correlation, as is seen in our data, is a conservative criterion for the inference of positive selection.
Marais et al. (2004) observed a slightly negative correlation between dN and recombination rate for 630 genes compared between D. melanogaster and D. yakuba. Their analysis, however, did not consider male-, female-, and nonsex-biased genes separately. Because the vast majority of genes are female or nonsex biased (table 1), the negative correlation in these genes would obscure any positive correlation present in the male genes.
Our results are similar to those seen for male-specific Acp genes. However, the genes analyzed here came from a random expressed sequence tag (EST) survey and were classified only by their degree of sex-biased expression. They are not enriched for genes of a particular functional class. In fact, none of the male-biased genes used in our analysis match annotated Acp genes or putative Acp genes identified in an accessory gland-specific EST screen (Swanson et al. 2001). Thus, it appears that rapid evolution due to positive selection is a general feature of male-biased genes and is not limited to a relatively small set of Acp genes.
References
Betancourt, A. J., and D. C. Presgraves. 2002. Linkage limits the power of natural selection in Drosophila. Proc. Natl. Acad. Sci. USA 99:13616–13620.
Civetta, A., and R. S. Singh. 1999. Broad-sense sexual selection, sex gene pool evolution, and speciation. Genome 42:1033–1041.
Darwin, C. 1871. The descent of man, and selection in relation to sex. John Murray, London.
Hill, W. G., and A. Robertson. 1966. The effect of linkage on limits to artificial selection. Genet. Res. 8:269–294.
Marais, G., and B. Charlesworth. 2003. Genome evolution: recombination speeds up adaptive evolution. Curr. Biol. 13:R68–R70.
Marais, G., T. Domazet-Loso, D. Tautz, and B. Charlesworth. 2004. Correlated evolution of synonymous and nonsynonymous sites in Drosophila. J. Mol. Evol. 59:771–779.
Parisi, M., R. Nuttal, D. Naiman, G. Bouffard, J. Malley, J. Andrews, S. Eastman, and B. Oliver. 2003. Paucity of genes on the Drosophila X chromosome showing male-biased expression. Science 299:697–700.
Ranz, J. M., C. I. Castillo-Davis, C. D. Meiklejohn, and D. L. Hartl. 2003. Sex-dependent gene expression and evolution of the Drosophila transcriptome. Science 300:1742–1745.
Singh, R. S., and R. J. Kulathinal. 2000. Sex gene pool evolution and speciation: a new paradigm. Genes Genet. Syst. 75:119–130.
Swanson, W. J., A. G. Clark, H. M. Waldrip-Dail, M. F. Wolfner, and C. F. Aquadro. 2001. Evolutionary EST analysis identifies rapidly evolving male reproductive proteins in Drosophila. Proc. Natl. Acad. Sci. USA 98:7375–7379.
Swanson, W. J., and V. D. Vacquier. 2002. The rapid evolution of reproductive proteins. Nat. Rev. Genet. 3:137–144.
Wu, C. I., and A. W. Davis. 1993. Evolution of postmating reproductive isolation: the composite nature of Haldane's rule and its genetic bases. Am. Nat. 142:187–212.
Zhang, Z., T. M. Hambuch, and J. Parsch. 2004. Molecular evolution of sex-biased genes in Drosophila. Mol. Biol. Evol. 21:2130–2139.(Zhi Zhang and John Parsch)