Positive Selection at Reproductive ADAM Genes with Potential Intercellular Binding Activity
http://www.100md.com
分子生物学进展 2004年第5期
Department of Biology, University of Winnipeg, Winnipeg, Manitoba, Canada
E-mail: a.civetta@uwinnipeg.ca.
Abstract
Many genes with a role in reproduction, including those implicated in fertilization and spermatogenesis, have been shown to evolve at a faster rate relative to genes associated with other functions and tissues. These survey studies usually group a wide variety of genes with different characteristics and evolutionary histories as reproductive genes based on their site of expression or function. We have examined the molecular evolution of the ADAM (a disintegrin and metalloprotease) gene family, a structurally and functionally diverse group of genes expressed in reproductive and somatic tissue to test whether a variety of protein characteristics such as phylogenetic clusters, tissue of expression, and proteolytic and adhesive function can group fast evolving ADAM genes. We found that all genes were evolving under purifying selection (dN/dS < 1), although reproductive ADAMs, including those implicated in fertilization and spermatogenesis, evolved at the fastest rate. Genes with a role in binding to cell receptors in endogenous tissue appear to be evolving under purifying selection, regardless of the tissue of expression. In contrast, positive selection of codon sites in the disintegrin/cysteine-rich adhesion domains was detected exclusively in ADAMs 2 and 32, two genes expressed in the testis with a potential role in sperm-egg adhesion. Positive selection was detected in the transmembrane/cytosolic tail region of ADAM genes expressed in a variety of tissues.
Key Words: ADAM gene family ? reproductive proteins ? rapid evolution ? positive selection ? mouse-human comparisons
Introduction
Genes with a role in reproduction have been suggested to show patterns of rapid evolution in a wide variety of organisms (Civetta and Singh 1998, 1999; Singh and Kulathinal 2000; Wyckoff, Wang, and Wu 2000; Swanson et al. 2001). The most extreme examples of rapid evolution driven by positive selection come from studies on sperm proteins in marine invertebrates, where selection is thought to act on sites involved in gamete recognition and sperm-egg fusion (Swanson and Vacquier 1995; Metz and Palumbi 1996).
In mammals, sperm proteins in general (Wyckoff, Wang, and Wu 2000; Torgerson, Kulathinal, and Singh 2002) and those with a role in binding to the zona pellucida (Swanson, Nielsen, and Yang 2003) show signs of interspecific sequence divergence driven by positive selection. In these studies, sequence analysis was done by comparing rates of nonsynonymous and synonymous substitutions among species from a pool of mammalian genes that have a common site of expression (testes and/or sperm) to others. In such a diverse pool of reproductive genes, other characteristics may contribute to the patterns of interspecific divergence experienced by these proteins. Focusing the analysis on a specific gene family can minimize problems that arise from comparing genes with a wide variety of functions and evolutionary history.
The ADAM (a disintegrin and metalloprotease) family is a multigene family comprised of members with a wide variety of functions such as myogenesis, neurogenesis, cell-cycle control and spermatogenesis. Different domains can be identified within ADAM proteins and these domains usually have a defined function in at least one member of the gene family; however, ADAM proteins are characterized as a group by the presence of at least a metalloprotease and a disintegrin domain (Primakoff and Myles 2000). Not all members of the ADAM protein family retain a metalloprotease domain in the mature protein. The presence of a zinc-binding catalytic consensus sequence within the metalloprotease domain defines proteolytic ADAM proteins capable of cleaving substrates, while some inactive proteases lack the catalytic consensus sequence or have a mutated version of it (Schl?ndorff and Blobel 1999; Primakoff and Myles 2000). The disintegrin domain of ADAM proteins provides adhesive properties for those ADAMs having an integrin-binding motif at the tip of a flexible protein loop (Primakoff and Myles 2000).
Little is known of the evolutionary relationship among ADAM genes, although there is some evidence to suggest that ADAM genes expressed in sperm and ADAM gene domains that function in sperm-egg interactions may be more diverged between species than other ADAM genes (Civetta 2003). There is heterogeneity in evolutionary rates within fertilin (ADAM 1) and fertilin ? (ADAM 2) with the prodomain and the metalloprotease domains more conserved than other regions, which was suggested to be a result of their roles in spermatogenesis or sperm capacitance, and positive selection detected in the disintegrin and cysteine-rich adhesion domains (Civetta 2003).
ADAMs are a group of widely expressed genes with potential roles in cell adhesion. Members of the ADAM family may also function as proteases based on whether a metalloprotease domain is present and active in the protein (Primakoff and Myles 2000). To date, 32 ADAM genes have been identified in the mouse, of which 19 have confirmed metalloprotease activity as mature proteins and are expressed either in reproductive tissue (testis or epididimys; n = 10) or somatic tissue (n = 9) (Primakoff and Myles 2000; Brachvogel et al. 2002; Gunn et al. 2002; Choi et al. 2003). At least 12 ADAM genes have been suggested to function in cell adhesion based on the presence of predictive amino acid motifs and/or in vitro binding assays to potential integrin ligands (Sagane et al. 1999; Evans 2002; Frayne, Hurd, and Hall 2002; Choi et al. 2003; table 1). Although few receptors have been confirmed, the disintegrin domain is implicated as the principal region mediating binding between ADAM ligands and integrins (Primakoff and Myles 2000).
Table 1 Average dN/dS Estimates and Proportions of Nonsynonymous (dN) and Synonymous (dS) Substitutions Among ADAM Genes in Mouse-Human Comparisons.
In this study, we collected 17 different mouse and human ADAM gene sequences and tested whether protein characteristics, other than the site of protein expression, could group fast evolving ADAM genes. We used the phylogenetic analysis by maximum likelihood approach (PAML; Yang 1997) on tree topologies from different ADAM genes to test for variable rates of evolution along tree branches and to determine the protein domain distribution (i.e., amino end, metalloprotease, disintegrin/cysteine-rich, carboxy tail) of positively selected codons.
Materials and Methods
Sequence Data Analysis
The following 17 complete ADAM gene sequences were obtained from GenBank from Mus musculus and Homo sapiens respectively: ADAM 2 (u38806; u38805), ADAM 7 (nm_007402; af215824), ADAM 8 (nm_007403; nm_001109), ADAM 9 (nm_007404; nm_003816), ADAM 10 (nm_007399; af009615), ADAM 11 (nm_009613; ab009675), ADAM 12 (nm_007400; af023477), ADAM 15 (ab022089; nm_003815), ADAM 17 (nm_009615; nm_003183), ADAM 18 (nm_010084; nm_014237), ADAM 19 (nm_009616; af311317), ADAM 21 (nm_020330; af158644), ADAM 22 (ab009674; nm_021722), ADAM 23 (nm_011780; nm_003812), ADAM 28 (af153350; nm_021778), ADAM 32 (af513715, nm_145004), ADAM 33 (nm_033615; nm_025220).
Amino acid sequence alignments were obtained using the ClustalW program within the DAMBE software package and used to generate nucleotide alignments (Xia 2000). An ADAM gene phylogeny for the 17 mouse and human sequences was constructed using distance (Neighbor-Joining), maximum parsimony (MP), and maximum likelihood (ML) methods (Mega 2.1; PAUP*4.0) (Swafford 1992; Kumar et al. 2001). The most parsimonious consensus tree was generated under the Tree Bisection-Reconnection (TBR) branch-swapping algorithm. The MP tree was evaluated using maximum likelihood methods under the heuristic model with 25 random-addition replicates and a transition/ transversion ratio = 2:1. The reliability of the inferred tree was assessed by bootstrapping with 1,000 replicates, and bootstrap values with scores greater than 80% were retained. For the maximum likelihood tree reconstruction, we used nucleotide substitution models that consider only transition/transversion bias (F84 model) or both transition/ transversion bias and transition rate heterogeneity (Tn93 model). The log likelihood ratio test (LR = 2 [InLTn93–InLF84]) (Nei and Kumar 2000, p. 70) was used to assess which model best fit the underlying data.
Proportion of nonsynonymous substitutions per nonsynonymous site (dN) and synonymous substitutions per synonymous site (dS) between mouse and human were calculated for each ADAM gene using the modified Nei-Gojobori Jukes Cantor method (Nei and Gojobori 1986; Nei and Kumar 2000, pp. 57–60), which takes into account deviations from a transition/transversion (R) = 1. We compared dN/dS ratios by grouping ADAM genes based on phylogenetic clustering, protease activity, potential ligand binding, and tissue of expression. Statistical comparisons of average dN/dS ratios among groups were tested by approximate randomization analysis with 25,000 permutations (Manly 1991).
Variation in Selective Pressures Across Lineages
For eight of the seventeen ADAM genes used to reconstruct the gene family phylogeny, sequences are available from GenBank for at least rat (Rattus novergicus) and/or macaque (Macaca fascicularis) (accession numbers available upon request). Tree topologies were generated using groups of three species, for a total of four possible combinations: human/rat/macaque, human/rat/mouse, macaque/rat/mouse, and human/mouse/macaque. We used tree topologies obtained from the clustal alignment to test for variation in selective pressures measured as dN/dS per codon () across lineages using the CODEML software package in PAML (Yang 1997). Likelihood estimates for each tree topology under the assumption of a unique ratio for all tree branches (Model 0) was compared to estimates from a free-ratio model (Model 1, variable among branches). This was achieved by changing the parameters for model to options 0 or 1 in the CODEML control file while keeping the NSsite option that allows for variation in across codons fixed at one ratio value (M0). A likelihood ratio test was calculated by multiplying twice the difference between log likelihood scores () obtained under each model and comparing the result against a 2 distribution with degrees of freedom equal to the difference in the number of parameters estimated in each model (Nielsen and Yang 1998; Yang et al. 2000). To maximize statistical power the analysis was repeated with ADAM genes for which four or more different mammalian species sequences were available (accession numbers available upon request).
Variation in Selective Pressure Across Codon Sites
We used the CODEML software package in PAML to detect specific codon sites under positive selection among sites experiencing variable selective pressures by comparing likelihood estimates of a tree topology under a model that assumes a distribution of values restricted between 0 and 1 (M7, no positive selection) with an alternative model that allows for values greater than 1 (M8 model). For this purpose, the parameter for model in the CODEML control file was set to 0 while changing the NSsite option to 7 (M7) or 8 (M8). Comparing estimates between M7 and M8 depicts the existence of codons under positive selection (Yang et al. 2000). If the log-likelihood test suggests the presence of sites under positive selection, then these sites can be identified by using a Bayesian method to estimate posterior probabilities (P) that particular sites have experienced a value greater than 1. Positively selected sites identified using this approach were assigned to different protein domains previously identified by using the Motif scan tool available in the PROSITE database of protein families and domains (http://hits.isb-sib.ch/cgi-bin/PFSCAN).
Results
Human-Mouse dN/dS Comparisons
The ancestral relationships among ADAM genes generated by distance, parsimony, and likelihood methods correspond to previous partial phylogenies of the gene family (Poindexter et al. 1999).
Phylogenetic analysis identified three ancestral gene clusters based on bootstrap estimates: cluster 1 is comprised exclusively of nonproteolytic genes expressed in the brain, and clusters 2 and 3 are a mix of proteolytic and nonproteolytic, somatic, and sex-related genes (fig. 1; see also online Supplementary Material). dN/dS ratio estimates between mouse and human varied across genes (one-sample t-test: t16 = 7.23, P < 0.001; table 1). When dN/dS estimates are grouped on the basis of their phylogenetic clustering, proteolytic activity, adhesion, or tissue of expression, results show a significantly higher average dN/dS for ADAM genes expressed in reproductive tissue versus nonsexual tissues (D6,11 = 0.193; P = 0.001; table 2). Interestingly, phylogenetic cluster 1 comprised exclusively of nonproteolytic genes expressed in the brain shows the lowest dN/dS ratio estimate (fig. 1 and table 2).
FIG. 1. Unrooted maximum parsimonious tree of ancestral phylogeny generated by PAUP*4.0 using Homo sapiens (Hs) and Mus musculus (Mm) ADAM gene sequences. Bootstrap values were 100% for all mouse-human nodes and are not shown. Other bootstrap values higher than 80% are shown in bold type
Table 2 Approximate Randomization of dN/dS on ADAM Gene Sequences Differentiated on the Basis of Phylogenetic Clusters, Proteolytic Structure, Tissue of Expression, and Ligand Function.
Although the average dN/dS rate is significantly higher for ADAM genes expressed in reproductive tissue compared to others, all ADAMs show dN/dS ratios less than 1 that are indicative of purifying selection. It is possible that some ADAM genes have experienced episodes of positive selection and that such signals remained undetected by using standard dN/dS estimates for the entire gene sequences.
Variation in Selective Pressures Across Lineages
Only four ADAM genes (ADAM 2, 17, 18, and 28) show evidence of significant variation in selective pressures across branches in the tree (table 3). Higher values seem to be found within primate branches (fig. 2) and this is supported by a significantly higher for either of the primate branches compared to the rodent tree branch in human, macaque, and mouse (or rat) three-species comparisons (paired t: t6 = –2.83, P = 0.030). This observation could be a consequence of true differences in selective pressures or an underestimation of the ratio in the rodent tree branch associated with larger rodent populations (Yang and Nielsen 1998). It is also possible that failure to reject the one-ratio model for phylogenies in which rodent lineages outnumber primate lineages is due to lack of power (Anisimova, Bielawski, and Yang 2001).
Table 3 Variation in Selective Pressures Along Different ADAM Gene Lineages.
FIG. 2. Phylogeny and branch estimates for (a) ADAM 7, (b) ADAM 2, (c) ADAM18 and (d) ADAM10 using Bos taurus (Bt), Cavia porcellus (Cp), Homo sapiens (Hs), Macaca fascicularis (Mf), Mus musculus (Mm), Oryctolagus cuniculus (Oc), Rattus norvegicus (Rn), and Sus scrofa (Ss)
Variation in Selective Pressures Across Codon Sites
The phylogenetic analysis by maximum likelihood (PAML) was used to estimate the likelihood of a phylogeny under models that make alternative assumptions about the dN/dS rate of change among codon sites () (Nielsen and Yang 1998; Yang et al. 2000). The tree topologies for groups of more than three species used for the analysis are shown in figure 2. We tested for the presence of positively selected sites by comparing the log-likelihood of a tree under a model that assumes a distribution of values constrained between 0 and 1 (M7) with the estimate obtained under an alternative model that adds a class of sites with ratios >1.0 (M8). The likelihood ratio tests obtained from comparing M7 and M8 shows that M8 fits the data better than M7 for ADAM genes expressed in both testis (ADAMs 2, 32) and somatic tissue (ADAMs 17, 28) indicating the occurrence of positive selection at codon sites within genes (table 4). However, a higher proportion of positively selected sites were found for sperm ADAMs 2 and 32 than ADAM 17 and 28 (table 4). There is also a difference in the distribution of positively selected sites across these genes. While all ADAM genes show a high concentration of positively selected sites in the carboxy tail region, only sperm ADAM genes show the highest proportion of positively selected sites in the disintegrin/cysteine-rich domain with potential adhesion function (fig. 3 and table 5).
Table 4 Variation in Selective Pressure Across Codon Sites for ADAM Gene Lineages in Groups of Three or More Species Comparisons.
FIG. 3. Posterior probability of positively selected codon sites ( > dN/dS) and their distribution among the amino, metalloprotease (grey rectangle), disintegrin/cysteine-rich (black rectangle), and carboxy tail domains of ADAM 2, ADAM 17, ADAM 28, and ADAM 32. The horizontal line indicates a posterior probability of 0.95
Table 5 Proportion of Positively Selected Codon Sites per Domain (P > 0.5) and Amino Acids in Mus musculus with a Posterior Probability >0.95 of Being Under Positive Selection.
Discussion
Approximate randomization analysis supports previous results showing that reproductive genes, including those implicated in fertilization and spermatogenesis, evolve at a faster rate relative to genes associated with other tissues, in agreement with numerous other studies (Civetta and Singh 1988, 1999; Singh and Kulathinal 2000; Wyckoff, Wang, and Wu 2000; Swanson et al. 2001). Maximum likelihood analysis revealed that ADAM genes expressed in the testis with the potential for sperm-egg adhesion are subjected to positive selection.
Genes expressed in both the testis and somatic tissue showed positive selection of codon sites in the carboxy tail, situated on the transmembrane and cytoplasmic segment of the ADAM gene. The role of the carboxy tail in ADAM proteins is poorly understood, although it may be important for localization of the sperm head prior to sperm-egg adhesion (Hunnicut, Koppel, and Myles 1997; Myles and Primakoff 1997; Cowan et al. 2001) or in signal transduction (Evans, Schultz, and Kopf 1998; Kang, Cao, and Zolkiewska 2000). There is also a potential adhesion role identified for the carboxy tail region of ADAMs 17 and 28 that may involve binding of proline-rich regions on the carboxy tail to SH3-domain-containing proteins (Howard, Maciewicz, and Blobel 2003) or of ADAM 17 to mitotic arrest deficient 2 (MAD2), a component of the mitotic machinery (Nelson, Schl?ndorff, and Blobel 1999), although these functions remain speculative.
Among ADAMs showing signs of positive selection, we found the highest proportion of positively selected codon sites in the disintegrin/cysteine-rich region of genes expressed in the testis (ADAMs 2 and 32) (fig. 3). Similar results were found for ADAM 1, a heterodimer of ADAM 2 that plays a direct role in sperm-egg adhesion (Civetta 2003). ADAM 2 is supposed to be implicated in sperm-egg adhesion as a ligand of integrins located on the surface of the egg (Evans, Kopf, and Schultz 1997). Such adhesive function is mediated by sites located in the disintegrin domain of ADAM 2 (Zhu et al. 2000). Mice with an ADAM 2 knockout are sterile; in in vitro assays the null sperm is defective in binding to the egg zona pellucida and is rarely found in the oviduct of females despite normal motility. ADAM 2 null males also show a reduced ability to fuse the egg membrane, suggesting that ADAM 2 is required but not essential in sperm-egg membrane binding (Cho et al. 1998). ADAM 32 shares a close sequence homology to ADAM 2, including a sequence similar to the conserved binding motif of the disintegrin loop, suggesting a potential role in gamete adhesion (Choi et al. 2003).
ADAM genes with the potential for mediating cell-cell adhesion within endogenous tissues and/or with a role in embryonic development were found to be under purifying selection. Included in this category are ADAMs 11, 22, and 23, potential ligands of integrins in neural brain tissue (Sagane et al. 1999); ADAMs 9, 12, and 15, genes with adhesion roles in fibroblasts, myoblasts and haemopoietic cells (Nath et al.1999, 2000; Eto et al. 2000); and ADAMs 7 and 18, genes expressed in the epididymis with roles in sperm development (Cornwall and Hsia 1997; Frayne, Hurd, and Hall 2002).
Our study is the first to examine the evolution of ADAM genes in the context of a disintegrin multigene family. The ADAM proteins are thought to have evolved as a multidomain structure from a proteolytic ancestor and for which pressure on the disintegrin adhesion domain is the suggested mechanism promoting functional diversification (Moura-da-Silva, Theakston, and Crampton 1996). This process is similar to selection for antigen recognition sites of major histocompatability loci (MHC) (Hughes and Nei 1988; Hughes, Ota, and Nei 1990) and heterogeneity at binding sites within the antifreeze protein (AFP) multigene family (Swanson and Aquadro 2002).
Evidence for rapid evolution of reproductive and immune system genes has led to speculation that genes binding to "foreign" substances may be generally under positive selection (e.g., Wyckoff, Wang, and Wu 2000). Our results show that positive Darwinian selection is restricted to the adhesion domains of ADAM genes involved in sperm-egg plasma membrane adhesion, as opposed to binding domains of all ADAM genes. This is best illustrated by ADAM 28, for which we found no evidence of positive selection at adhesion sites, despite the potential binding role inferred by recognition of the disintegrin domain by integrin 4?1 on human lymphocytes (Bridges et al. 2002).
Our study highlights the faster evolution of reproductive genes as a group and the adaptive diversification of sites within binding domains of sperm-egg adhesion ADAMs. Further functional studies are needed to determine the mechanism and significance of the actual changes driven by positive selection.
Acknowledgements
We would like to thank Ed Byard and Brian Golding for comments on the manuscript. Thanks to Scott Forbes for kindly providing his approximate randomization program. This work was supported by an Establishment Manitoba Health Research Council grant and a research grant from the Natural Sciences and Engineering Research Council of Canada to A.C.
Literature Cited
Anisimova, M., J. P. Bielawski, and Z. Yang. 2001. Accuracy and power of the likelihood ratio test in detecting adaptive molecular evolution. Mol. Biol. Evol. 18:1585-1592.
Brachvogel, B., D. Reichenberg, S. Beyer, B. Jehn, K. von der Mark, and W. Bielke. 2002. Molecular cloning and expression analysis of a novel member of the disintegrin and metalloprotease-domain (ADAM) family. Gene 288:203-210.
Bridges, L. C., P. H. Tani, K. R. Hanson, C. M. Roberts, M. B. Judkins, and R. D. Bowditch. 2002. The lymphocyte metalloprotease MDC-L (ADAM 28) is a ligand for the integrin 4?1. J. Biol. Chem. 277:3784-3792.
Cho, C., D. O'Dell Bunch, J-E. Faure, E. H. Goulding, E. M. Eddy, P. Primakoff, and D. G. Myles. 1998. Fertilization defects in sperm from mice lacking fertilin ?. Science 281:1857-1859.
Choi, I., J-M. Woo, S. Hong, Y. K. Jung, D. H. Kim, and C. Cho. 2003. Identification and characterization of ADAM32 with testis-predominant gene expression. Gene 304:151-62.
Civetta, A. 2003. Positive selection within sperm-egg adhesion domains of fertilin: an ADAM gene with a potential role in fertilization. Mol. Biol. Evol. 20:21-29.
Civetta, A., and R. S. Singh. 1998. Sex-related genes, directional sexual selection, and speciation. Mol. Biol. Evol. 15:901-909.
Civetta, A., and R. S. Singh. 1999. Broad-sense sexual selection, sex gene pool evolution, and speciation. Genome 42:1033-1041.
Cornwall, G. A., and N. Hsia. 1997. ADAM7, A member of the ADAM (A Disintegrin And Metalloprotease) gene family is specifically expressed in the mouse anterior pituitary and epididymis. Endocrinology 138:4262-4272.
Cowan, A. E., D. E. Koppel, L. A. Vargas, and G. R. Hunnicut. 2001. Guinea pig fertilin exhibits restricted lateral mobility in epididymal sperm and becomes freely diffusing during capacitation. Dev. Biol. 236:502-509.
Eto, K., W. Puzon-McLaughlin, D. Sheppard, A. Sehara-Fujisawa, X-P. Zhang, and Y. Takada. 2000. RGD-independent binding of integrin 9?1 to the ADAM-12 and -15 disintegrin domains mediates cell-cell interaction. J. Biol. Chem. 275:34922-34930.
Evans, J. P., G. S. Kopf, and R. M. Schultz. 1997. Characterization of the binding of recombinant mouse sperm fertilin ? subunit to mouse eggs: Evidence for adhesive activity via an egg ?1 integrin-mediated interaction. Dev. Biol. 187:79-93.
Evans, J. P., R. M. Schultz, and G. S. Kopf. 1998. Roles of the disintegrin domains of mouse fertilins and ? in fertilization. Biol. Reprod. 59:145-152.
Evans, J. P. 2002. The molecular basis of sperm-oocyte membrane interactions during mammalian fertilization. Hum. Reprod. Update 8:297-311.
Frayne J., E. A. Hurd, and L. Hall. 2002. Human tMDC III: a sperm protein with a potential role in oocyte recognition. Mol. Hum. Reprod. 8:817-22.
Gunn, T. M., A. Azarani, P. H. Kim, R. W. Hyman, R. W. Davis, and G. S. Barsh. 2002. Identification and preliminary characterization of mouse ADAM33. BMC Genet. 3:2-9.
Howard, L., R. A. Maciewicz, and C. P. Blobel. 2003. Cloning and characterization of ADAM28: evidence for autocatalytic pro-domain removal and for cell surface localization of mature ADAM28. Biochem. J. 348:21-27.
Hughes, A. L., and M. Nei. 1988. Pattern of nucleotide substitution at MHC class I loci reveals overdominant selection. Nature 335:167-170.
Hughes, A. L., T. Ota, and M. Nei. 1990. Positive Darwinian selection promotes charge profile diversity in the antigen-binding cleft of class I major-histocompatibility-complex molecules. Mol. Biol. Evol. 7:515-524.
Hunnicut, G. R., D. E. Koppel, and D. G. Myles. 1997. Analysis of the process of localization of fertilin to the sperm posterior head plasma membrane domain during sperm maturation in the epididymis. Dev. Biol. 191:146-159.
Kang, Q., Y. Cao, and A. Zolkiewska. 2000. Metalloprotease-disintegrin ADAM 12 binds to the SH3 domain of Src and activates Src tyrosine kinase in C2C12 cells. Biochem. J. 352:883-892.
Kumar, S., K. Tamura, I. B. Jacobsen, and M. Nei. 2001. MEGA2: molecular evolutionary genetics analysis software. Bioinformatics 17:1244-1245.
Manly, B. J. F. 1991. Randomization and Monte Carlo methods in biology. Chapman and Hall, London.
Metz, E. C., and S. R. Palumbi. 1996. Positive selection and sequence rearrangements generate extensive polymorphism in the gamete recognition protein bindin. Mol. Biol. Evol. 13:397-406.
Moura-da-Silva, A. M., R. D. G. Theakston, and J. M. Crampton. 1996. Evolution of disintegrin cysteine-rich and mammalian matrix-degrading metalloproteinases: gene duplication and divergence of a common ancestor rather than convergent evolution. J. Mol. Evol. 43:263-269.
Myles, D. G., and P. Primakoff. 1997. Why did the sperm cross the cumulus? To get to the oocyte. Functions of the sperm surface proteins PH-20 and fertilin in arriving at, and fusing with, the egg. Biol. Reprod. 56:320-327.
Nath, D., P. M. Slocombe, P. E. Stephens, A. Warn, G. R. Hutchinson, K. M. Yamada, A. J. Docherty, and G. Murphy. 1999. Interaction of metargidin (ADAM-15) with v?3 and 5?1 integrins on different haemopoietic cells. J. Cell Sci. 112:579-587.
Nath, D., P. M. Slocombe, A. Webster, P. E. Stephens, A. J. P. Docherty, and G. Murphy. 2000. Meltrin (ADAM-9) mediates cellular adhesion through 6?1 integrin, leading to a marked induction of fibroblast cell motility. J. Cell Sci. 113:2319-2328.
Nei, M., and T. Gojobori. 1986. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol. Biol. Evol. 3:418-426.
Nei, M., and S. Kumar. 2000. Molecular evolution and phylogenetics. Oxford University Press, New York.
Nielsen, R., and Z. Yang. 1998. Likelihood models for detecting positively selected amino acid sites and applications to the HIV-1 envelope gene. Genetics 148:929-936.
Nelson, K. I., J. Schl?ndorff, and C. P. Blobel. 1999. Evidence for an interaction of the metalloprotease–disintegrin tumour necrosis factor convertase (TACE) with mitotic arrest deficient 2 (MAD2), and of the metalloprotease–disintegrin MDC9 with a novel MAD2-related protein, MAD2?. Biochem. J. 343:673-680.
Poindexter, K., N. Nelson, R. F. DuBose, R. A. Black, and D. P. Ceretti. 1999. The identification of seven metalloproteinase-disintegrin (ADAM) genes from genomic libraries. Gene 237:61-70.
Primakoff, P., and D. Myles. 2000. The ADAM gene family: surface proteins with adhesion and protease activity. Trends Genet. 16:83-87.
Sagane, K., K. Yamazaki, Y. Mizui, and I. Tanaka. 1999. Cloning and chromosomal mapping of mouse ADAM11, ADAM22, and ADAM23. Gene 236:79-86.
Schl?ndorff, J., and C. P. Blobel. 1999. Metalloprotease-disintegrins: modular proteins capable of promoting cell-cell interactions and triggering signals by protein-ectodomain shedding. J. Cell Sci. 112:3603-3617.
Singh, R. S., and R. J. Kulathinal. 2000. Sex gene pool evolution and speciation: A new paradigm. Genes Genet. Sys. 75:119-130.
Swafford, D. L. 1992. PAUP*: phylogenetic analysis using parsimony (*and other methods). Version 4. Sinauer Associates, Sunderland, Mass.
Swanson, W. J., and V. D. Vacquier. 1995. Extraordinary divergence and positive Darwinian selection in a fusagenic protein coating the acrosomal process of abalone spermatozoa. Proc. Natl. Acad. Sci. USA 92:4957-61.
Swanson, W. J., Z. Yang, M. F. Wolfner, and C. F. Aquadro. 2001. Positive Darwinian selection drives the evolution of several female reproductive proteins in mammals. Proc. Natl. Acad. Sci. USA 98:2509-2514.
Swanson, W. J., and C. F. Aquadro. 2002. Positive Darwinian selection promotes heterogeneity among members of the antifreeze protein multigene family. J. Mol. Evol. 54:403-410.
Swanson, J. W., R. Nielsen, and Q. Yang. 2003. Pervasive adaptive evolution in mammalian fertilization proteins. Mol. Biol. Evol. 20:18-20.
Torgerson, D. G., R. J. Kulathinal, and R. S. Singh. 2002. Mammalian sperm proteins are rapidly evolving: evidence of positive selection in functionally diverse genes. Mol. Biol. Evol. 19:1973-1980.
Wyckoff, G. J., W. Wang, and C-I Wu. 2000. Rapid evolution of male reproductive genes in the descent of man. Nature 403:304-308.
Xia, X. 2000. Data analysis in molecular biology and evolution. Kluwer Academic Publishers, Boston.
Yang, Z. 1997. PAML: a program package for phylogenetic analysis by maximum likelihood. Comput. Appl. Biosci. 13:555-556.
Yang, Z., and R. Nielsen. 1998. Synonymous and nonsynonymous rate variation in nuclear genes of mammals. J. Mol. Evol. 46:409-418.
Yang, Z., R. Nielsen, N. Goldman, and A-M. K. Pedersen. 2000. Codon-substitution models for heterogeneous selection pressure at amino acid sites. Genetics 155:431-449.
Zhu, X., N. P. Bansal, and J. P. Evans. 2000. Identification of key amino acids of the mouse fertilin ? (ADAM2) disintegrin loop for cell-cell adhesion during fertilization. J. Biol. Chem. 275:7677-7683.(Barb Glassey1 and Alberto)
E-mail: a.civetta@uwinnipeg.ca.
Abstract
Many genes with a role in reproduction, including those implicated in fertilization and spermatogenesis, have been shown to evolve at a faster rate relative to genes associated with other functions and tissues. These survey studies usually group a wide variety of genes with different characteristics and evolutionary histories as reproductive genes based on their site of expression or function. We have examined the molecular evolution of the ADAM (a disintegrin and metalloprotease) gene family, a structurally and functionally diverse group of genes expressed in reproductive and somatic tissue to test whether a variety of protein characteristics such as phylogenetic clusters, tissue of expression, and proteolytic and adhesive function can group fast evolving ADAM genes. We found that all genes were evolving under purifying selection (dN/dS < 1), although reproductive ADAMs, including those implicated in fertilization and spermatogenesis, evolved at the fastest rate. Genes with a role in binding to cell receptors in endogenous tissue appear to be evolving under purifying selection, regardless of the tissue of expression. In contrast, positive selection of codon sites in the disintegrin/cysteine-rich adhesion domains was detected exclusively in ADAMs 2 and 32, two genes expressed in the testis with a potential role in sperm-egg adhesion. Positive selection was detected in the transmembrane/cytosolic tail region of ADAM genes expressed in a variety of tissues.
Key Words: ADAM gene family ? reproductive proteins ? rapid evolution ? positive selection ? mouse-human comparisons
Introduction
Genes with a role in reproduction have been suggested to show patterns of rapid evolution in a wide variety of organisms (Civetta and Singh 1998, 1999; Singh and Kulathinal 2000; Wyckoff, Wang, and Wu 2000; Swanson et al. 2001). The most extreme examples of rapid evolution driven by positive selection come from studies on sperm proteins in marine invertebrates, where selection is thought to act on sites involved in gamete recognition and sperm-egg fusion (Swanson and Vacquier 1995; Metz and Palumbi 1996).
In mammals, sperm proteins in general (Wyckoff, Wang, and Wu 2000; Torgerson, Kulathinal, and Singh 2002) and those with a role in binding to the zona pellucida (Swanson, Nielsen, and Yang 2003) show signs of interspecific sequence divergence driven by positive selection. In these studies, sequence analysis was done by comparing rates of nonsynonymous and synonymous substitutions among species from a pool of mammalian genes that have a common site of expression (testes and/or sperm) to others. In such a diverse pool of reproductive genes, other characteristics may contribute to the patterns of interspecific divergence experienced by these proteins. Focusing the analysis on a specific gene family can minimize problems that arise from comparing genes with a wide variety of functions and evolutionary history.
The ADAM (a disintegrin and metalloprotease) family is a multigene family comprised of members with a wide variety of functions such as myogenesis, neurogenesis, cell-cycle control and spermatogenesis. Different domains can be identified within ADAM proteins and these domains usually have a defined function in at least one member of the gene family; however, ADAM proteins are characterized as a group by the presence of at least a metalloprotease and a disintegrin domain (Primakoff and Myles 2000). Not all members of the ADAM protein family retain a metalloprotease domain in the mature protein. The presence of a zinc-binding catalytic consensus sequence within the metalloprotease domain defines proteolytic ADAM proteins capable of cleaving substrates, while some inactive proteases lack the catalytic consensus sequence or have a mutated version of it (Schl?ndorff and Blobel 1999; Primakoff and Myles 2000). The disintegrin domain of ADAM proteins provides adhesive properties for those ADAMs having an integrin-binding motif at the tip of a flexible protein loop (Primakoff and Myles 2000).
Little is known of the evolutionary relationship among ADAM genes, although there is some evidence to suggest that ADAM genes expressed in sperm and ADAM gene domains that function in sperm-egg interactions may be more diverged between species than other ADAM genes (Civetta 2003). There is heterogeneity in evolutionary rates within fertilin (ADAM 1) and fertilin ? (ADAM 2) with the prodomain and the metalloprotease domains more conserved than other regions, which was suggested to be a result of their roles in spermatogenesis or sperm capacitance, and positive selection detected in the disintegrin and cysteine-rich adhesion domains (Civetta 2003).
ADAMs are a group of widely expressed genes with potential roles in cell adhesion. Members of the ADAM family may also function as proteases based on whether a metalloprotease domain is present and active in the protein (Primakoff and Myles 2000). To date, 32 ADAM genes have been identified in the mouse, of which 19 have confirmed metalloprotease activity as mature proteins and are expressed either in reproductive tissue (testis or epididimys; n = 10) or somatic tissue (n = 9) (Primakoff and Myles 2000; Brachvogel et al. 2002; Gunn et al. 2002; Choi et al. 2003). At least 12 ADAM genes have been suggested to function in cell adhesion based on the presence of predictive amino acid motifs and/or in vitro binding assays to potential integrin ligands (Sagane et al. 1999; Evans 2002; Frayne, Hurd, and Hall 2002; Choi et al. 2003; table 1). Although few receptors have been confirmed, the disintegrin domain is implicated as the principal region mediating binding between ADAM ligands and integrins (Primakoff and Myles 2000).
Table 1 Average dN/dS Estimates and Proportions of Nonsynonymous (dN) and Synonymous (dS) Substitutions Among ADAM Genes in Mouse-Human Comparisons.
In this study, we collected 17 different mouse and human ADAM gene sequences and tested whether protein characteristics, other than the site of protein expression, could group fast evolving ADAM genes. We used the phylogenetic analysis by maximum likelihood approach (PAML; Yang 1997) on tree topologies from different ADAM genes to test for variable rates of evolution along tree branches and to determine the protein domain distribution (i.e., amino end, metalloprotease, disintegrin/cysteine-rich, carboxy tail) of positively selected codons.
Materials and Methods
Sequence Data Analysis
The following 17 complete ADAM gene sequences were obtained from GenBank from Mus musculus and Homo sapiens respectively: ADAM 2 (u38806; u38805), ADAM 7 (nm_007402; af215824), ADAM 8 (nm_007403; nm_001109), ADAM 9 (nm_007404; nm_003816), ADAM 10 (nm_007399; af009615), ADAM 11 (nm_009613; ab009675), ADAM 12 (nm_007400; af023477), ADAM 15 (ab022089; nm_003815), ADAM 17 (nm_009615; nm_003183), ADAM 18 (nm_010084; nm_014237), ADAM 19 (nm_009616; af311317), ADAM 21 (nm_020330; af158644), ADAM 22 (ab009674; nm_021722), ADAM 23 (nm_011780; nm_003812), ADAM 28 (af153350; nm_021778), ADAM 32 (af513715, nm_145004), ADAM 33 (nm_033615; nm_025220).
Amino acid sequence alignments were obtained using the ClustalW program within the DAMBE software package and used to generate nucleotide alignments (Xia 2000). An ADAM gene phylogeny for the 17 mouse and human sequences was constructed using distance (Neighbor-Joining), maximum parsimony (MP), and maximum likelihood (ML) methods (Mega 2.1; PAUP*4.0) (Swafford 1992; Kumar et al. 2001). The most parsimonious consensus tree was generated under the Tree Bisection-Reconnection (TBR) branch-swapping algorithm. The MP tree was evaluated using maximum likelihood methods under the heuristic model with 25 random-addition replicates and a transition/ transversion ratio = 2:1. The reliability of the inferred tree was assessed by bootstrapping with 1,000 replicates, and bootstrap values with scores greater than 80% were retained. For the maximum likelihood tree reconstruction, we used nucleotide substitution models that consider only transition/transversion bias (F84 model) or both transition/ transversion bias and transition rate heterogeneity (Tn93 model). The log likelihood ratio test (LR = 2 [InLTn93–InLF84]) (Nei and Kumar 2000, p. 70) was used to assess which model best fit the underlying data.
Proportion of nonsynonymous substitutions per nonsynonymous site (dN) and synonymous substitutions per synonymous site (dS) between mouse and human were calculated for each ADAM gene using the modified Nei-Gojobori Jukes Cantor method (Nei and Gojobori 1986; Nei and Kumar 2000, pp. 57–60), which takes into account deviations from a transition/transversion (R) = 1. We compared dN/dS ratios by grouping ADAM genes based on phylogenetic clustering, protease activity, potential ligand binding, and tissue of expression. Statistical comparisons of average dN/dS ratios among groups were tested by approximate randomization analysis with 25,000 permutations (Manly 1991).
Variation in Selective Pressures Across Lineages
For eight of the seventeen ADAM genes used to reconstruct the gene family phylogeny, sequences are available from GenBank for at least rat (Rattus novergicus) and/or macaque (Macaca fascicularis) (accession numbers available upon request). Tree topologies were generated using groups of three species, for a total of four possible combinations: human/rat/macaque, human/rat/mouse, macaque/rat/mouse, and human/mouse/macaque. We used tree topologies obtained from the clustal alignment to test for variation in selective pressures measured as dN/dS per codon () across lineages using the CODEML software package in PAML (Yang 1997). Likelihood estimates for each tree topology under the assumption of a unique ratio for all tree branches (Model 0) was compared to estimates from a free-ratio model (Model 1, variable among branches). This was achieved by changing the parameters for model to options 0 or 1 in the CODEML control file while keeping the NSsite option that allows for variation in across codons fixed at one ratio value (M0). A likelihood ratio test was calculated by multiplying twice the difference between log likelihood scores () obtained under each model and comparing the result against a 2 distribution with degrees of freedom equal to the difference in the number of parameters estimated in each model (Nielsen and Yang 1998; Yang et al. 2000). To maximize statistical power the analysis was repeated with ADAM genes for which four or more different mammalian species sequences were available (accession numbers available upon request).
Variation in Selective Pressure Across Codon Sites
We used the CODEML software package in PAML to detect specific codon sites under positive selection among sites experiencing variable selective pressures by comparing likelihood estimates of a tree topology under a model that assumes a distribution of values restricted between 0 and 1 (M7, no positive selection) with an alternative model that allows for values greater than 1 (M8 model). For this purpose, the parameter for model in the CODEML control file was set to 0 while changing the NSsite option to 7 (M7) or 8 (M8). Comparing estimates between M7 and M8 depicts the existence of codons under positive selection (Yang et al. 2000). If the log-likelihood test suggests the presence of sites under positive selection, then these sites can be identified by using a Bayesian method to estimate posterior probabilities (P) that particular sites have experienced a value greater than 1. Positively selected sites identified using this approach were assigned to different protein domains previously identified by using the Motif scan tool available in the PROSITE database of protein families and domains (http://hits.isb-sib.ch/cgi-bin/PFSCAN).
Results
Human-Mouse dN/dS Comparisons
The ancestral relationships among ADAM genes generated by distance, parsimony, and likelihood methods correspond to previous partial phylogenies of the gene family (Poindexter et al. 1999).
Phylogenetic analysis identified three ancestral gene clusters based on bootstrap estimates: cluster 1 is comprised exclusively of nonproteolytic genes expressed in the brain, and clusters 2 and 3 are a mix of proteolytic and nonproteolytic, somatic, and sex-related genes (fig. 1; see also online Supplementary Material). dN/dS ratio estimates between mouse and human varied across genes (one-sample t-test: t16 = 7.23, P < 0.001; table 1). When dN/dS estimates are grouped on the basis of their phylogenetic clustering, proteolytic activity, adhesion, or tissue of expression, results show a significantly higher average dN/dS for ADAM genes expressed in reproductive tissue versus nonsexual tissues (D6,11 = 0.193; P = 0.001; table 2). Interestingly, phylogenetic cluster 1 comprised exclusively of nonproteolytic genes expressed in the brain shows the lowest dN/dS ratio estimate (fig. 1 and table 2).
FIG. 1. Unrooted maximum parsimonious tree of ancestral phylogeny generated by PAUP*4.0 using Homo sapiens (Hs) and Mus musculus (Mm) ADAM gene sequences. Bootstrap values were 100% for all mouse-human nodes and are not shown. Other bootstrap values higher than 80% are shown in bold type
Table 2 Approximate Randomization of dN/dS on ADAM Gene Sequences Differentiated on the Basis of Phylogenetic Clusters, Proteolytic Structure, Tissue of Expression, and Ligand Function.
Although the average dN/dS rate is significantly higher for ADAM genes expressed in reproductive tissue compared to others, all ADAMs show dN/dS ratios less than 1 that are indicative of purifying selection. It is possible that some ADAM genes have experienced episodes of positive selection and that such signals remained undetected by using standard dN/dS estimates for the entire gene sequences.
Variation in Selective Pressures Across Lineages
Only four ADAM genes (ADAM 2, 17, 18, and 28) show evidence of significant variation in selective pressures across branches in the tree (table 3). Higher values seem to be found within primate branches (fig. 2) and this is supported by a significantly higher for either of the primate branches compared to the rodent tree branch in human, macaque, and mouse (or rat) three-species comparisons (paired t: t6 = –2.83, P = 0.030). This observation could be a consequence of true differences in selective pressures or an underestimation of the ratio in the rodent tree branch associated with larger rodent populations (Yang and Nielsen 1998). It is also possible that failure to reject the one-ratio model for phylogenies in which rodent lineages outnumber primate lineages is due to lack of power (Anisimova, Bielawski, and Yang 2001).
Table 3 Variation in Selective Pressures Along Different ADAM Gene Lineages.
FIG. 2. Phylogeny and branch estimates for (a) ADAM 7, (b) ADAM 2, (c) ADAM18 and (d) ADAM10 using Bos taurus (Bt), Cavia porcellus (Cp), Homo sapiens (Hs), Macaca fascicularis (Mf), Mus musculus (Mm), Oryctolagus cuniculus (Oc), Rattus norvegicus (Rn), and Sus scrofa (Ss)
Variation in Selective Pressures Across Codon Sites
The phylogenetic analysis by maximum likelihood (PAML) was used to estimate the likelihood of a phylogeny under models that make alternative assumptions about the dN/dS rate of change among codon sites () (Nielsen and Yang 1998; Yang et al. 2000). The tree topologies for groups of more than three species used for the analysis are shown in figure 2. We tested for the presence of positively selected sites by comparing the log-likelihood of a tree under a model that assumes a distribution of values constrained between 0 and 1 (M7) with the estimate obtained under an alternative model that adds a class of sites with ratios >1.0 (M8). The likelihood ratio tests obtained from comparing M7 and M8 shows that M8 fits the data better than M7 for ADAM genes expressed in both testis (ADAMs 2, 32) and somatic tissue (ADAMs 17, 28) indicating the occurrence of positive selection at codon sites within genes (table 4). However, a higher proportion of positively selected sites were found for sperm ADAMs 2 and 32 than ADAM 17 and 28 (table 4). There is also a difference in the distribution of positively selected sites across these genes. While all ADAM genes show a high concentration of positively selected sites in the carboxy tail region, only sperm ADAM genes show the highest proportion of positively selected sites in the disintegrin/cysteine-rich domain with potential adhesion function (fig. 3 and table 5).
Table 4 Variation in Selective Pressure Across Codon Sites for ADAM Gene Lineages in Groups of Three or More Species Comparisons.
FIG. 3. Posterior probability of positively selected codon sites ( > dN/dS) and their distribution among the amino, metalloprotease (grey rectangle), disintegrin/cysteine-rich (black rectangle), and carboxy tail domains of ADAM 2, ADAM 17, ADAM 28, and ADAM 32. The horizontal line indicates a posterior probability of 0.95
Table 5 Proportion of Positively Selected Codon Sites per Domain (P > 0.5) and Amino Acids in Mus musculus with a Posterior Probability >0.95 of Being Under Positive Selection.
Discussion
Approximate randomization analysis supports previous results showing that reproductive genes, including those implicated in fertilization and spermatogenesis, evolve at a faster rate relative to genes associated with other tissues, in agreement with numerous other studies (Civetta and Singh 1988, 1999; Singh and Kulathinal 2000; Wyckoff, Wang, and Wu 2000; Swanson et al. 2001). Maximum likelihood analysis revealed that ADAM genes expressed in the testis with the potential for sperm-egg adhesion are subjected to positive selection.
Genes expressed in both the testis and somatic tissue showed positive selection of codon sites in the carboxy tail, situated on the transmembrane and cytoplasmic segment of the ADAM gene. The role of the carboxy tail in ADAM proteins is poorly understood, although it may be important for localization of the sperm head prior to sperm-egg adhesion (Hunnicut, Koppel, and Myles 1997; Myles and Primakoff 1997; Cowan et al. 2001) or in signal transduction (Evans, Schultz, and Kopf 1998; Kang, Cao, and Zolkiewska 2000). There is also a potential adhesion role identified for the carboxy tail region of ADAMs 17 and 28 that may involve binding of proline-rich regions on the carboxy tail to SH3-domain-containing proteins (Howard, Maciewicz, and Blobel 2003) or of ADAM 17 to mitotic arrest deficient 2 (MAD2), a component of the mitotic machinery (Nelson, Schl?ndorff, and Blobel 1999), although these functions remain speculative.
Among ADAMs showing signs of positive selection, we found the highest proportion of positively selected codon sites in the disintegrin/cysteine-rich region of genes expressed in the testis (ADAMs 2 and 32) (fig. 3). Similar results were found for ADAM 1, a heterodimer of ADAM 2 that plays a direct role in sperm-egg adhesion (Civetta 2003). ADAM 2 is supposed to be implicated in sperm-egg adhesion as a ligand of integrins located on the surface of the egg (Evans, Kopf, and Schultz 1997). Such adhesive function is mediated by sites located in the disintegrin domain of ADAM 2 (Zhu et al. 2000). Mice with an ADAM 2 knockout are sterile; in in vitro assays the null sperm is defective in binding to the egg zona pellucida and is rarely found in the oviduct of females despite normal motility. ADAM 2 null males also show a reduced ability to fuse the egg membrane, suggesting that ADAM 2 is required but not essential in sperm-egg membrane binding (Cho et al. 1998). ADAM 32 shares a close sequence homology to ADAM 2, including a sequence similar to the conserved binding motif of the disintegrin loop, suggesting a potential role in gamete adhesion (Choi et al. 2003).
ADAM genes with the potential for mediating cell-cell adhesion within endogenous tissues and/or with a role in embryonic development were found to be under purifying selection. Included in this category are ADAMs 11, 22, and 23, potential ligands of integrins in neural brain tissue (Sagane et al. 1999); ADAMs 9, 12, and 15, genes with adhesion roles in fibroblasts, myoblasts and haemopoietic cells (Nath et al.1999, 2000; Eto et al. 2000); and ADAMs 7 and 18, genes expressed in the epididymis with roles in sperm development (Cornwall and Hsia 1997; Frayne, Hurd, and Hall 2002).
Our study is the first to examine the evolution of ADAM genes in the context of a disintegrin multigene family. The ADAM proteins are thought to have evolved as a multidomain structure from a proteolytic ancestor and for which pressure on the disintegrin adhesion domain is the suggested mechanism promoting functional diversification (Moura-da-Silva, Theakston, and Crampton 1996). This process is similar to selection for antigen recognition sites of major histocompatability loci (MHC) (Hughes and Nei 1988; Hughes, Ota, and Nei 1990) and heterogeneity at binding sites within the antifreeze protein (AFP) multigene family (Swanson and Aquadro 2002).
Evidence for rapid evolution of reproductive and immune system genes has led to speculation that genes binding to "foreign" substances may be generally under positive selection (e.g., Wyckoff, Wang, and Wu 2000). Our results show that positive Darwinian selection is restricted to the adhesion domains of ADAM genes involved in sperm-egg plasma membrane adhesion, as opposed to binding domains of all ADAM genes. This is best illustrated by ADAM 28, for which we found no evidence of positive selection at adhesion sites, despite the potential binding role inferred by recognition of the disintegrin domain by integrin 4?1 on human lymphocytes (Bridges et al. 2002).
Our study highlights the faster evolution of reproductive genes as a group and the adaptive diversification of sites within binding domains of sperm-egg adhesion ADAMs. Further functional studies are needed to determine the mechanism and significance of the actual changes driven by positive selection.
Acknowledgements
We would like to thank Ed Byard and Brian Golding for comments on the manuscript. Thanks to Scott Forbes for kindly providing his approximate randomization program. This work was supported by an Establishment Manitoba Health Research Council grant and a research grant from the Natural Sciences and Engineering Research Council of Canada to A.C.
Literature Cited
Anisimova, M., J. P. Bielawski, and Z. Yang. 2001. Accuracy and power of the likelihood ratio test in detecting adaptive molecular evolution. Mol. Biol. Evol. 18:1585-1592.
Brachvogel, B., D. Reichenberg, S. Beyer, B. Jehn, K. von der Mark, and W. Bielke. 2002. Molecular cloning and expression analysis of a novel member of the disintegrin and metalloprotease-domain (ADAM) family. Gene 288:203-210.
Bridges, L. C., P. H. Tani, K. R. Hanson, C. M. Roberts, M. B. Judkins, and R. D. Bowditch. 2002. The lymphocyte metalloprotease MDC-L (ADAM 28) is a ligand for the integrin 4?1. J. Biol. Chem. 277:3784-3792.
Cho, C., D. O'Dell Bunch, J-E. Faure, E. H. Goulding, E. M. Eddy, P. Primakoff, and D. G. Myles. 1998. Fertilization defects in sperm from mice lacking fertilin ?. Science 281:1857-1859.
Choi, I., J-M. Woo, S. Hong, Y. K. Jung, D. H. Kim, and C. Cho. 2003. Identification and characterization of ADAM32 with testis-predominant gene expression. Gene 304:151-62.
Civetta, A. 2003. Positive selection within sperm-egg adhesion domains of fertilin: an ADAM gene with a potential role in fertilization. Mol. Biol. Evol. 20:21-29.
Civetta, A., and R. S. Singh. 1998. Sex-related genes, directional sexual selection, and speciation. Mol. Biol. Evol. 15:901-909.
Civetta, A., and R. S. Singh. 1999. Broad-sense sexual selection, sex gene pool evolution, and speciation. Genome 42:1033-1041.
Cornwall, G. A., and N. Hsia. 1997. ADAM7, A member of the ADAM (A Disintegrin And Metalloprotease) gene family is specifically expressed in the mouse anterior pituitary and epididymis. Endocrinology 138:4262-4272.
Cowan, A. E., D. E. Koppel, L. A. Vargas, and G. R. Hunnicut. 2001. Guinea pig fertilin exhibits restricted lateral mobility in epididymal sperm and becomes freely diffusing during capacitation. Dev. Biol. 236:502-509.
Eto, K., W. Puzon-McLaughlin, D. Sheppard, A. Sehara-Fujisawa, X-P. Zhang, and Y. Takada. 2000. RGD-independent binding of integrin 9?1 to the ADAM-12 and -15 disintegrin domains mediates cell-cell interaction. J. Biol. Chem. 275:34922-34930.
Evans, J. P., G. S. Kopf, and R. M. Schultz. 1997. Characterization of the binding of recombinant mouse sperm fertilin ? subunit to mouse eggs: Evidence for adhesive activity via an egg ?1 integrin-mediated interaction. Dev. Biol. 187:79-93.
Evans, J. P., R. M. Schultz, and G. S. Kopf. 1998. Roles of the disintegrin domains of mouse fertilins and ? in fertilization. Biol. Reprod. 59:145-152.
Evans, J. P. 2002. The molecular basis of sperm-oocyte membrane interactions during mammalian fertilization. Hum. Reprod. Update 8:297-311.
Frayne J., E. A. Hurd, and L. Hall. 2002. Human tMDC III: a sperm protein with a potential role in oocyte recognition. Mol. Hum. Reprod. 8:817-22.
Gunn, T. M., A. Azarani, P. H. Kim, R. W. Hyman, R. W. Davis, and G. S. Barsh. 2002. Identification and preliminary characterization of mouse ADAM33. BMC Genet. 3:2-9.
Howard, L., R. A. Maciewicz, and C. P. Blobel. 2003. Cloning and characterization of ADAM28: evidence for autocatalytic pro-domain removal and for cell surface localization of mature ADAM28. Biochem. J. 348:21-27.
Hughes, A. L., and M. Nei. 1988. Pattern of nucleotide substitution at MHC class I loci reveals overdominant selection. Nature 335:167-170.
Hughes, A. L., T. Ota, and M. Nei. 1990. Positive Darwinian selection promotes charge profile diversity in the antigen-binding cleft of class I major-histocompatibility-complex molecules. Mol. Biol. Evol. 7:515-524.
Hunnicut, G. R., D. E. Koppel, and D. G. Myles. 1997. Analysis of the process of localization of fertilin to the sperm posterior head plasma membrane domain during sperm maturation in the epididymis. Dev. Biol. 191:146-159.
Kang, Q., Y. Cao, and A. Zolkiewska. 2000. Metalloprotease-disintegrin ADAM 12 binds to the SH3 domain of Src and activates Src tyrosine kinase in C2C12 cells. Biochem. J. 352:883-892.
Kumar, S., K. Tamura, I. B. Jacobsen, and M. Nei. 2001. MEGA2: molecular evolutionary genetics analysis software. Bioinformatics 17:1244-1245.
Manly, B. J. F. 1991. Randomization and Monte Carlo methods in biology. Chapman and Hall, London.
Metz, E. C., and S. R. Palumbi. 1996. Positive selection and sequence rearrangements generate extensive polymorphism in the gamete recognition protein bindin. Mol. Biol. Evol. 13:397-406.
Moura-da-Silva, A. M., R. D. G. Theakston, and J. M. Crampton. 1996. Evolution of disintegrin cysteine-rich and mammalian matrix-degrading metalloproteinases: gene duplication and divergence of a common ancestor rather than convergent evolution. J. Mol. Evol. 43:263-269.
Myles, D. G., and P. Primakoff. 1997. Why did the sperm cross the cumulus? To get to the oocyte. Functions of the sperm surface proteins PH-20 and fertilin in arriving at, and fusing with, the egg. Biol. Reprod. 56:320-327.
Nath, D., P. M. Slocombe, P. E. Stephens, A. Warn, G. R. Hutchinson, K. M. Yamada, A. J. Docherty, and G. Murphy. 1999. Interaction of metargidin (ADAM-15) with v?3 and 5?1 integrins on different haemopoietic cells. J. Cell Sci. 112:579-587.
Nath, D., P. M. Slocombe, A. Webster, P. E. Stephens, A. J. P. Docherty, and G. Murphy. 2000. Meltrin (ADAM-9) mediates cellular adhesion through 6?1 integrin, leading to a marked induction of fibroblast cell motility. J. Cell Sci. 113:2319-2328.
Nei, M., and T. Gojobori. 1986. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol. Biol. Evol. 3:418-426.
Nei, M., and S. Kumar. 2000. Molecular evolution and phylogenetics. Oxford University Press, New York.
Nielsen, R., and Z. Yang. 1998. Likelihood models for detecting positively selected amino acid sites and applications to the HIV-1 envelope gene. Genetics 148:929-936.
Nelson, K. I., J. Schl?ndorff, and C. P. Blobel. 1999. Evidence for an interaction of the metalloprotease–disintegrin tumour necrosis factor convertase (TACE) with mitotic arrest deficient 2 (MAD2), and of the metalloprotease–disintegrin MDC9 with a novel MAD2-related protein, MAD2?. Biochem. J. 343:673-680.
Poindexter, K., N. Nelson, R. F. DuBose, R. A. Black, and D. P. Ceretti. 1999. The identification of seven metalloproteinase-disintegrin (ADAM) genes from genomic libraries. Gene 237:61-70.
Primakoff, P., and D. Myles. 2000. The ADAM gene family: surface proteins with adhesion and protease activity. Trends Genet. 16:83-87.
Sagane, K., K. Yamazaki, Y. Mizui, and I. Tanaka. 1999. Cloning and chromosomal mapping of mouse ADAM11, ADAM22, and ADAM23. Gene 236:79-86.
Schl?ndorff, J., and C. P. Blobel. 1999. Metalloprotease-disintegrins: modular proteins capable of promoting cell-cell interactions and triggering signals by protein-ectodomain shedding. J. Cell Sci. 112:3603-3617.
Singh, R. S., and R. J. Kulathinal. 2000. Sex gene pool evolution and speciation: A new paradigm. Genes Genet. Sys. 75:119-130.
Swafford, D. L. 1992. PAUP*: phylogenetic analysis using parsimony (*and other methods). Version 4. Sinauer Associates, Sunderland, Mass.
Swanson, W. J., and V. D. Vacquier. 1995. Extraordinary divergence and positive Darwinian selection in a fusagenic protein coating the acrosomal process of abalone spermatozoa. Proc. Natl. Acad. Sci. USA 92:4957-61.
Swanson, W. J., Z. Yang, M. F. Wolfner, and C. F. Aquadro. 2001. Positive Darwinian selection drives the evolution of several female reproductive proteins in mammals. Proc. Natl. Acad. Sci. USA 98:2509-2514.
Swanson, W. J., and C. F. Aquadro. 2002. Positive Darwinian selection promotes heterogeneity among members of the antifreeze protein multigene family. J. Mol. Evol. 54:403-410.
Swanson, J. W., R. Nielsen, and Q. Yang. 2003. Pervasive adaptive evolution in mammalian fertilization proteins. Mol. Biol. Evol. 20:18-20.
Torgerson, D. G., R. J. Kulathinal, and R. S. Singh. 2002. Mammalian sperm proteins are rapidly evolving: evidence of positive selection in functionally diverse genes. Mol. Biol. Evol. 19:1973-1980.
Wyckoff, G. J., W. Wang, and C-I Wu. 2000. Rapid evolution of male reproductive genes in the descent of man. Nature 403:304-308.
Xia, X. 2000. Data analysis in molecular biology and evolution. Kluwer Academic Publishers, Boston.
Yang, Z. 1997. PAML: a program package for phylogenetic analysis by maximum likelihood. Comput. Appl. Biosci. 13:555-556.
Yang, Z., and R. Nielsen. 1998. Synonymous and nonsynonymous rate variation in nuclear genes of mammals. J. Mol. Evol. 46:409-418.
Yang, Z., R. Nielsen, N. Goldman, and A-M. K. Pedersen. 2000. Codon-substitution models for heterogeneous selection pressure at amino acid sites. Genetics 155:431-449.
Zhu, X., N. P. Bansal, and J. P. Evans. 2000. Identification of key amino acids of the mouse fertilin ? (ADAM2) disintegrin loop for cell-cell adhesion during fertilization. J. Biol. Chem. 275:7677-7683.(Barb Glassey1 and Alberto)