Molecular Evolution of the Avian CHD1 Genes on the Z and W Sex Chromosomes

Corresponding author: Hans Ellegren, Department of Evolutionary Biology, Evolutionary Biology Ctr., Uppsala University, Norbyvägen 18D, SE-752-36 Uppsala, Sweden., hans.ellegren{at}ebc.uu.se (E-mail)

Communicating editor: P. D. KEIGHTLEY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Genes shared between the nonrecombining parts of the two types of sex chromosomes offer a potential means to study the molecular evolution of the same gene exposed to different genomic environments. We have analyzed the molecular evolution of the coding sequence of the first pair of genes found to be shared by the avian Z (present in both sexes) and W (female-specific) sex chromosomes, CHD1Z and CHD1W. We show here that these two genes evolve independently but are highly conserved at nucleotide as well as amino acid levels, thus not indicating a female-specific role of the CHD1W gene. From comparisons of sequence data from three avian lineages, the frequency of nonsynonymous substitutions (K_a) was found to be higher for CHD1W (1.55 per 100 sites) than for CHD1Z (0.81), while the opposite was found for synonymous substitutions (K_s, 13.5 vs. 22.7). We argue that the lower effective population size and the absence of recombination on the W chromosome will generally imply that nonsynonymous substitutions accumulate faster on this chromosome than on the Z chromosome. The same should be true for the Y chromosome relative to the X chromosome in XY systems. Our data are compatible with a male-biased mutation rate, manifested by the faster rate of neutral evolution (synonymous substitutions) on the Z chromosome than on the female-specific W chromosome.

THE underlying factors affecting the molecular evolution of sex-linked genes differ in some important ways as compared to those governing the evolution of autosomal genes. First, the effective population size of sex-linked genes is always smaller than that of autosomal genes, implying different fixation probabilities of a given selection coefficient (CHARLESWORTH et al. 1987 ; LI 1997 ). Second, while autosomal genes spend an equal amount of time in the male as in the female germline, sex-linked genes show a bias with respect to their transmission through the two sexes. The mammalian X chromosome, for example, is two-thirds of the time in the female germline. Moreover, genes from the nonrecombining part of one of the sex chromosomes are exclusively transmitted by a single sex. This means that if the patterns of mutation or selection differ between sexes, sex-linked genes will evolve in a "sex-biased" fashion (MIYATA et al. 1987 ). For instance, several lines of evidence from several organisms indicate that the mutation rate of males is higher than that of females, a situation commonly attributed to the many more mitotic germline cell divisions in spermatogenesis than in oogenesis (MIYATA et al. 1987 ; SHIMMIN et al. 1993 ; ELLEGREN and FRIDOLFSSON 1997 ; HURST and ELLEGREN 1998 ). Moreover, the degree of methylation of CpG sites, which increases the mutability of such sites, may differ between sexes (DRISCOLL and MIGEON 1990 ). Third, while there as yet is no evidence for chromosome-specific mutation rates of vertebrate autosomes, a lowered mutation rate has been suggested for the mammalian X chromosome, which could be adaptive by reducing the effect of slightly deleterious mutations being exposed in hemizygote males (MCVEAN and HURST 1997 ). Fourth, possible dosage and dominance effects might act differentially on genes on sex chromosomes (CHARLESWORTH et al. 1987 ) and their recombination rates may also differ. Thus, a number of sex- or chromosome-specific factors may be manifested in the molecular evolution of sex-linked genes.

Since the selection pressure on individual genes varies enormously, empirically addressing the effects of sex- and chromosome-specific factors in molecular evolutionary processes ideally requires analyses of the same gene exposed to different genomic environments. This is obviously not possible for single-copy genes and most multigene families are either autosomal or sex-linked, at least with respect to expressed gene copies. However, a very special class of genes offers a possibility to study these factors, namely, genes shared between the nonrecombining parts of the two types of sex chromosomes. In principle, sex chromosomes are thought to evolve from an ancestral pair of autosomes, where, following the arrest of recombination, one of the chromosomes gradually becomes degraded and devoid of most genes (CHARLESWORTH 1996 ; RICE 1996 ). Since degradation will in most cases not be complete, a few genes will remain on the smaller sex chromosome (e.g., the mammalian Y chromosome) and will thus be present in a copy both on this and on the larger nondegraded sex chromosome (e.g., the X chromosome). As shown for mammals, some of these genes are associated with male-specific or male-enhancing functions and may actually become silenced or deleted from the X chromosome (GRAVES 1995 ). Others, however, will be expressed from both sex chromosomes. Only a limited number of such genes have yet been identified (LAHN and PAGE 1997 ).

In birds, the female is the heterogametic sex and she has one Z and one W chromosome, whereas the male has two Z chromosomes. Physically, the W chromosome resembles the mammalian Y chromosome in several respects; it is small, gene-poor, and mainly heterochromatic (STEFOS and ARRIGHI 1971 ). Studies of genes shared by the Z and W chromosomes would be important for reasons discussed above, and the avian sex chromosome system makes it possible to distinguish between some sex- and chromosome-specific factors that confound analyses in mammals (cf. CROW 1997 ; ELLEGREN and FRIDOLFSSON 1997 ; LESSELS 1997 ). Moreover, such studies are also motivated by the fact that the role of the W chromosome in avian sex determination is still unclear (ELLEGREN 2000 ). The critical issue is whether it is the W chromosome that is required for female development or if it is the number of Z chromosomes that regulates male development, i.e., a dominance (as in mammals) or a balance (as in Drosophila and Caenorhabditis elegans) mode of genic sex determination. If it is the latter, and circumstantial evidence lends some support to this idea (CREW 1954 ; HALVERSON and DVORAK 1993 ; RAYMOND et al. 1999 ; SMITH et al. 1999 ), the question is what selective constraints act on W-linked genes and why they have been retained on this chromosome.

Two avian genes have recently been shown to exist in a copy on both the Z and the W chromosome, the CHD1Z/CHD1W gene pair (ELLEGREN 1996 ; GRIFFITHS et al. 1996 ; GRIFFITHS and KORN 1997 ; FRIDOLFSSON et al. 1998 ) and the ATP5A1Z/ATP5A1W pair (DVORAK et al. 1992 ; FRIDOLFSSON et al. 1998 ; CARMICHAEL et al. 2000 ). The avian CHD1 genes belong to a family of genes composed of a chromatin organization modifier (chromo) domain, a SNF2-related helicase/ATPase domain, and a DNA-binding domain and the protein has been named CHD to denote these domains. Functional studies in model organisms have indicated that CHD1 alters the chromatin structure and thereby facilitates gene expression (STOKES and PERRY 1995 ; STOKES et al. 1996 ). It is not yet known if avian CHD1Z and CHD1W are functionally differentiated.

In this study, we present a detailed analysis of the molecular evolution of avian CHD1Z and CHD1W genes. Based on sequence data from three avian species, we show that the two genes are highly conserved both with respect to CHD1 genes in other organisms and to each other. However, the two genes appear to evolve independently, without signs of genetic exchange through recombination. CHD1Z has a lower frequency of nonsynonymous (K_a) but a higher frequency of synonymous (K_s) substitutions compared with CHD1W. We attribute these differences to the respective characteristics of effective population size, recombination, and sex-specific mutation rates associated with the two types of sex chromosomes.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

PCR and cloning:
mRNA was prepared from 25 µl of fresh whole blood from one male and one female of Tengmalm's owl (Aegolius funerus) and of cockatiel (Lutino cockatiel), with a Quick Prep Micro mRNA purification kit (Pharmacia Biotech, Piscataway, NJ). The Access reverse transcriptase PCR (RT-PCR) system (Promega, Madison, WI) was used with 1/500 of each mRNA preparation, together with the primer combinations described below, to amplify overlapping fragments of the CHD1 genes. Obtaining CHD1Z was straightforward since amplification of male mRNA yields only this gene, even when using primer sequences conserved between CHD1Z and CHD1W. The following five pairs were used to amplify CHD1Z: 1090F (ELLEGREN 1996 ) and 2128R (FRIDOLFSSON et al. 1998 ), 1628F (FRIDOLFSSON et al. 1998 ) and 2469R (ELLEGREN 1996 ), 2421F (ELLEGREN 1996 ) and 3112R (ELLEGREN and FRIDOLFSSON 1997 ), 2895F (5'-CGGCTAGTCACAAAAGGATC-3') and 3681R (ELLEGREN and FRIDOLFSSON 1997 ), and finally P3 (GRIFFITHS and TIWARI 1995 ) and 4104R (ELLEGREN 1996 ).

Specific amplification of CHD1W in female birds is complicated by the fact that CHD1Z and CHD1W are very similar and are both expressed in females. We used a combined strategy of W-specific primers (underlined below), on the basis of sequence information from chicken CHD1W, and single-strand conformation polymorphism (SSCP) analysis to identify CHD1W products in amplifications of female DNA. The primers used were: 1275F (ELLEGREN 1996 ) and 1869R (5'-CATCCATTCATGAGTTCTTAT-3'), 1628F and 2469R, 2421F and 3112R, 2987F (ELLEGREN and FRIDOLFSSON 1997 ) and 3829R (5'-GCCAACTCTTCTTCGTGAGAA-3'), and 3468F (ELLEGREN 1996 ) and 4105R. RT-PCR conditions were 48° for 45 min, then an initial denaturation step of 94° for 2 min followed by a 10-cycle touchdown profile consisting of 94° for 30 sec, 60–50° (lowering the temperature by 1° per cycle) for 30 sec, and 68° for 1 min. Then 30 cycles of the same profile were run at a constant annealing temperature of 50°, and a final extension step of 68° for 10 min was added after the last cycle.

Amplification products were separated by agarose gel electrophoresis (1.5% agarose, Sea Kem) and fragments of the expected size were excised and purified (Qiaex II gel extraction kit, QIAGEN, Hilden, Germany) and ligated into pGEM-T vector (pGEM-T easy vector systems, Promega). For CHD1W, 10 clones of each ligation were reamplified with the same primers and were analyzed with SSCP, together with clones known to contain the Z copy of the fragment (i.e., amplified from males). Clones containing CHD1W could thereby be identified on the basis of the contrasting SSCP patterns of CHD1Z and CHD1W sequences. Clones were sequenced with vector primers using BigDye terminator cycle sequencing chemistry (Perkin Elmer, Norwalk, CT), followed by analysis on an ABI377 automated sequencing instrument (Perkin Elmer, Foster City, CA). The fact that we used overlapping fragments allowed us to ensure that correct clones had always been identified.

Genes were named with a prefix denoting the Latin name of the species of origin (chicken, Gg; Tengmalm's owl, Af; cockatiel, Lc). For use in analyses we obtained from GenBank chicken CHD1Z (AF004397), mouse (Mm, L10410), human (Hs, AF006513), Drosophila melanogaster (Dm, X99021), Saccharomyces cerevisiae (Sc, L10718), and Arabidopsis thaliana (At, AC007209) CHD1 gene sequences. Sequences obtained in this study have been deposited in GenBank under accession nos. AF181824, AF181825, AF181826, AF181827, AF181828.

Sequence analysis:
Contigs of the coding sequence of CHD1Z and CHD1W from each species were constructed using Sequencher 3.0 (Gene Codes, Ann Arbor, MI). Avian sequences were aligned with Sequence Navigator (Applied Biosystems, Foster City, CA) and MEGA (KUMAR et al. 1993 ) was used for translation and analyses of amino acid (aa) distances and base composition. Phylogenetic analyses were done by maximum parsimony (MP) and maximum likelihood (ML) as implemented in PAUP* 4.0b2A (SWOFFORD 1998 ). PHYLIP version 3.5c (FELSENSTEIN 1991 ) was used for UPGMA clustering of sequences based on synonymous substitutions. The frequency of synonymous and nonsynonymous substitutions and their standard errors were calculated by combining the information from twofold and fourfold degenerate sites and using the Kimura two-parameter model to correct for multiple hits (LI 1993 ; PAMILO and BIANCHI 1993 ). Patterns of variation in the K_a/K_s ratio across genes were calculated by dividing the gene into 18 nonoverlapping sections, each containing 51 codons. Spearman rank correlation was used to test if patterns were significantly repeatable. To test for positive selection in individual CHD1W lineages or among sites, the program codeml included in PAML (YANG 1999 ) was used. Analysis of CpG sites followed the method described by SMITH and HURST 1999 , where sites on both strands were included in the analysis. In several analyses, we present means of K_a and K_s in comparisons of different CHD1Z and CHD1W sequences. However, since only three avian species were studied, it should be noted that the three possible comparisons (chicken vs. Tengmalm's owl, chicken vs. cockatiel, and Tengmalm's owl vs. cockatiel) do not represent independent observations.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Independent evolution of CHD1W and CHD1Z:
Based on overlapping fragments amplified by RT-PCR of mRNA prepared from blood, we sequenced 2754 bp of the coding region of the CHD1Z and CHD1W genes from two divergent bird species, Tengmalm's owl and cockatiel. This continuous region covers most of the three functional domains of the CHD protein, i.e., the chromo domain, the helicase domain, and the DNA-binding domain. The obtained sequences could be aligned with chicken CHD1Z (GRIFFITHS and KORN 1997 ) and CHD1W (ELLEGREN 1996 ) without gaps.

As a starting point for further analysis, we first asked whether CHD1Z and CHD1W genes are evolving independently. Phylogenetic analysis with both MP and ML, using mouse and human CHD1 as outgroups, clustered the three CHD1Z and the three CHD1W genes separately (Fig 1). The ML tree has a stronger bootstrap support (84/100) than the MP tree (66/100), which is not unexpected given that maximum-likelihood analysis is less sensitive to long-branch attraction (HUELSENBECK 1997 ; the outgroups are only distantly related to birds and the branch leading to GgCHD1Z is the longest within the avian CHD1 tree; moreover, the rate of evolution varies among the branches; see below). The best alternative MP tree (two steps longer) places the root on the GgCHD1Z branch, and the best alternative ML tree (ln L= -5.34) places the root on the AgCHD1Z/LcCHD1Z branch; both alternatives are indeed unlikely. These results indicate that the CHD1Z and CHD1W genes of the three avian lineages under study have evolved without signs of genetic exchange (e.g., through recombination) between the Z and the W chromosomes. Importantly, the respective molecular evolution of CHD1Z and CHD1W should therefore reflect the intrinsic and different evolutionary forces operating on the two sex chromosomes.

View larger version (20K): In this window In a new window Download PPT slide   Figure 1. Phylogenetic tree showing the relationship between six avian CHD1 gene sequences, using human and mouse CHD1 as outgroups. Numbers indicate bootstrap support for branches using maximum-parsimony and maximum-likelihood analysis (1000 replicates in each case).

High degree of amino acid conservation in CHD1 genes:
The frequency of aa replacements between different gene copies was derived from alignments of avian CHD1Z and CHD1W sequences and of CHD1 from mouse, Drosophila, yeast, and Arabidopsis (Table 1). While avian sequences could be aligned to the mouse sequence without gaps, gaps had to be introduced relative to the more distantly related species. Overall levels of conservation were very high, with, for instance, about five replacements per 100 sites between avian and mouse genes.

Comparisons of avian CHD1Z and CHD1W aa sequences revealed that the two proteins are very similar to each other (mean = 3.2 ± 0.6 aa replacements per 100 sites, range 2.5–3.9; Fig 2), suggesting shared functional properties. Within the respective class of genes, CHD1Z proteins (mean = 1.2 ± 0.1, range 1.2–1.3) are more slowly evolving than CHD1W proteins (3.4 ± 0.5, range 2.8–3.7).

View larger version (171K): In this window In a new window Download PPT slide   Figure 2. Amino acid alignment of avian CHD1 genes, with mouse CHD1 as master sequence. Identical positions are denoted by dots and positions for which data are lacking are denoted by dashes. There are no gaps. Positions are numbered according to the complete aa sequence of mouse. Known functional domains or motifs are boxed.

K_a and K_a/K_s ratios of avian CHD1Z and CHD1W genes:
In accordance with the aa data, K_a was lower for CHD1Z (mean = 0.81 ± 0.08 nonsynonymous nucleotide substitutions per 100 sites) than for CHD1W (1.55 ± 0.30; Table 2), which in turn was only marginally less than that for CHD1Z vs. CHD1W (1.85 ± 0.31). However, since the overall mutation rate may differ between the Z and W chromosomes (ELLEGREN and FRIDOLFSSON 1997 ), a more appropriate measure of the evolutionary forces operating on CHD1Z and CHD1W should be their K_a/K_s ratios. Mean K_a/K_s for CHD1Z (0.037 ± 0.01) was considerably lower than for CHD1W (0.11 ± 0.01).

Selective forces upon replacement substitutions can obviously be different for different parts of a gene, leading to variation in the pattern of K_a/K_s across genes (ALVAREZ-VALIN et al. 1998 ). Repeatability of K_a/K_s patterns in comparisons of independent pairs of gene lineages is an indication of nonrandom substitution rates (SMITH and HURST 1998 ) and is suggestive of different gene copies sharing functional properties. The patterns of K_a/K_s variation across avian CHD1 genes were roughly similar in the three possible comparisons of CHD1Z and CHD1W genes (Fig 3). For instance, K_a/K_s was particularly low in the 3' end of the helicase domain of both CHD1Z and CHD1W. Repeatability was statistically significant for Tengmalm's owl vs. cockatiel (R_s = 0.60, P = 0.013), but not so in the two other comparisons.

View larger version (21K): In this window In a new window Download PPT slide   Figure 3. Variation in Ka/Ks across CHD1Z (solid squares) and CHD1W (open squares) genes. Each data point represents 102 codons, with an overlapping window of 51 codons. (a) Tengmalm's owl vs. chicken; (b) Tengmalm's owl vs. cockatiel; and (c) chicken vs. cockatiel.

Although the fact that K_a/K_s never exceeded 0.35 suggests an absence of positive selection, a higher K_a/K_s ratio in CHD1W than in CHD1Z genes might be indicative of adaptive changes in individual lineages or in parts of the CHD1W gene. To investigate this further, we used a likelihood-ratio test implemented in PAML (YANG and NIELSEN 1998 ; YANG 1999 ). However, this failed to reject a null hypothesis of equal K_a/K_s ratios in individual lineages, tested in all possible topologies of CHD1W trees [2l = 0.92, d.f. = 2, not significant (NS)]. Similarly, a likelihood-ratio test failed to reject a null hypothesis of equal K_a/K_s ratios among sites (2l = -27, d.f. = 2, NS).

A higher K_s in CHD1Z than in CHD1W:
The frequency of synonymous substitutions (K_s) was higher for CHD1Z (mean = 22.70 ± 6.62) than for CHD1W (mean = 13.48 ± 2.06; Table 3), which contrasts to the situation for K_a. This indicates an underlying sex difference in the mutation rate, assuming that synonymous substitutions in CHD1Z and CHD1W are selectively neutral or are at least under the same constraints. Pairwise comparisons of K_s revealed estimates of the male-to-female mutation rate ratio (_m) of 2.1 ± 0.3 (Tengmalm's owl vs. chicken), 2.1 ± 0.3 (cockatiel vs. chicken), and 1.5 ± 0.2 (cockatiel vs. Tengmalm's owl). A mean value of _m 1.7 was estimated from the branch lengths of a dendrogram based on K_s distances (cf. SHIMMIN et al. 1993 ). Since the phylogenetic relationship of the CHD1W genes was unresolved, this mean value is only an approximation.

Low influence of CpG sites on K_s:
The GC content of CHD1Z (mean = 40.5 ± 0.1) and CHD1W genes (39.1 ± 0.1) was lower than an average of 53.2% estimated from 399 chicken genes (OLIVIER and MARIN 1996 ), but did not differ between the two types of genes (² = 0.04, NS). The GC3 content was even lower (CHD1Z, 36.0 ± 0.5; CHD1W, 33.0 ± 0.4; chicken average, 69.4%; BERNARDI et al. 1988 ), but again did not differ between CHD1Z and CHD1W (² = 0.20, NS). The observed number of CpG sites was about five times lower than expected based on base composition in both CHD1Z (ratio of observed/expected = 0.17) and CHD1W (0.20). This underrepresentation is of the same magnitude as the average for genes in the human genome (SCHORDERET and GARTLER 1992 ).

In separate analyses of CHD1Z and CHD1W, we counted the number of synonymous and nonsynonymous sites where all three sequences had a CpG dinucleotide. This number was compared to the number of sites where at least one sequence had a TpG dinucleotide while the other/s had a CpG dinucleotide, i.e., possible cases of C–T transitions at methylated CpG sites. Since both the total number of CpG sites (CHD1Z, 15; CHD1W, 15) and the number of sites with possible C–T transitions (CHD1Z, 4; CHD1W, 5) were low, and did not differ between genes, we conclude that methylated CpG sites seem not to affect the molecular evolution of CHD1Z and CHD1W evolution in a contrasting way.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Very few aa changes distinguish avian CHD1Z and CHD1W proteins (eight positions represent fixed differences, six of which are conservative changes). Similarly, comparisons of eukaryotic CHD1 genes, including avian CHD1Z and CHD1W, reveal extensive conservation, particularly in the functional domains. For instance, only one fixed amino acid difference distinguishes birds from mammals over a region of 180 aa residues in the helicase domain (Fig 2). In fact, the helicase domain is highly conserved even between different members of the CHD gene family (WOODAGE et al. 1997 ), indicating strong functional constraints. The DNA-binding activity of the CHD1 protein has been located to a domain of 229 aa residues and within this region a sequence of 11 aa is essential for DNA binding by A · T minor-groove interactions (STOKES and PERRY 1995 ). This sequence motif is identical between avian and mouse CHD1 genes. Overall, this suggests (i) that CHD1Z and CHD1W share similar functional properties and (ii) that this function should be more or less the same as in other organisms.

Comparative analyses of nonsynonymous substitution rates are preferably made using the K_a/K_s ratio to account for local variation in the mutation rate. In our study, we found K_a/K_s to be higher for CHD1W (0.11) than for CHD1Z (0.04), which in turn was higher than for CHD1 in mammals (0.025). Since K_a/K_s for CHD1W is <<1, which is the strict requirement for demonstration of positive selection, we found no overall suggestion that CHD1W would be rapidly diverging in an adaptive way. Likelihood-ratio tests similarly failed to detect signs of positive selection in terms of variation in K_a/K_s among CHD1W lineages or among CHD1W sites. Moreover, the patterns of K_a/K_s variation across the gene were similar between CHD1Z and CHD1W genes. This, together with the high degree of aa conservation seen between CHD1Z and CHD1W, strongly argues against a female-specific role of CHD1W. In fact, it might be argued that CHD1Z and CHD1W act in concert and in a sense should be seen as allelic variants of the same functional protein. It should be noted that positive selection has been recognized in male-specific and Y-linked sequences in mammals. For example, the mammalian SRY gene shows a K_a/K_s ratio of 1.3 (TUCKER and LUNDRIGAN 1993 ; WHITFIELD et al. 1993 ).

We argue that the difference in K_a/K_s between CHD1Z and CHD1W is associated with differences in effective population size and recombination characteristics of the two types of sex chromosomes. First, selection is more effective in removing slightly deleterious mutations in a population of larger size (NEI 1970 ; LI 1997 ). Other factors being equal, this should imply that such mutations are more easily removed from the Z chromosome since its effective population size is three times that of the W chromosome. Second, since most parts of the W chromosome do not recombine and are thus clonally transmitted, slightly deleterious mutations should be expected to accumulate faster than on the Z chromosome (CHARLESWORTH 1996 ; RICE 1996 ). The expectation is in both cases a higher K_a/K_s ratio on the W chromosome than on the Z chromosome, as we observe. The same should be true for genes on the Y (analogous to W) and X (analogous to Z) chromosomes of mammals and is indeed supported by available data. The K_a/K_s ratio is higher for Ube1y (0.19) than for Ube1x (0.0; CHANG and LI 1995 ), for ZFY (0.42) than for ZFX (0.13; PAMILO and BIANCHI 1993 ), and for SMCY (0.17) than for SMCX (0.02; AGULNIK et al. 1997 ).

In contrast to the situation for K_a, K_s was higher for CHD1Z than for CHD1W. From a similar observation based on partial sequence data, we recently interpreted this as evidence for a male-biased mutation rate, given that W is exclusively transmitted through the female germline (ELLEGREN and FRIDOLFSSON 1997 ). Applying the formula of MIYATA et al. 1987 , present data suggest a male bias in the mutation rate of _m 1.7 in the lineages studied, which is lower than our previous estimate of _m 3.9 derived from the coding regions of CHD1 genes of two passerine bird species (ELLEGREN and FRIDOLFSSON 1997 ). It is not clear if this suggests variation in _m between avian lineages, since the validity of statistical analyses is uncertain due to the difficulty in estimating confidence intervals of _m. Importantly, all presently available bird data indicate more mutations among males than females.

Does the excess of male mutations tie in quantitatively with the difference in the number of germline cell divisions between males and females? This question is difficult to address due to the lack of detailed cytological data, although it seems quite clear that spermatogenesis involves more cell generations than oogenesis in birds (JONES and LIN 1993 ). Moreover, if there is an intrinsic reduction in the Z chromosome mutation rate, as has been suggested for the mammalian X chromosome (MCVEAN and HURST 1997 ), comparisons of the rate of neutral evolution on Z and W chromosomes would tend to underestimate _m. On the other hand, _m could overestimate the difference in the number of cell divisions in male and female germlines if the per cell generation mutation rate differs between sexes. One such potential factor is the degree of germline methylation, which affects the mutability of CpG sites (LI 1997 ). For example, methylation has been invoked to explain the male-biased mutation rate at hemophilia A CpG sites, which are more strongly methylated in male than in female germline (OLDENBURG et al. 1993 ; SOMMER and KETTERLING 1996 ). According to the present data, however, a potential role of methylated CpG sites in explaining the male-biased mutation rate of avian CHD1 genes could be excluded.

In summary, the genomic location of the CHD1Z and CHD1W genes on the avian sex chromosomes is likely to have affected the molecular evolution of these two genes in distinct ways. While the two proteins are highly conserved and do not seem functionally differentiated, they differ with respect to frequency of synonymous and nonsynonymous nucleotide substitutions. Since the respective factors contributing to these differences (effective population size, recombination, and sex-specific mutation rates) should be valid for sex chromosomes in general, we anticipate the observed patterns of molecular evolution to be general characteristics of sex-linked genes.

	ACKNOWLEDGMENTS

We thank Tim Hipkiss and Jan Högberg for providing avian blood samples. We also thank Nick Smith and Bengt-Olle Bengtsson for valuable comments on the manuscript. Financial support was obtained from the Swedish Natural Sciences Research Council and Lars Hiertas Minne.

Manuscript received December 8, 1999; Accepted for publication April 14, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AGULNIK, A. I., C. E. BISHOP, J. L. LERNER, S. I. AGULNIK, and V. V. SOLOVYEV, 1997  Analysis of mutation rates in the SMCY/SMCX genes shows that mammalian evolution is male driven. Mamm. Genome 8:134-138[Medline].

ALVAREZ-VALIN, F., K. JABBARI, and G. BERNARDI, 1998  Synonymous and nonsynonymous substitutions in mammalian genes: intragenic correlations. J. Mol. Evol. 46:37-44[Medline].

BERNARDI, G., D. MOUCHIROUD, C. GAUTIER, and G. BERNARDI, 1988  Compositional patterns in vertebrate genomes: conservation and change in evolution. J. Mol. Evol. 28:7-18[Medline].

CARMICHAEL, A. N., A. K. FRIDOLFSSON, J. HALVERSON, and H. ELLEGREN, 2000  Male-biased mutation rates revealed from Z- and W-chromosome linked ATP synthase -subunit (ATP5A1) sequences in birds. J. Mol. Evol. 50:443-446[Medline].

CHANG, B. H. and W.-H. LI, 1995  Estimating the intensity of male-driven evolution in rodents by using X-linked and Y-linked Ube1 genes and pseudogenes. J. Mol. Evol. 40:70-77[Medline].

CHARLESWORTH, B., 1996  The evolution of chromosomal sex determination and dosage compensation. Curr. Biol. 6:149-162[Medline].

CHARLESWORTH, B., J. A. COYNE, and N. H. BARTON, 1987  The relative rates of evolution of sex chromosomes and autosomes. Am. Nat. 130:113-146.

CREW, F. A. E., 1954 Sex Determination. John Wiley & Sons, New York.

CROW, J. F., 1997  Molecular evolution—who is in the driver's seat? Nat. Genet. 17:129-130[Medline].

DRISCOLL, D. J. and B. R. MIGEON, 1990  Sex difference in methylation of single-copy genes in human meiotic germ cells: implications for X chromsome inactivation, parental imprinting and origin of CpG mutations. Somat. Cell Mol. Genet. 16:267-282[Medline].

DVORAK, J., J. L. HALVERSON, P. GULICK, K. A. RAUEN, and U. K. ABBOTT et al., 1992  cDNA cloning of a Z- and W-linked gene in Gallinaceous birds. J. Hered. 83:22-25[Abstract/Free Full Text].

ELLEGREN, H., 1996  First gene on the avian W chromosome provides a tag for universal sexing of non-ratite birds. Proc. R. Soc. Lond. Ser. B 263:1635-1641[Medline].

ELLEGREN, H., 2000  Evolution of the avian sex chromosomes and their role in sex determination. Trends Ecol. Evol. 15:188-192[Medline].

ELLEGREN, H. and A.-K. FRIDOLFSSON, 1997  Male-driven evolution of DNA sequences in birds. Nat. Genet. 17:182-184[Medline].

FELSENSTEIN, J., 1991 PHYLIP: phylogeny inference package. Version 3.4. University of Washington, Seattle.

FRIDOLFSSON, A.-K., H. CHENG, N. G. COPELAND, N. A. JENKINS, and H. C. LIU et al., 1998  Evolution of the avian sex chromosomes from an ancestral pair of autosomes. Proc. Natl. Acad. Sci. USA 95:8147-8152[Abstract/Free Full Text].

GRAVES, J. A. M., 1995  The evolution of mammalian sex chromosomes and the origin of sex determining genes. Philos. Trans. R. Soc. Lond. Ser. B 350:305-312[Medline].

GRIFFITHS, R. and R. M. KORN, 1997  A CHD1 gene is Z chromosome linked in the chicken Gallus domesticus.. Gene 197:225-229[Medline].

GRIFFITHS, R. and B. TIWARI, 1995  Sex of the last wild Spix's macaw. Nature 375:454[Medline].

GRIFFITHS, R., S. DAAN, and C. DIJKSTRA, 1996  Sex identification in birds using two CHD genes. Proc. R. Soc. Lond. Ser. B 263:1251-1256[Medline].

HALVERSON, J. J. and J. DVORAK, 1993  Genetic control of sex determination in birds and the potential for its manipulation. Poult. Sci. 72:890-896[Medline].

HUELSENBECK, J. P., 1997  Is the Felsenstein zone a fly trap? Syst. Biol. 46:69-74[Medline].

HURST, L. D. and H. ELLEGREN, 1998  Sex bias in the mutation rate. Trends Genet. 14:446-451[Medline].

JONES, R. C., and M. LIN, 1993 Spermatogenesis in birds, pp. 233–264 in Oxford Review of Reproductive Biology, edited by S. R. MILLIGAN. Oxford University Press, Oxford.

KUMAR, S., K. TAMURA and M. NEI, 1993 MEGA: Molecular Evolutionary Genetic Analysis, version 1.0. Pennsylvania State University, University Park, PA.

LAHN, B. T. and D. C. PAGE, 1997  Functional coherence of the human Y chromosome. Science 278:675-680[Abstract/Free Full Text].

LESSELS, K., 1997  More mutations in males. Nature 390:236-237[Medline].

LI, W.-H., 1993  Unbiased estimation of the rates of synonymous and nonsynonymous substitution. J. Mol. Evol. 36:96-99[Medline].

LI, W.-H., 1997 Molecular Evolution. Sinauer Associates, Sunderland, MA.

MCVEAN, G. T. and L. D. HURST, 1997  Evidence for a selectively favourable reduction in the mutation rate of the X chromosome. Nature 386:388-395[Medline].

MIYATA, T., H. HAYASHIDA, K. KUMA, K. MITSUYASU, and T. YASUNAGA, 1987  Male-driven molecular evolution: a model and nucleotide sequence analysis. Cold Spring Harbor Symp. Quant. Biol. 52:863-867[Abstract/Free Full Text].

NEI, M., 1970  Accumulation of nonfunctional genes on a sheltered chromosome. Am. Nat. 104:311-322.

OLDENBURG, J., R. SCWAAB, T. GRIMM, K. ZERRES, and K. HAKENBERG et al., 1993  Direct and indirect estimation of the sex ratio of mutation frequencies in hemophilia A. Am. J. Hum. Genet. 53:1229-1238[Medline].

OLIVIER, J. L. and A. MARIN, 1996  A relationship between GC content and coding sequence length. J. Mol. Evol. 43:216-223[Medline].

PAMILO, P. and N. O. BIANCHI, 1993  Evolution of the Zfx and Zfy genes: rates and interdependence between the genes. Mol. Biol. Evol. 10:271-281[Abstract].

RAYMOND, C. S., J. R. KETTLEWELL, B. HIRSCH, V. J. BARDWELL, and D. ZARKOWER, 1999  Expression of Dmrt1 in the genital ridge of mouse and chicken embryos suggests a role in vertebrate sexual development. Dev. Biol. 215:208-220[Medline].

RICE, W. R., 1996  Evolution of the Y sex chromosome in animals. BioScience 46:331-343.

SCHORDERET, D. F. and S. M. GARTLER, 1992  Analysis of CpG suppression in methylated and nonmethylated species. Proc. Natl. Acad. Sci. USA 89:957-961[Abstract/Free Full Text].

SHIMMIN, L. C., B. H. CHANG, and W.-H. LI, 1993  Male-driven evolution of DNA sequences. Nature 362:745-747[Medline].

SMITH, N. G. C. and L. D. HURST, 1998  Molecular evolution of an imprinted gene: repeatability of patterns of evolution within the mammalian insulin-like growth factor type II receptor. Genetics 150:823-833[Abstract/Free Full Text].

SMITH, N. G. C. and L. D. HURST, 1999  The causes of synonymous rate variation in rodent genome: can substitution rates be used to estimate the sex bias in mutation rate? Genetics 152:661-673[Abstract/Free Full Text].

SMITH, C. A., P. J. MCCLIVE, P. S. WESTERN, K. J. REED, and A. H. SINCLAIR, 1999  Conservation of a sex-determining gene. Nature 402:601-602[Medline].

SOMMER, S. S. and R. P. KETTERLING, 1996  The factor IX gene as a model for analysis of human germline mutations: an update. Hum. Mol. Genet. 5:1505-1514[Abstract].

STEFOS, K. and F. E. ARRIGHI, 1971  Heterochromatic nature of W chromosome in birds. Exp. Cell Res. 68:228-231[Medline].

STOKES, D. G. and R. P. PERRY, 1995  DNA-binding and chromatin localization properties of CHD1.. Mol. Cell. Biol. 5:2745-2753.

STOKES, D. G., K. D. TARTOF, and R. P. PERRY, 1996  CHD1 is concentrated in interbands and puffed regions of Drosophila polytene chromosomes. Proc. Natl. Acad. Sci. USA 93:7137-7142[Abstract/Free Full Text].

SWOFFORD, D. L., 1998 PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods), version 4.0b2a. Sinauer Associates, Sunderland, MA.

TUCKER, P. K. and B. L. LUNDRIGAN, 1993  Rapid evolution of the sex determining locus in Old World mice and rats. Nature 364:715-717[Medline].

WHITFIELD, S. L., R. LOVELL-BADGE, and P. N. GOODFELLOW, 1993  Rapid sequence evolution of the mammalian sex-determining gene SRY. Nature 364:713-715[Medline].

WOODAGE, T., M. A. BASRA, A. D. BAXEVANIS, P. HIETER, and F. S. COLLINS, 1997  Characterization of the CHD family of proteins. Proc. Natl. Acad. Sci. USA 94:11472-11477[Abstract/Free Full Text].

YANG, Z., 1999 Phylogenetic Analysis by Maximum Likelihood (PAML), version 2.0. University College, London.

YANG, Z. and R. NIELSEN, 1998  Synonymous and nonsynonymous rate variation in nuclear genes of mammals. J. Mol. Evol. 46:409-418[Medline].

This article has been cited by other articles:

				D. E. Janes, T. Ezaz, J. A. Marshall Graves, and S. V. Edwards Recombination and Nucleotide Diversity in the Sex Chromosomal Pseudoautosomal Region of the Emu, Dromaius novaehollandiae J. Hered., March 1, 2009; 100(2): 125 - 136. [Abstract] [Full Text] [PDF]

				H. Ellegren Characteristics, causes and evolutionary consequences of male-biased mutation Proc R Soc B, January 7, 2007; 274(1606): 1 - 10. [Abstract] [Full Text] [PDF]

				D. Bachtrog Sex chromosome evolution: Molecular aspects of Y-chromosome degeneration in Drosophila Genome Res., October 1, 2005; 15(10): 1393 - 1401. [Abstract] [Full Text] [PDF]

				D. T. Gerrard and D. A. Filatov Positive and Negative Selection on Mammalian Y Chromosomes Mol. Biol. Evol., June 1, 2005; 22(6): 1423 - 1432. [Abstract] [Full Text] [PDF]

				E. Axelsson, N. G.C. Smith, H. Sundstrom, S. Berlin, and H. Ellegren Male-Biased Mutation Rate and Divergence in Autosomal, Z-Linked and W-Linked Introns of Chicken and Turkey Mol. Biol. Evol., August 1, 2004; 21(8): 1538 - 1547. [Abstract] [Full Text] [PDF]

				L.-J. L. Handley, H. Ceplitis, and H. Ellegren Evolutionary Strata on the Chicken Z Chromosome: Implications for Sex Chromosome Evolution Genetics, May 1, 2004; 167(1): 367 - 376. [Abstract] [Full Text] [PDF]

				H. Ceplitis and H. Ellegren Adaptive Molecular Evolution of HINTW, a Female-Specific Gene in Birds Mol. Biol. Evol., February 1, 2004; 21(2): 249 - 254. [Abstract] [Full Text] [PDF]

				R. J. Agate, M. Choe, and A. P. Arnold Sex Differences in Structure and Expression of the Sex Chromosome Genes CHD1Z and CHD1W in Zebra Finches Mol. Biol. Evol., February 1, 2004; 21(2): 384 - 396. [Abstract] [Full Text] [PDF]

				D. A. Filatov and D. Charlesworth Substitution Rates in the X- and Y-Linked Genes of the Plants, Silene latifolia and S. dioica Mol. Biol. Evol., June 1, 2002; 19(6): 898 - 907. [Abstract] [Full Text] [PDF]

				H. Montell, A.-K. Fridolfsson, and H. Ellegren Contrasting Levels of Nucleotide Diversity on the Avian Z and W Sex Chromosomes Mol. Biol. Evol., November 1, 2001; 18(11): 2010 - 2016. [Abstract] [Full Text] [PDF]

				H. Ellegren and A. Carmichael Multiple and Independent Cessation of Recombination Between Avian Sex Chromosomes Genetics, May 1, 2001; 158(1): 325 - 331. [Abstract] [Full Text]