ESP as a Mediator of Sperm-Egg Binding and Fusion
Early studies by Yanagimachi and Noda (1970) led to the hypothesis that the equatorial segment is an acrosomal domain specifically involved in binding and fusion of sperm with the oolemma (Yanagimachi, 1994). Subsequently, correlative evidence has supported this long-held paradigm. Indeed, addition of equatorial segment—positive monoclonal antibodies, such as the anti-hamster M1 (Noor and Moore, 1999) or the anti-murine MN9 (Toshimori et al, 1992), to in vitro bioassays involving intact and zona-free eggs inhibits the binding and fusion steps of fertilization. In the experiments presented here, the use of polyclonal antiserum generated against ESP revealed a significant inhibition of both sperm binding to and fusion with the egg membrane in comparison with the preimmune serum in the hamster egg penetration assay. Additionally, although a significant decrease in both the binding and fusion was noted, the greater reduction in fusion independent of binding suggests that ESP may be particularly important for this latter step.
It is particularly interesting that all the human sperm that tightly bound to the hamster oocyte surface were ESP positive. One tentative interpretation of this observation may be that an intact equatorial segment may be required to achieve tight binding. Now that a probe specific for the human equatorial segment is available, it may be important to extend such observations to human oocytes.
Regions of ESP protein sequence with homology to other molecules (Figure 7) suggest functionalities that relate both to the location of ESP in the equatorial segment (Wolkowicz et al, 2003) and to binding and fusion events. ESP's N-terminal portion is homologous to the cytolysin family of proteins, which function in oligomerization, membrane insertion, cholesterol binding, and pore formation (Palmer, 2001). A C-terminal domain in ESP homologous to a bacterial type II membrane-binding protein is overlaid on top of a portion of the osteoglycin domain. Type II membrane-binding proteins are used by bacteria in attaching their chromosome to the interior wall of the cell. Lastly, the C-terminus of ESP is homologous to a 68-amino acid region of osteoglycin. Osteoglycin belongs to a large family of leucine-rich repeat proteoglycans that includes decorin, keratocan, fibromodulin, and biglycan (Matsushima et al, 2000) that act through protein-protein interactions and appear to mediate such diverse cellular processes as signal transduction, cell adhesion, and recombination. In fact, ESP does contain a conserved serine residue (ESP amino acid 306) thought to enable covalent cross linking (osteoglycin amino acid 101). This may be important for the final binding and fusion steps of fertilization and may serve to anchor and hold the 2 opposing gametes together, facilitating membrane fusion.
Figure 7. . ESP sequence and homology domains. Figure to scale. Numbering above designates the amino acids at the beginning and end of each domain. In the osteoglycin domain, the location of the serine residue responsible for chondroitin sulfate binding is noted (S). Cytolysin domain, aa58-99; Type 2 membrane-binding protein domain, aa274-304; Osteoglycin domain, aa272-337.
Download figure to PowerPoint
Mechanisms of action of ASA-positive seminal fluids in males and females following ejaculation have been well described (Bronson, 1999). In intact sperm, ESP lies within the acrosomal matrix (Wolkowicz et al, 2003). However, acrosome-reacted sperm may reveal exposed ESP in the cleft between inner and outer acrosomal membranes, thus providing sufficient antibody binding to cause the inhibitory effects observed if antibodies to ESP are present in oviductal fluids. Likewise, spontaneous acrosome reactions in both the male and female reproductive tracts may carry some intra-acrosmal ESP onto the plasma membrane. Anti-ESP antibodies present in seminal plasma may be predicted to affect fertility through this mechanism. Finally, the zona pellucida is known to be permeable to IgG. ASAs present in the zona and perivitelline space of sufficient titer may exert immunologic effects that are localized at the egg following the zona-induced acrosome reaction. Fine structural immunolocalization studies to date have not been successful in simultaneously preserving membranes and retaining immunogenicity of ESP sufficient to allow its precise localization with respect to the plasma membrane, inner and outer equatorial segment membranes, and matrix following the acrosome reaction and after egg binding.
ESP: Relationship to Other Equatorial Segment Proteins
Several proteins found to be associated with the equatorial region have been hypothesized to be involved in fertilization. One group thought to be involved in the final steps of fertilization is the ADAM family of proteins (Primakoff and Myles, 2000). This family, a subset of which is testis specific, contains both disintegrin and metalloprotease domains and therefore is potentially involved in both cell adhesion and protease activities. Indeed, antibodies to the testis-specific ADAM 2 (fertilin β) protein (Primakoff et al, 1987) and peptides representing the disintegrin domain of ADAM 2 (RGD motif; Evans et al, 1995) and ADAM 3 (cyritestin; Yuan et al, 1997) were found to block the adhesion and fusion steps of fertilization. More recently, knockout experiments with both ADAM 2 (Cho et al, 2000) and cyritestin (Nishimura et al, 2001) have demonstrated greatly reduced sperm-egg adhesion but no apparent involvement in the fusion step. ADAM 2 has been localized to the posterior of the head (Hunnicutt et al, 1997), not the equatorial segment.
The cysteine-rich secretory protein (CRISP) family of acidic proteins is thought to regulate calcium channels (Kirchoff, 1998), and some members are transcribed and secreted by the epididymis, where they associate with maturing spermatozoa. One family member, rat epididymal glycoprotein DE (named for isoforms “D” and “E”), associates with the dorsal sperm surface but migrates to the equatorial segment upon capacitation (Rochwerger and Cuasnicu, 1992). Antisera raised against recombinant forms of the mouse and human orthologs (Cohen et al, 2001) of this protein appear to inhibit sperm-egg fusion. However, unlike ESP's immunolocalization, the human ortholog reveals staining of the head as well as the principal and midpieces of the tail (Hayashi et al, 1996). Recently, targeted deletion of the intra-acrosomal protein IZUMO has been shown to result in complete infertility, with sperm entering the perivitelline space without difficulty but unable to fuse with the oolemma (Inoue et al, 2005). Antiserum to human IZUMO blocks human sperm fusion with the oolemma in the hamster egg model.
Previous investigations have provided evidence for subcompartments of the acrosomal matrix and membranes. During mouse spermiogenesis, the Golgi markers giantin, βCOP, golgin 97, and mannosidase II are localized to an acrosomal subcompartment that surrounds the region occupied by acrosin within the acrosomal granule (Ramalho-Santos et al, 2001). Similarly, the vSNARE, VAMP, and tSNARE syntaxin are present in a subdomain surrounding the acrosomal granule in mouse sperm (Ramalho-Santos et al, 2000) and are concentrated in human and rhesus sperm in the equatorial segment of non—acrosome-reacted sperm and persist on the equatorial segment following the acrosome reaction (Ramalho-Santos et al, 2000). Likewise, the cystatin-related CRES gene product has been shown to be present in the equatorial segment of human sperm (Wassler et al, 2002). Coimmunoprecipitation experiments are currently underway using the α–rec-hum-ESP serum to determine whether ESP is complexed or associated with any other equatorial segment proteins and to identify other constituents of the specific equatorial segment compartment that ESP occupies.
Useful ASA reagents have been generated to study sperm subcompartments (Bronson and Tung, 1992; Noor and Moore, 1999; Auer et al, 2000). Monoclonal antibodies HS1A.1 (Villaroya and Scholler, 1986) and MA1–3 (Isahakia and Alexander, 1984) recognize proteins residing only in the principal segment of the human acrosome, whereas D3 (Hinrichsen-Kohane et al, 1985) and 21D3 (Le et al, 1984) bind antigens located only in the equatorial segment. Olson et al (1998) have demonstrated immunohistochemically that 2 major acrosomal matrix proteins, AM22 and AM29, are excluded from the equatorial segment of mouse sperm, whereas the MN9 monoclonal antibody (Toshimori et al, 1992) recognizes 38- and 48-kDa proteins in mouse that are segregated specifically to the equatorial segment. Other researchers reported that the M1 (Noor and Moore, 1999) and P36 (Auer et al, 2000) monoclonal antibodies, which recognize sperm protein bands of 38 and 34 kDa in hamster and humans, respectively, localize to the equatorial segment and inhibit fertilization. However, while the identities and functions of the proteins corresponding to these immunoreagents remain unknown, there is a striking similarity in molecular weight to isoforms of ESP.
Clinical Implications for Infertility
ASAs have been identified in 10%–15% of men experiencing infertility and 15%–20% of women with unexplained infertility (Ghazeeri and Kutteh, 2001). ASAs persist for years following vasectomy and are thought to be one reason for the infertility that may follow vasectomy reversal (for reviews, see Bronson et al, 1994; Marshburn and Kutteh, 1994; Bronson, 1999). The presence of ASAs in clinical cases of infertility, therefore, indicates a possible etiology for the inability to conceive on the part of either the male or female partner. The present experiments demonstrate that ASA-positive infertile male and, to a lesser degree, infertile female sera reacted positively with the 36- to 38-kDa ESP region on 2D gels of human sperm protein extracts, indicating that this antigen is both accessible to and immunogenic for the human immune system. Our live staining, immunofluorescence, and electron microscopy observations (Wolkowicz et al, 2003) have shown ESP to be present within the acrosome, bringing into question how the ESP protein could elicit an immune response in males. It will be noted that some sperm undergo a spontaneous acrosome reaction, thus releasing their acrosomal contents prematurely. Indeed, exposure of the male immune system to the acrosomal contents in the urethra or elsewhere in the postepididymal male reproductive tract may cause the production of antibodies and contribute in some cases to an antifertility effect in the male. Likewise, this same mechanism of premature, spontaneous acrosome reaction could elicit antibodies in the female's lower reproductive tract and infertility via vaginal secretion of antibodies. Endogenously generated ASA-positive sera recognized both native and recombinant ESP on Western blots, indicating the rec-h-ESP retains sufficient primary and secondary structure to be recognized by antibodies to the native antigen. Polyclonal immune sera raised against the cloned recombinant ESP showed that endogenous ESP was localized to the equatorial segment, where it was retained in acrosome-reacted sperm. This same anti-ESP antiserum significantly blocked sperm binding and, more importantly, fusion when tested in vitro. This result is in accord with the finding that ESP is retained in the acrosomes of all sperm tightly bound to the oolemma.
Taken together, these data suggest ESP is an antigen that may be involved in immune infertility in humans and that recombinant ESP may be useful to monitor infertility-associated antibody responses. Intracytoplasmic sperm injection (ICSI) is now the standard of care in many cases of ASA as well as other forms of male infertility. Presently, the incidence of transmission of infertilities of genetic origins (such as motility defects) to the male offspring after ICSI is poorly understood, because the first ICSI males are only now reaching puberty. Interest in differential diagnosis of the causes of male immunoinfertility, such as ASA, may become more important in the future if the acceptability of ICSI in treating some forms of infertility were to wane due to a high rate of transmission of infertilities of genetic origin. In the future, ICSI might continue to be the management tool of choice in the case of ASA, whereas it might not be the management tool of choice for certain infertilities with transmissible genetic origins. A differential diagnosis of immunoinfertile patients with ESP and other molecules would be helpful in deciding which patients should be treated with ICSI therapy. Moreover, ESP's testis-abundant expression, presence on acrosome-reacted sperm and on sperm bound to eggs, as well as the inhibition of fertilization by antibodies to recombinant ESP, indicate that ESP may also be a candidate contraceptive vaccinogen. Currently, we have a number of sperm proteins, including ESP, undergoing testing in macaques to ascertain their immunogenic properties (Herr, 1996; Diekman and Herr, 1997; Kurth et al, 2007).
In addition to previous strategies for establishing the identities of the target antigens recognized by ASAs (Herr et al, 1985; Tsuji et al, 1988; Diekman et al, 1999; Li et al, 2000), the present report shows that screening of 2D Western blots with ASA-positive sera and mass spectrometry can result in the characterization of relevant antigens with key roles in fertilization and demonstrates that proteomic strategies are a powerful approach to dissect the molecular mediators of immunoinfertility.