Immunoinhibition of fertilization in vitro by antibodies to the cumulus oophorus intercellular matrix J. Tesa\l=r%v\\l=i'\k

Centre for , Purkyne University Medical Faculty, nam Sov. hrdinu 11, CS-65677 Brno, Czechoslovakia

Summary. Rabbit polyvalent antiserum raised against the solubilized cumulus matrix was a powerful inhibitor of human fertilization in vitro, affecting \p=n-\zonapellucida interaction. Both sperm binding to, and penetration of, the were severely impaired by the anti\p=n-\cumulusmatrix antiserum, whereas no effects of this antiserum on cumulus matrix solubilization or penetration of zona-free human were evident. Moreover, the anti-cumulus antiserum partly neutralized the reaction-inducing activity of the cumulus matrix. These results warrant further research into the functional specificity of different cumulus matrix components and into the effects of the respective antibodies on reproductive function as a possible lead to a new approach to contraceptive vaccine development.

Keywords: contraceptive vaccine; cumulus oophorus antigens; sperm\p=n-\zonapellucida interaction; sperm ; immunoinhibition of fertilization; man

Introduction

There is general consensus that the contraceptive methods currently available are far from covering the increasing demand for control, especially in Third World countries. Accordingly, a cheap and simple method, affording prolonged yet reversible protection against conception and preferen¬ tially acting before or during the early stages of fertilization, is needed. Since the early seventies, when the concept of immunological interference with fertilization was born and seemed to comply with the above criteria (Shivers, 1974), research in this field has been focussed on both sperm and antigens. Ongoing studies involving sperm antigens deal with enzymes essential to sperm energetic metabolism and with numerous types of membrane antigens (for a recent review see Carrón et al, 1988). On the other hand, the parallel research dealing with egg-derived antigens is virtually limited to antigens from the zona pellucida (reviewed by Henderson et al, 1988). Even though antibodies raised against antigens of the pig zona pellucida cross-react with the human zona pellucida and produce an effective block to fertilization under conditions in vitro (Mori et al, 1985; Henderson et al, 1987), the usefulness of zona-derived antigenic determinants as a basis for a human contraceptive vaccine has been somewhat compromised by the possibility that anti-zona antibodies might also attack ovarian components other than the zona pellucida, with resulting adverse effects on ovarian function. In fact, some in-vivo studies have confirmed the binding of anti-zona antibodies not only to ovulated , but also to those still in their follicles (Sacco, 1979), resulting in serious disturbances of follicular growth with ultimate reduction of the developing pool (Gwatkin et al, 1980; Wood et al, 1981; Gulyas et al, 1983; Skinner et al, 1984). Moreover, adverse effects of anti-zona pellucida antibodies on hormonal status of actively immunized animals have been described (Skinner et al, 1984; Bamezai et al, 1986). Although many of the above side-effects may apparently be overcome by purification of the antigen, the *Present address: INSERM Unite 187, Hôpital Antoine-Béclère, 92141 Clamart, France.

Downloaded from Bioscientifica.com at 09/29/2021 09:42:33PM via free access possibility of developing a contraceptive vaccine based on antigens other than those of the zona pellucida should be reconsidered, especially in the light of the recent data concerning the effects of the cumulus oophorus intercellular matrix on human sperm function (Tesafík, 1985; Tesafík et al, 1988a) and on sperm-zona pellucida interaction (Tesafík et al, 1988b). It has to be stressed that, unlike the zona pellucida, this material is produced late in the preovulatory period and the respect¬ ive antibodies can therefore be expected not to react with components of younger follicles. In the present study, I examined the effects of antiserum against the cumulus intercellular matrix on human sperm-egg interaction in vitro.

Materials and Methods

Source of oocytes and cumuli oophori. Oocytes (n = 45) were obtained from women treated in an in-vitro fertilization programme and came from cases in which the treatment had to be cancelled after oocyte recovery, either due to the inability of the husband to produce a semen sample or to unexpected severe asthenozoospermia. Ovarian stimulation and follicle puncture were carried out as described by Tesafík et al. (1988b). Parts of cumuli oophori were dissected from oocyte-cumulus complexes (other than those subsequently used for testing the effects of anti-cumulus matrix antisera on fertilization) just after recovery using sharp hypodermic needles. Solubilization ofcumulus matrix. Parts of cumuli oophori were washed thoroughly from follicular fluid by incubat¬ ing each cumulus fragment for 10 min in 5 ml Dulbecco's phosphate-buffered saline (DPBS; Serva, Heidelberg, FRG). After this incubation, the cumuli were transferred into a fresh washing solution and the whole procedure was repeated three times. The intercellular matrix was solubilized by incubation of cumuli with 10 units of bovine testicular hyaluronidase (Spofa, Prague, Czechoslovakia) dissolved in 1 ml DPBS at 37°C for 30 min. The basic proteinase inhibitor aprotinin (Antilysin; Spofa) and the SH proteinase inhibitor 4-aminomercury benzoate Na (Serva) were added at concentrations of 100pg/ml and 10pg/ml, respectively. The digested matrix was then separated from cumulus cells by centrifugation at 500g for 15 min. The supernatants were decanted and freeze-dried. Preparation ofzona-free human eggs. After the removal of the cumulus oophorus, some eggs were stripped of their zona pellucida by the combined action of pronase and mechanical forces (Tesafík et al, 1988b). Production ofanti-cumulus matrix antiserum. Two female rabbits were bled before immunization to obtain normal rabbit serum for control incubations. The rabbits were then injected subcutaneously with 250 pg of solubilized cumulus matrix preparations in Freund's complete adjuvant (Gibco, Grand Island, NY, USA). Three booster injec¬ tions of the same preparation emulsified with Freund's incomplete adjuvant were given at 8-week intervals. Rabbits were bled by heart puncture 4 weeks after the last injection, the blood was allowed to coagulate and serum, obtained by centrifugation, was pooled, heated at 56°C for 30 min, absorbed by normal human serum in a 1:1 ratio and stored at 20°C until used. Antibody titres were monitored by ELISA using the solubilized antigen preparation (as above). The— reactivity of the antiserum with intact cumuli oophori spread on microscope slides was checked by indirect immunofluorescence. The antiserum used in this study exhibited a titre between 1:2000 and 1:2500. The tissue speci¬ ficity of the anti-cumulus matrix antiserum was checked by indirect immunofluorescence after the reaction with cryostat sections of human ovaries. There was no specific reaction with ovarian stromal and thecal cells. A slight reaction with mural granulosa cells was occasionally observed, but only with undiluted antiserum. In-vitro . Spermatozoa from normospermic donors with proven were used in all experiments. Sperm preparation and in-vitro insemination were performed using standard protocols (Tesafík et al, 1988b). Fertili¬ zation outcomes were evaluated 18-20 h after in-vitro insemination by inspection for pronuclei and polar bodies in an inverted microscope equipped with differential interference-contrast optics. An egg was considered penetrated when at least 2 pronuclei were visible in the egg cytoplasm. Eggs were then fixed and processed for electron microscopy using standard procedures (Tesafík et al, 1988b). Incubations with anti-cumulus matrix antisera. For the evaluation of the effects of the anti-cumulus matrix anti¬ bodies on fertilization and sperm-zona pellucida interaction, freshly obtained or briefly cultured (1-6 h) oocyte- cumulus complexes were preincubated for 90 min at 37°C in the presence of anti-cumulus matrix antiserum, followed by thorough washing of the oocyte-cumulus complexes in medium and in-vitro insemination. Each oocyte-cumulus complex was placed in 5 ml medium and incubated for 10 min at 37°C. This step was repeated four times using fresh medium to remove all unbound antibody and so to prevent any potential direct reaction of non-specific (in relation to the cumulus) antibodies with subsequently added spermatozoa. Other oocytes inseminated under the same conditions but preincubated with preimmune rabbit serum instead of the antiserum served as controls. For testing the effects of anti-cumulus matrix antisera on the cumulus matrix-induced acrosome reaction, samples of solubilized cumulus matrix were mixed with anti-cumulus matrix antisera and incubated at 37°C for 2 h. Aliquants of this mixture (100 pi) were then added to sperm suspensions (about 1 IO7 spermatozoa in 1 ml medium) preincu¬ bated in capacitating medium for 5 or 16 h. After an additional 1 h spermatozoa were prepared for quantitation of the acrosome reaction (see below). Sperm cultures in which either the anti-cumulus matrix antiserum was replaced with Downloaded from Bioscientifica.com at 09/29/2021 09:42:33PM via free access normal rabbit serum or in which the solubilized cumulus matrix was deleted served as the positive and negative controls, respectively. A third control incubation was performed with the antiserum to the mixture of hyaluronidase and proteinase inhibitor preparations (see above). All incubations were done in three replicates. Quantitation of sperm-zona pellucida interaction. The counts of total zona pellucida-associated spermatozoa and the percentages of surface-bound spermatozoa and of those having at least begun zona penetration were determined as previously described (Tesarik et al, 1988b). Briefly, total counts were made on living eggs when they were examined for the presence of pronuclei. Relative proportions of spermatozoa on the surface of, and within, the zona pellucida were evaluated later, by examination of ultrathin sections in the electron microscope. Quantitation of the acrosome reaction. Acrosome reactions were evaluated using the triple stain technique (Talbot & Chacon, 1981): 200 living spermatozoa (excluding trypan blue) were evaluated for each experimental protocol. Statistical analyses. Statistical significane of differences between data were evaluated by 2 tests.

Results

Effect ofanti-cumulus matrix antiserum on in-vitrofertilization All 15 oocytes preincubated with anti-cumulus matrix antiserum failed to be fertilized. By contrast, 10 out of 12 control oocytes preincubated with normal rabbit serum were penetrated and developed pronuclei ( < 0001). All oocytes of both groups showed a normal disaggregation of their cumuli oophori with complete cumulus matrix dispersal and only a few corona cells adhering to the zona pellucida surface. To determine the site of antibody action, the same experiment was performed with oocytes from which the zona pellucida had been previously removed. Such oocytes usually develop various degrees of upon incubation with capacitated spermatozoa (Tesarik et al, 1988b). In the absence of the zona pellucida, neither the proportion of penetrated oocytes nor the degree of polyspermy were noticeably influenced by the anti-cumulus matrix antibodies (P > 005). From 10 zona-free oocytes inseminated in the presence of anti-cumulus matrix antiserum, 8 became penetrated, as compared with 7 penetrated control oocytes out of 8 inseminated. The mean numbers of pronuclei per penetrated egg in the antibody-treated and control groups of oocytes were 5-4 and 5-7, respectively. The anti-fertilization effect of anti- cumulus matrix antibodies is therefore dependent on the presence of the zona pellucida.

Effect of anti-cumulus matrix antiserum on human sperm-zona pellucida interaction Prior incubation of human oocyte-cumulus complexes with anti-cumulus matrix antiserum markedly reduced the number of zona pellucida-associated spermatozoa (Table 1). Virtually all spermatozoa associated with the zona pellucida of the antiserum-pretreated oocytes were localized on the surface of the zona pellucida, while almost one-third of zona-associated spermatozoa in the control oocytes had at least begun zona penetration (Table 1). The proportion of spermatozoa within the zona pellucida of the control oocytes was similar to that previously observed for meta- phase II oocytes inseminated in vitro using a standard in-vitro fertilization protocol (Tesarik et al, 1988b).

Table 1. Effect of anti-cumulus matrix antiserum on sperm inter¬ action with the human zona pellucida (ZP)

ZP-associated spermatozoa

No. of % surface- % Treatment oocytes No./oocyte bound penetrated

Antiserum 15 3-3 + 2-8* 100 Control 12 58-3 ± 10-4 71 29

Values are means ± s.d. *P < 0001; compared with control. Downloaded from Bioscientifica.com at 09/29/2021 09:42:33PM via free access Effect of anti-cumulus matrix antiserum on cumulus matrix-induced acrosome reaction

When spermatozoa were preincubated in capacitating medium for 5 or 16 h and then incubated with the solubilized cumulus matrix, the percentages of acrosome-reacted spermatozoa (evaluated only for living sperm populations) were significantly (P < 001) increased above the values ascertained for sperm samples incubated in the absence of cumulus material (Fig. 1). This enhance¬ ment of the acrosome reaction frequency was similar to that previously described for capacitated human sperm samples incubated directly with whole cumuli oophori (Tesarik, 1985). The neutrali¬ zation of cumulus matrix antigens by preincubation of solubilized cumulus matrix with anti- cumulus matrix antisera significantly (P < 001) diminished the proportion of acrosome-reacted spermatozoa as compared with sperm samples incubated with solubilized cumulus matrix alone (Fig. 1). As also apparent from Fig. 1, the neutralization of cumulus matrix antigens by anti- cumulus matrix antiserum before the addition of this mixture to capacitated spermatozoa did not return the percentage of acrosome-reacted spermatozoa to the level of sperm samples incubated in the absence of cumulus materials and the difference between the latter two protocols still remained significant (P < 0-05). This observation indicates that the solubilized cumulus matrix contains some acrosome reaction-stimulating components that are not neutralized by anti-cumulus matrix antibodies. The presence of antiserum raised to a mixture of bovine testicular hyaluronidase and proteinase inhibitors, which were used for cumulus-matrix antigen preparation, did not show any quantitative effect on the cumulus matrix-induced acrosome reaction (Fig. 1), suggesting that the antagonistic effect of the anti-cumulus matrix antiserum on this process was not due to a non¬ specific action of antibodies to these contaminating components.

Fig. 1. Effect of anti-cumulus matrix antiserum on the cumulus matrix-induced acrosome reac¬ tion of human spermatozoa. After preincubation in capacitating medium, spermatozoa were treated for 1 h with normal rabbit serum alone ( ), with solubilized cumulus matrix and nor¬ mal rabbit serum ( D ), with solubilized cumulus matrix and anti-cumulus matrix antiserum ( D ) or with solubilized cumulus matrix and antiserum against the mixture of hyaluronidase and proteinase inhibitors as used for cumulus antigen solubilization (D). Downloaded from Bioscientifica.com at 09/29/2021 09:42:33PM via free access Discussion

The experiments described in this study have demonstrated that antiserum raised to the solubilized human cumulus oophorus intercellular matrix have the ability to block the fertilization process. While it is true that the antigen preparations used for immunization also contained some contami¬ nating components, particularly the commercial preparations of hyaluronidase and proteinase inhibitors used for antigen solubilization and possibly also incompletely washed blood plasma- derived components of follicular fluid, the experiments were constructed so as to minimize the potential non-specific effects of antibodies to these materials. After the thorough washing of freshly obtained cumuli oophori, traces of serum albumin were the only remaining serum proteins detec¬ table by h.p.l.c. (J. Tesarik, unpublished observation). Absorption of the antiserum with human serum was performed to remove antibodies to serum components. After the reaction with anti- serum, the cumuli were also washed very thoroughly before in-vitro insemination, in order to avoid any direct effect of unbound anti-hyaluronidase antibodies on spermatozoa. Finally, antiserum to a mixture of hyaluronidase and proteinase inhibitors did not show any effect on the frequency of cumulus matrix-induced acrosome reactions. Hence, antigens from the cumulus matrix are capable of inducing the production of antibodies interfering with fertilization and so they meet the principal requirement for a contraceptive vaccine target. The exact mechanism of the fertilization block caused by anti-cumulus matrix antibodies is not known. Surprisingly, the anti-fertilization effect of these antibodies seems to be due rather to impairment of sperm interaction with the zona pellucida than to defective penetration of the cumu¬ lus itself. It is possible that the relatively firm association of the cumulus-derived antigen with the zona pellucida, as demonstrated for cumulus cell-secreted glycoproteins (Tesarik & Kopecny, 1986), prevents the complete removal of these components by sperm action and, consequently, it also accounts for the retention of the anti-cumulus antibody in the zona pellucida. Two possible mechanisms of anti-cumulus antibody action can be proposed. First, the zona pellucida of late preovulatory and early post-ovulatory human oocytes is a dynamic system, into which both the oocyte and cumulus oophorus cells secrete newly synthesized materials (Tesarik & Kopecny, 1986). The materials secreted by the oocyte may have a hardening effect on the human zona pellucida, while those secreted by the cumulus cells apparently cause zona pellucida softening and an increase in its penetrability by homologous spermatozoa (Tesarik et al, 1988b). Anti-cumulus matrix anti¬ bodies may destabilize the equilibrium between these two opposite effects in favour of the oocyte- secreted hardening factors and so impede sperm penetration. Continuing secretion from the oocyte may be required for maintenance of the zona pellucida barrier function, since salt-stored zonae pellucidae surrounding dead human oocytes become penetrable even by heterologous spermatozoa (Yoshimatsu et al, 1988). Even though there are no comparative data concerning the penetrability by heterologous spermatozoa of zonae pellucidae surrounding living human oocytes, the zona pellucida has been shown to exert a block to heterologous sperm penetration in a number of mammalian species and there is no reason to suppose that the human species would be different. Second, anti-cumulus matrix antibodies may interfere with the appropriate timing of the conversion of proacrosin into the active form, acrosin. Components of the human cumulus matrix possessing a pronounced proacrosin-converting activity have been isolated and partly character¬ ized (J. Drahorád, unpublished work). Acrosin seems to be involved both in sperm-zona pellucida binding and zona penetration and the activation of this enzyme in the fertilizing spermatozoa must be exactly synchronized with the course of sperm-egg interaction (Tesarik et al, 1988c). The neu¬ tralization of cumulus-matrix components responsible for acrosin activation could therefore impair sperm-zona pellucida interaction at multiple sites of impact. We are at present investigating whether antibodies to cumulus matrix components do interfere with the proacrosin-converting activity of the cumulus matrix. Irrespective of the exact mechanism by which anti-cumulus matrix antibodies interfere with fertilization, the present results warrant further research into the cumulus-matrix antigens and Downloaded from Bioscientifica.com at 09/29/2021 09:42:33PM via free access their respective antibodies as potential tools for generating an efficient contraceptive vaccine with minimum side-effects. The suitability ofcumulus antigens for development of a human contraceptive vaccine will depend upon a careful evaluation of the effects of immunization on ovarian function in vivo as well as the demonstrated efficacy in an appropriate animal model.

I thank Dr L. Veselsky for preparation of anti-cumulus matrix antisera and Dr J. Testart for critically reading the manuscript.

References

Bamezai, A.K., Suman, ., Das, C. & Talwar, G.P. (1986) and inhibition of gonadotrophin-induced steroid Effects of immunization against pig zona pellucida secretion. Endocrinology 115, 2418-2432. (PZP) on steroid profiles and fertility in Talbot, P. & Chacon, R.S. (1981) A triple-stain technique primates. J. Reprod. Imuunol 9, Suppl. p. 85, abstr. evaluating normal acrosome reaction of human Carrón, C.P., Jarvis, H.W. & Saling, P.M. (1988) sperm. J. exp. Zool. 215, 201-208. Characterization of antibodies to determi¬ idiotypic Tesarik, J. (1985) Comparison of acrosome reaction- nants of monoclonal anti-sperm antibodies. Biol. inducing activities of human cumulus oophorus, folli¬ 1093-1103. Reprod. 38, cular fluid and ionophore A23187 in human sperm R.B.L. & L.C Gulyas, B.J., Gwatkin, Yuan, (1983) populations of proven fertilizing ability in vitro. J. Active immunization of Cynomolgus monkeys Reprod. Fert. 74, 383-388. with zonae (Macaca fascicularis) porcine pellucidae. Tesarik, J. & Kopecny, V. (1986) Late preovulatory syn¬ Res. 4, 299-307. thesis of proteoglycans by the human oocyte and Gwatkin, R.B.L., Andersen, O.F. & Williams, D.T. (1980) cumulus cells and their secretion into the oocyte- Immunization of female rabbits with heat-solubilized cumulus-complex intercellular matrices. Histo- bovine zonae: production of anti-zona antibody and chemistry 85, 523-528. inhibition of fertility. Gamete Res. 1, 19-26. Tesarik, J., Pilka, L., Drahorád, J., Cechova, D. & P. & Henderson, C.J., Braude, Aitken, R.J. (1987) L. The role of cumulus cell-secreted antibodies to 32-KDA Veselsky, (1988a) Polyclonal a deglycosylated proteins in the development ofhuman sperm fertilizing polypeptide from porcine zonae pellucidae will pre¬ in IVF. Hum. 3, 129-132. vent human interaction in vitro. Gamete Res. ability: implication Reprod. gamete Tesarik, L. & P. Zona 18,251-265. J., Pilka, Trávník, (1988b) pellu¬ cida resistance to before the com¬ Henderson, C.J., Hulme, M.J. & Aitken, R.J. sperm penetration (1988) pletion of human oocyte maturation. J. Reprod. Fert. Contraceptive potential of antibodies to the zona pel¬ 83,487^195. lucida. J. Fert. 83, 325-343. Reprod. Tesarik, J., Drahorád, J. & Peknicova, J. (1988c) Sub- Mori, T., Kamada, M., Yamano, S., Kinoshita, T., Kano, cellular acrosin in . & Mori, T. (1985) Production of monoclonal anti¬ immunochemical localization of human spermatozoa the acrosome reaction and bodies to porcine zona pellucida and their inhibition during zona Feri. Steril. 50, 133-141. of human zona pellucida penetration. sperm penetration through pellucida C. & in vitro. J. Immunol. 1-11. Wood, D.M., Liu, Dunbar, B.S. (1981) Effect of Reprod. 8, alloimmunization and Sacco, A.G. (1979) Inhibition of fertility in mice by heteroimmunization with immunization with antibodies to isolated zonae pellucidae on fertility in rabbits. Biol Reprod. passive 439^*50. zonae pellucidae. J. Reprod. Fert. 56, 533-537. 25, Shivers, CA. (1974) Immunological interference with Yoshimatsu, N., Yanagimachi, R. & Lopata, A. (1988) fertilization. Acta endocr., Copenh., Suppl. 194, Zonae pellucidae of salt-stored hamster and human 223-244. eggs: their penetrability by homologous and hetero¬ Skinner, S.M., Mills, T., Kirchick, H.J. & Dunbar, B.S. logous spermatozoa. Gamete Res. 21, 115-126. (1984) Immunization with zona pellucida proteins re¬ sults in an abnormal ovarian follicular differentiation Received 22 December 1988

Downloaded from Bioscientifica.com at 09/29/2021 09:42:33PM via free access