| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Physiologie de la Reproduction et des Comportements,3 INRA-CNRS-Université de Tours-Haras Nationaux, IFR 135, 37380 Nouzilly, France
Département de Biologie Structurale,4 IMPMC UMR 7590, Universités Pierre et Marie Curie-Paris 6 et Denis Diderot-Paris 7, CNRS, Campus Boucicaut, 75015 Paris, France
ABSTRACT
Vertebrate eggs are surrounded by an extracellular matrix with similar functions and conserved individual components: the zona pellucida (ZP) glycoproteins. In mammals, chickens, frogs, and some fish species, we established an updated list of the ZP genes, studied the relationships within the ZP gene family using phylogenetic analysis, and identified ZP pseudogenes. Our study confirmed the classification of ZP genes in six subfamilies: ZPA/ZP2, ZPB/ZP4, ZPC/ZP3, ZP1, ZPAX, and ZPD. The identification of a Zpb pseudogene in the mouse genome, Zp1 pseudogenes in the dog and bovine genomes, and Zpax pseudogenes in the human, chimpanzee, macaque, and bovine genomes showed that the evolution of ZP genes mainly occurs by death of genes. Our study revealed that the extracellular matrix surrounding vertebrate eggs contains three to at least six ZP glycoproteins. Mammals can be classified in three categories. In the mouse, the ZP is composed of three ZP proteins (ZPA/ZP2, ZPC/ZP3, and ZP1). In dog, cattle and, putatively, pig, cat, and rabbit, the zona is composed of three ZP proteins (ZPA/ZP2, ZPB/ZP4, and ZPC/ZP3). In human, chimpanzee, macaque, and rat, the ZP is composed of four ZP proteins (ZPA/ZP2, ZPB/ZP4, ZPC/ZP3, and ZP1). Our review provides new directions to investigate the molecular basis of sperm-egg recognition, a mechanism which is not yet elucidated.
evolution, fertilization, gamete biology, oocyte development, ovum, pseudogene, zona pellucida
Vertebrate eggs are surrounded by an extracellular matrix called the chorion in fish, the vitelline envelope in amphibians, the perivitelline envelope in reptiles and birds, and the zona pellucida (ZP) in mammals. These extracellular matrices have similar functions. They participate in taxon-specific sperm-egg binding during fertilization and protect the embryo during early development. Moreover, they have similar ultrastructures, composed of fibrous matrices with conserved individual components featuring common protein domains.
The ZP matrix that surrounds all mammalian oocytes is composed of three to four major glycoproteins. Remarkably, the different ZP glycoproteins share an apparent overall similar architecture, around a conserved C-terminal ZP domain, preceding their transmembrane segment. The ZP domain is a large domain (
260 residues) found in a wide variety of extracellular proteins with various functions [1]. Sequences downstream from the ZP domains are much variable and present marked differences between the ZP proteins (see below). The initial nomenclature of the ZP glycoproteins was based on the apparent molecular weight of the mouse ZP proteins after migration by SDS-PAGE. These proteins were named ZP1, ZP2, and ZP3 from the highest to the lowest apparent molecular weight, respectively [2]. Zona pellucida glycoproteins that have been subsequently described in other species have been named according to several criteria, including apparent molecular weight and charge following two-dimensional gel electrophoresis [3], the size of the cDNAs (with Zpa being the largest and Zpc the smallest) [4], as well as sequence identity comparison [5, 6]. This has resulted in a confused nomenclature.
The situation for the pig species was particularly confusing. Electrophoretic analysis suggested that the pig ZP family was composed of four distinct proteins [7]. Other investigators reported that the pig ZP family was composed of three proteins, referred to as PZI, PZII, and PZIII [8, 9], or ZP1, ZP2, and ZP3 [10]. Later studies suggested the presence of a fourth pig ZP protein, referred to as PZIV [11] or 25K [12]. Additional studies suggested that the ZP3 component was actually a mixture of two different proteins, referred to as ZP3alpha and ZP3beta [13].
In an attempt to clarify the relationship between the different classes of ZP genes, Harris et al. [4] proposed a unified system of nomenclature in which ZP genes were named in order of the length of their encoded protein sequences with a new letter-based system. Thus, ZP2 became ZPA, ZP1 became ZPB, and ZP3 became ZPC. However, the number system and the letter system are used concurrently by different groups, increasing the confusion. For example, in humans, ZPB and ZP1 are two distinct genes [14]. In the pig, ZP1 and ZP2, described as separate genes, are two names of the same gene [15, 16]. Later, Conner et al. [17] proposed a simplified naming system in which ZP genes are termed ZP1, ZP2, ZP3, and ZP4. They focused on the four main classes of ZP protein, omitting ZPAX and ZPD subfamilies. Spargo and Hope [15] have also created a logical system of nomenclature on the basis of which an updated classification can be proposed.
Here, we have established an updated list of the ZP genes in mammals, chickens, and frogs. In these species and some fish species we have established the relationships within the ZP gene family using phylogeny. Finally, we have identified ZP pseudogenes in several species. This will help to understand the characteristic evolution of these reproductive proteins.
ANNOTATION OF VERTEBRATE ZP GENES BY PHYLOGENETIC ANALYSIS
The nomenclature used to describe ZP genes and proteins from different mammalian species is very confusing. We established an updated list of the ZP genes in mammals, chickens, and frogs, and we identified orthologous and paralogous genes on the basis of phylogenetic analyses. Phylogenetic analyses were carried out using FIGENIX software (http://www.up.univ-mrs.fr/evol/figenix) [18], which produces a consensus tree according to the three approaches, the Neighbor Joining method [19], the Maximum Parsimony method [20], and the Maximum Likelihood method [21]. The dataset of putative homologous sequences used for phylogenetic reconstruction is built by BLAST that performs local alignment of peptide sequences. Then, sequences are submitted to the MUSCLE software (http://www.drive5.com/muscle) [22, 23], which performs a multiple alignment of complete peptide sequences. The sequences of interest used as query were the amino acid sequences of the pig ZPA/ZP2 and ZPC/ZP3.
Table 1 shows an updated list of the ZP genes in mammals, chickens, and frog, as well as the Swissprot and GenBank accession numbers and the current name of all of these ZP genes. Three to six ZP genes are present in these species. To avoid confusion, fish ZP sequences have been omitted, as there have been both genome and gene duplications giving rise to multiple gene copies [24].
|
An Updated Annotation of ZP Genes
The results of the phylogenetic analyses are shown in Figures 1 and 2 and in the supplemental data (Supplementary Fig. 1 and Supplementary Fig. 2, available online at www.biolreprod.org). When the phylogenetic analysis is built with FIGENIX using the amino acid sequence of the porcine ZPA as query, the phylogenetic tree contains the genes from the ZPA/ZP2, ZPB/ZP4, ZP1, ZPD, and ZPAX subfamilies, but not the ZPC/ZP3 subfamily (Fig. 1 and Supplementary Fig. 1). As previously suggested [5], a gene duplication gave rise to two paralogous groups of genes within the ZPB subfamily, ZP1 and ZPB/ZP4, since the phylogenetic tree shows that they share a common ancestral gene. This common ancestral gene shares a common ancestral gene with the ZPA/ZP2 subfamily, and the latter shares a common ancestral gene with the ZPAX subfamily. These genes share a common ancestral gene with the ZPD subfamily. This ZPD subfamily shares an ancestral gene with the CUZD1/DMBT1 gene subfamily (CUB and zona pellucida-like domains 1/Deleted in Malignant Brain Tumors 1). When the phylogenetic analysis is built using the amino acid sequence of the porcine ZPC as query, the phylogenetic tree only contains the genes from the ZPC/ZP3 subfamily (Fig. 2 and Supplementary Fig. 2). This subfamily shares an ancestral gene with the DMBT1 gene subfamily. We never could obtain any consensus tree containing the six ZP subfamilies. This raises the question of an ancestral gene with ZPC. One could hypothesize that the first event in ZP evolution could be a gene duplication event, which gave rise to the ancestral ZPC gene and to the ancestral gene of the ZPA/ZP2, ZPB/ZP4, ZPD, ZP1, and ZPAX subfamilies. Spargo and Hope [15] provide some evidence that the first event in ZP evolution was a gene duplication event, which gave rise to the ancestral ZPC gene and to the precursor of all other ZP gene subfamilies. In fish species, there are several ZPC/ZP3 genes. For example, the phylogenetic tree exhibits four Oryzias latipes ZPC-ZP3 genes—ZPC2, ZPC3, ZPC4, and ZPC5—and three Danio rerio ZPC/ZP3 genes—ZP3, ZP3a and ZP3b. As stated above, due to the degree of diversity of ZP in these species, the clarification of fish ZP gene classification requires further analysis that is not presented here because it would have increased the complexity of our data.
|
|
The ZP subfamilies share an ancestral gene with the CUZD1/DMBT1 gene subfamily. The CUZD1/DMBT1 subfamily contains proteins that incorporate two domains, the CUB domain and the ZP domains [25–29]. These two domains are present in two separate families of proteins implicated in sperm-egg recognition: the CUB domain is found in the spermadhesins [30], and the ZP domain is found in the ZP proteins [31]. Moreover, the CUZD1/DMBT1 proteins are expressed in the female genital tract under estrogen regulation [25–27, 32]. This suggests that these proteins may possess a role in fertilization.
The phylogenetic analysis allowed us to classify ZP genes in six subfamilies: the ZPA/ZP2 subfamily, the ZPB/ZP4 subfamily, the ZPC/ZP3 subfamily, the ZP1 subfamily, the ZPAX subfamily, and the ZPD subfamily. It should be noted that two so-called Macaque (Mra) ZP1 were identified, one being reclassified here in the ZPB/ZP4 subfamily [33], and the other in the ZP1 subfamily. Moreover, the so-called marmoset (Cja) ZP1 gene [34] was reclassified here in the ZPB/ZP4 subfamily, and the so-called pig ZP1 gene [35] was reclassified here in the ZPA/ZP2 subfamily, because the genes previously named ZP1 [35], ZP2 [36], and ZPA [37] in this species encode for the same protein. There are actually three distinct ZP genes in the pig: the ZPA/ZP2 gene previously called ZP1 [10], the ZPB/ZP4 gene previously called ZP3alpha [13], and the ZPC/ZP3 gene previously called ZP3beta [13].
The murine ZP is composed of three major glycoproteins, ZP1, ZP2, and ZP3 [2], and it was suggested that the human ZP was composed of three distinct proteins, ZPA, ZPB, and ZPC [4]. The complete cDNA sequence of the mouse Zp1 gene was determined subsequently to the sequence of the human ZPB cDNA sequence [4, 38]. It was assumed that the mouse ZP1 was orthologous to human ZPB [38] because it shared greater amino acid sequence identity with human ZPB than with human ZPA or ZPC, until a human genomic sequence orthologous to the mouse Zp1 gene and paralogous to the human ZPB gene was identified [39]. Soon after, Bausek et al. [5] suggested that a gene duplication gave rise to two paralogous groups of genes within the ZPB subfamily: ZP1 and ZPB. Our bioinformatic analysis now clarifies the classification of the ZP1 and ZPB genes, with the ZPB/ZP4 subfamily containing the ortholog of human ZPB/ZP4, and the ZP1 subfamily containing the human, chimpanzee, macaque, mouse, rat, and chicken ZP1.
The avian oocyte is surrounded by the perivitelline membrane, which is equivalent to the ZP. Whereas the mammalian ZP has three to four main glycoproteins, the chicken perivitelline membrane is composed of six glycoproteins: ZP1/ZPB1, ZPA/ZP2, ZPB/ZP4/ZPB2, ZPC/ZP3, ZPD/ZPX2, and ZPAX/ZPX1 [40]. The genetic and physical mapping of these genes has been performed [40]. The Xenopus egg envelope contains five main glycoprotein components: ZPA, ZPB, ZPC, ZPD/ZPX2, and ZPAX/ZPX1 [15]. As stated above, the fish gene family is complex, and all of the ZP genes were not present in the tree that we generated, likely due to the high degree of divergence with mammalian ZP genes.
The automatic alignments provided by the BLAST2 software between paralogous genes, which mainly focused on the much conserved ZP domains, show relatively high levels of divergence. These alignments did not take into account the N-terminal extensions which exist in ZP proteins, with the exception of ZPC (as well as the chicken ZPD). We recently showed that these extensions are made of divergent copies of ZP-N domains [41]. ZP-N domains consist in the N-terminal part of ZP domains, which have been shown to exist independently of the C-terminal regions in proteins, such as PLAC1 and Oosp1, and fold autonomously [42]. ZP1 and ZPB possess one divergent copy of the ZP-N domain in their N-termini, whereas ZPA and ZPAX have several (the first one being much more related to the single ZP-N copy of ZP1 and ZPB; Fig. 3).
|
Verification of Orthology Relationship by Localization of Genes in Syntenic Chromosomal Regions
To further ascertain the identity of ZPA, ZPB, ZPC, and ZP1 orthologous genes in the rat, mouse, and human, we verified that they are all mapped in syntenic regions in these three species (Table 2). Mapping of the human, mouse, and rat ZP genes in the corresponding genomes was performed using both the Ensembl genome browser software (http://www.ensembl.org/; release 47, October 2007) [43] and the BLAT Search software from the UCSC Genome Bioinformatics address (http://genome.ucsc.edu) [44]. In particular, the human chromosome 16 region that contains ZPA/ZP2 is syntenic to the murine chromosome 7 region that contains Zpa/Zp2. The chromosomal localization of the rat Zpa/Zp2 is not established. However, the ZPA/ZP2 gene on human chromosome 16 is localized close to (approximately 200 kb) the dynein axonemal heavy polypeptide 3 gene (DNAH3, GenBank accession no. NM_017539, gene map locus 16p12) [45, 46]. The rat gene orthologous to human DNAH3 maps to chromosome 1 (gene map locus 1q35), which shows conserved synteny with the region of human chromosome 16, in which both DNAH3 and ZPA are localized. Moreover, the human chromosome 1 region that contains ZPB/ZP4 is syntenic to the rat chromosome 17 region that contains Zpb/Zp4. The human chromosome 7 region that contains ZPC/ZP3 is syntenic to the murine chromosome 5 region and to the rat chromosome 12 region that contain Zpc/Zp3. The human chromosome 11 region that contains ZP1 is syntenic to the murine chromosome 19 region and to the rat chromosome 1 region that contain Zp1.
|
Table 3 summarizes the diversity of ZP genes expressed in mammals, chicken, and Xenopus.
|
IDENTIFICATION OF ZP PSEUDOGENES
The phylogenetic analysis suggested that in several mammals the presence of ZP1, ZPB, ZPD, and/or ZPAX gene is lacking. Moreover, a search in EST databases revealed that there are no sequences with a high level of identity with human ZP1 or chicken ZPAX in cattle, pig, dog, and rabbit, and no sequence with a high level of identity with rat Zpb/Zp4 in the mouse. This suggested the presence in the genome of these species of pseudogenes (genes that have evolved by generating stop codon and/or insertion/deletion disrupting the reading frame and resulting in the loss of their protein-coding ability).
Identification of a Zpb Pseudogene in the Mouse Genome
To identify an eventual Zpb/Zp4 pseudogene in the mouse, we submitted the nucleotide and the amino acid sequences of the rat Zpb/Zp4 to the BLASTn and the tBLASTn softwares against the mouse genome. We observed on the mouse genome an alignment between the nucleotide sequence of rat Zpb and sequences from the mouse syntenic region (Supplementary Fig. 3, available online at www.biolreprod.org). Interestingly, we observed a microdeletion of 19 nucleotides beginning at 438 147 bp, leading to a frame shift in the open reading frame and the appearance of premature stop codons at 438 187 and 438 220 bp (Fig. 4). This microdeletion is likely responsible for the loss of ZPB protein in the mouse. We also verified that the newly identified mouse Zpb pseudogene, located in the region A2 of chromosome 13, mapped in a genome region syntenic to human and rat genome regions carrying corresponding Zpb/Zp4 genes (gene map locus: 1q43 for human and 17q12.1 for rat). Overall, these in silico evidences are supported by recent data using mass spectrometry analysis, which failed to identify mouse ZPB/ZP4 [47]. Numerous peptides from ZP1, ZPA/ZP2, and ZPC/ZP3 were identified, but the authors failed to identify any peptides that could correspond to ZPB/ZP4.
|
Identification of ZP1 Pseudogenes in the Dog and Bovine Genome
By using the same strategy, we searched for a pseudogene of ZP1 in the dog genome. A tBLASTn analysis with the human ZP1 protein sequence revealed a significant alignment with the exons 2, 3, 5, 7, 8, and 9 (from 42% to 82% similarity). This analysis also revealed the existence of at least three stop codons at positions 151, 279, and 421, replacing, respectively, a proline in exon 3, a glutamine in exon 5, and a tyrosine in exon 8 of the human ZP1 protein sequence (Fig. 5). The alignment also revealed the existence of microinsertions/deletions in the same exons, but it was not possible with a BLASTn analysis to describe them precisely. We also verified that the ZP1 dog pseudogene is located on chromosome 18 in the dog genome, near GPR44 and PRPF19 canine genes, in a region corresponding to a syntenic region of chromosome 11 in human genome, near the GPR44 and PRPF19 human counterpart genes.
|
A similar work allowed the identification of a ZP1 pseudogene on chromosome 29 in the bovine genome, with microinsertions and a stop codon (Supplementary Fig. 3).
Identification of ZPAX Pseudogenes in the Human, Chimpanzee, Rhesus Macaque, and Bovine Genome
A tBLASTn analysis with the chicken ZPAX protein sequence on the human genome showed an alignment between the nucleotide sequence of chicken ZPAX and sequences from the human syntenic region (Supplementary Fig. 3). We observed a stop codon in the exon 5 (Fig. 6) and an insertion of six nucleotides and a stop codon in exon 3. We also verified that the ZPAX human pseudogene is located on chromosome 2p24.2 in the human genome, between the MSGN1 and KCNS3 human genes and in a region corresponding to a syntenic region of chromosome 3 in the chicken genome, near the MSGN1 and KCNS3 chicken counterpart genes.
|
A similar work allowed the identification of a ZPAX pseudogene located on chromosome 2A in the chimpanzee genome, on chromosome 13 in the Rhesus macaque (Macaca mulata) genome, and on chromosome 11 in the bovine genome, between the MSGN1 and KCNS3 genes (Supplementary Fig. 3). An alignment of the human ZPAX pseudogene and the chimpanzee ZPAX pseudogene showed the great homology between these two pseudogenes (88% identity). Moreover, the mutations observed in these two pseudogenes have similar localizations.
We did not find any traces of ZPD pseudogenes in any mammalian genomes. In the chicken genome, the ZPD gene is located on chromosome 11. There is no predicted coding sequence in the syntenic region of all mammalian genomes examined, suggesting that the ZPD gene has completely disappeared in these species.
A summary of the pseudogenes that have been identified is presented in Table 3. Note that for those that have not been found, either they are present but not yet revealed because the genome is not completely sequenced, or the genomic fragment has been completely lost. The complete sequencing of other vertebrate genomes will definitively elucidate which species have lost ZP1, ZPAX, and ZPD genes during evolution.
ZP GENE EVOLUTION: DEATH OF GENES FROM LOWER VERTEBRATES TO MAMMALS
The phylogenetic analysis allowed us to classify ZP genes in six subfamilies: the ZPA/ZP2 subfamily, the ZPB/ZP4 subfamily, the ZPC/ZP3 subfamily, the ZP1 subfamily, the ZPAX subfamily, and the ZPD subfamily. The mammalian genome contains three to four ZP genes. The Xenopus genome contains five ZP genes. The chicken genome contains six ZP genes. In fish, there are at least seven genes encoding ZP proteins, ZPC being in particular highly duplicated in Medaka and Zebrafish. The identification of pseudogenes showed that ZP gene evolution mainly occurs by death of genes.
Since the chicken genome contains a ZP1 gene, the duplication of the ancestor of this gene can be dated before the divergence between birds and mammals (300Mya). After this divergence, ZPAX and ZPD genes seem to have disappeared in mammals, not in birds. In particular, the death of the ZPAX gene occurred before the divergence between humans and monkeys, since similar mutations were observed in human and chimpanzee ZPAX pseudogenes. In mammals, only primates and rodents contain a ZP1 gene, whereas we observed a ZP1 pseudogene in cows and dogs, which suggests that the death of the ZP1 gene in those species occurred after the divergence between primate and rodent groups, and other mammals. Finally, we observed a ZPB pseudogene in the mouse, which shows that the death of this ZPB gene occurred after the divergence between the mouse and the rat. The persistence of both ZPA/ZP2 and ZPC/ZP3 genes across the vertebrates indicates that both genes have functional importance.
Overall, it seems that there is a sort of discontinuity in the loss of the ZP genes along the phylogeny, suggesting that this loss of ZP genes along the evolution of mammals was independent between species. The significance of this loss of ZP genes in mammals compared with other vertebrates remains obscure. It can be argued that a higher number of ZP genes are necessary for species for which fertilization is external to the female genital tract, whereas a lower number are necessary for species for which fertilization is internal. However, since fertilization is internal in chickens, this would not be the good or the sole explanation.
The amino acid sequences of the ZPA/ZP2, ZPB/ZP4, ZPC/ZP3, and ZP1 families present a high degree of identity between species. This implies a structural similarity in the ZP structures of these species and may suggest some similarities in the mechanisms of sperm-zona interaction. However, this review raises new insights into the mechanisms of sperm-ZP interaction.
Our analysis revealed that mammals can be classified in three categories (Fig. 7). In the first category, composed of human, chimpanzee, macaque, and rat, the ZP is composed of four ZP proteins: ZPA/ZP2, ZPB/ZP4, ZPC/ZP3, and ZP1. In the second category, composed of dog, cattle and, putatively, pig, cat, and rabbit, the ZP does not contain ZP1 protein. In the third category, composed of only the mouse, ZP does contain ZP1 but not ZPB/ZP4. So all ZPs in mammals share ZPA/ZP2 and ZPC/ZP3 proteins, and one or both of the ZP1 and ZPB/ZP4 proteins. Interestingly, these two latter genes share an ancestral gene, suggesting a closest similarity between the two corresponding proteins than with ZPA/ZP2 and ZPC/ZP3. Overall, this suggests that the mechanism of sperm-zona interaction requires the presence of both ZPA/ZP2 and ZPC/ZP3 proteins, and one or both of the ZP1 and ZPB/ZP4 proteins.
|
Current models for the structure of mammalian ZP are based upon the existence of three ZP proteins. In particular, data obtained in the mouse raise the hypothesis that the mammalian ZP has three proteins, of which ZPC/ZP3 is the primary sperm receptor [48]. Since four ZP genes are expressed in the human, chimpanzee, macaque, and rat oocyte, a re-evaluation would be required, both on the structure of the ZP and, potentially, on the mechanisms of sperm-ZP interaction. Moreover, in the pig, sperm membranes present a high affinity for ZPB-ZPC heterocomplexes but not free ZPB or ZPC glycoprotein subunits [49], whereas in the mouse, ZPC/ZP3 is the primary sperm receptor. This result was puzzling, since porcine ZPB/ZP4 and ZPC/ZP3 were thought to be orthologous to mouse ZP1 and ZPC/ZP3. Here, we show that the ZPB/ZP4 protein has been lost in the mouse species. This would explain that spermatozoa only interact with ZPC/ZP3 in this species. This also suggests that different mechanisms of sperm-egg interaction have evolved in mammals. Knockout experiments in which murine ZP proteins were replaced by human equivalents [50, 51] showed that although mouse sperm is able to bind mouse oocytes engineered to express human ZPA/ZP2 and ZPC/ZP3 proteins, human sperm did not bind. These findings suggest either that the species-specific recognition of the ZP is mediated by the mouse-specific glycosylation patterns, regardless of the amino acid sequence, or that human sperm has evolved to interact with a ZP composed of four, not three, proteins. The role of carbohydrate moieties remains to be clarified. In fact, the amino acid sequence could be sufficient for sperm recognition in some mammals, in which ZP proteins expressed in Escherichia coli, a species incapable of synthesizing glycoproteins, bind to homologous spermatozoa and stimulate acrosomal exocytosis. This is the case in the macaque [52, 53] and the bovine [54]. In humans, contradictory results have been observed [55, 56]. Moreover, if human sperm interacts with a ZP composed of four proteins, it may interact with rat ZP. However, sperm binding assays to ovulated eggs indicate that rat ZPA/ZP2, ZPB/ZP4, ZPC/ZP3, and ZP1 are not sufficient to support human sperm binding [57], whereas mouse and rat sperm bind to mouse ZP (composed of three proteins) and rat ZP (composed of four proteins) [57]. Moreover, mouse and rat sperm bind to zonae composed solely of mouse ZPA/ZP2 and ZPC/ZP3 [58, 59]. Thus, whether the number of glycoproteins in the ZP affects taxon-specific sperm-egg recognition remains to be determined.
In conclusion, the molecular basis of sperm binding to the ZP, an essential first step in egg fertilization, remains an enigma. Our review clarifies the classification and evolution of the ZP gene family and provides new directions to investigate sperm-egg recognition in mammals.
ACKNOWLEDGMENTS
We wish to thank Pierre Pontarotti and Philippe Gouret for their help in the phylogenetic study and for helpful discussion, and Lauranne Paillard-Laurence for technical assistance.
FOOTNOTES
1Supported by grants from Les Haras Nationaux in France and La Région Centre in France to S.M. 
Correspondence: 2FAX: 33 2 47 42 77 43; e-mail: goudet{at}tours.inra.fr
Received: 26 July 2007.
First decision: 29 August 2007.
Accepted: 7 November 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
