- Email: [email protected]
Use of a synthetic peptide adjuvant for the immunization of baboons with denatured and deglycosylated pig zona pellucida glycoproteins*
Vol. 52, No.2, August 1989
FERTILITY AND STERILITY
Printed in U.S.A.
Copyright <> 1989 The American Fertility Society
Use of a synthetic peptide adjuvant for the immunization of baboons with denatured and deglycosylated pig zona pellucida glycoproteins*
Bonnie S. Dunbar, Ph.D.t:j: Claire Lo, B.A. t John Powell, B.Sc§ Vernon C. Stevens, Ph.D.§ Baylor College of Medicine, Houston, Texas, and The Ohio State University Hospital, Columbus, Ohio
Baboons were immunized using a synthetic peptide adjuvant with two purified pig zona pellucida glycoproteins. The major zona pellucida glycoprotein (ZPI) was purified by preparative isoelectric focusing, and the 80 K deglycosylated zona pellucida protein (ZPIII) was purified by preparative sodium dodecyl sulfate polyacrylamide gel electrophoresis. The immunogenicity as well as the a;ntigenicity of these proteins were evaluated by characterizing antibodies using the enzyme-linked immunoassay and by immunoblotting of zona pellucida proteins separated by high-resolution two-dimensional polyacrylamide gel electrophoresis. Both groups of animals developed antibodies that recognize the major zona pellucida glycoprotein, (ZPI) and immunoblotting procedures provide evidence that two of the major porcine zona pellucida glycoprotein families (ZPI and ZPIII) contain shared antigenic determinants. The animals immunized with ZPI showed decreased levels of estrogen throughout their menstrual cycles, and two of the animals ceased ovulation. All animals in the group immunized with ZPIII had a significant reduction in the numbers of antral follicles as compared with control animals. Although ovarian cyclicity was not altered significantly within a few months after immunization, two of the five animals in this group became amenorrheic by 8 months. Histologic analysis of ovarian tissue shows that follicles were absent or frequently abnormal in animals of both groups following longterm immunization. These studies demonstrate that the synthetic adjuvant is effective in inducing antibodies (to purified zona pellucida glycoproteins) that recognize antigenic determinants to either denatured or deglycosylated zona pellucida glycoproteins, and that some of these antibodies may interfere with normal ovarian function. Fertil Steril 52: 311, 1989
It has been clearly established that active heteroimmunization of a variety of mammalian species with zona pellucida glycoproteins will cause the Received September 13, 1988; revised and accepted March 6, 1989. • This project was funded by contract no. 359 from Program Applied Research Fertility Regulation to B.S.D. and V.C.S. t Department of Cell Biology, Baylor College of Medicine. t Reprint requests: Bonnie S. Dunbar, Ph.D., Department of Cell Biology, Baylor College of Medicine, Houston, Texas 77030. § Department of Obstetrics and Gynecology, The Ohio State University Hospital. Vol. 52, No.2, August 1989
production of antibodies that inhibit fertilization by preventing the spermatozoa from binding to the surface of the zona pellucida in vitro. 1- 3 These studies have led to hopes that an effective i_m munologic contraceptive method could be developed using zona pellucida proteins as an immunogen. However, studies have been carried out that demonstrate that porcine zona pellucida proteins can result in altered cell types and ovarian function in the rabbit, bonnet monkeys, and dogs.2•4- 8 The studies carried out with rabbits by Skinner and colleagues2 have demonstrated that small primordial and primary ovarian follicles are damaged at the
Dunbar et al.
Immunization of baboons with ZP glycoproteins
311
stage at which the zona pellucida extracellular matrix is forming as the oocyte is growing and follicular cells are proliferating. Therefore, subsequent differentiation of steroid-producing granulosa cells does not differentiate in these animals. In studies by Sacco et al.,7 the major porcine zona pellucida glycoprotein was purified by chromatographic methods, and it was demonstrated that this form of the zona pellucida antigen also resulted in alteration of bonnet monkeys' ovarian function, in that fewer oocytes could be obtained by laparoscopy. These effects have more recently been reported to be reversible in squirrel monkeys.6 The use of monoclonal as well as polyclonal antibodies have shown that the antigenicity and immunogenicity of the mammalian zona pellucida is complex,9-13 and other biochemical studies have demonstrated that there are marked variations in the constituent glycoproteins of different mammalian zonae pellucidae, a factor that has made biochemical nomenclature of zona pellucida proteins complicated. 3 Multiple antigens are associated with the zona pellucida, and both shared and unique antigenic determinants are associated with the zonae pellucidae of different mammalian species. 9- 12 Further studies by Sacco et alY have shown that the carbohydrate moieties of the porcine zona pellucida can effect the immunogenicity and antigenicity of zonae pellucidae. If a practical and effective contraceptive method using a zona pellucida vaccine is to be developed for use in humans, it will most likely be through the use of proteins produced by recombinant deoxyribonucleic acid (DNA) technology or by synthetically produced peptides and synthetic adjuvants. Because these types of peptides will not contain the carbohydrate structures, nor the secondary or tertiary structures that are associated with the native zona pellucida matrix, it is important to evaluate the immunogenicity and antigenicity of the denatured and deglycosylated forms of these proteins. Furthermore, the use of acceptable adjuvants will be required for these purposes. One of these adjuvants (CGP-ll,637) has been reported to be effective in stimulating antibody production to both peptides as well as to human chorionic gonadotropin/ 4 and therefore provides an attractive candidate as an adjuvant for immunization with zona pellucida proteins. The present studies were therefore designed to use the denatured and deglycosylated forms of porcine zona pellucida proteins for the following reasons: 312
Dunbar et al.
1. To determine the immunogenicity of dena-
tured and deglycosylated zona pellucida proteins in nonhuman primate, the baboon; 2. To determine the efficacy of the use of a synthetic adjuvant in primates; 3. To characterize antigens using the established enzyme-linked immunoassay, and high-resolution two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) immunoblotting methods; and 4. To determine if purified zona pellucida proteins cause alteration in ovarian function, as do the native proteins, using a species that can be easily evaluated for ovarian cyclicity. MATERIALS AND METHODS Isolation of Zona Pellucida Proteins Used for Immunization
Porcine zonae pellucidae were isolated from frozen ovaries obtained from Health Products, Inc. (South Haven, MI), using methods described previously.s Zonae pellucidae were solubilized and deglycosylated using trifluoromethane sulfonic acid. 10 Preparative electrophoresis and electroelution methods were used to purify specific zona pellucida glycoproteins. The porcine zona pellucida glycoprotein terminology first described by Dunbar et al. 3 is used to describe the purified proteins used in this study. Sodium Dodecyl Sulfate (SDS)-P AGE and High-Resolution 2D-P AGE
One-dimensional electrophoresis of zona pellucida protein was carried out as previously described. 15,16 The methods used for high-resolution 2D-PAGE were those previously described. 15 ,16 Isoelectric focusing using 3.5 to 10 pH ampholines (LKB) (first-dimension electrophoresis) was carried out at 25°C for 16 hours at 750 V for a total of 12,000 volt-hours after solubilization at pH 9.5 (95°C, 10 minutes) in isoelectric focusing solubilization buffer containing 2% SDS, 2% /3-mercaptoethanol, 1 % cyclohexaminoethyl sulfonic acid, and 10% glycerol. Second-dimension gradient slab gels of 10% to 20% polyacrylamide (including bis-crosslinker) were prepared using the Pace Linear Gradient Maker (Isolabs, Akron, OH) and the MEGADALT gel casting system (Health Products, Rockford, IL); electrophoresis was carried out using the Anderson DALT electrophoresis tank (Electro-
Immunization of baboons with ZP glycoproteins
Fertility and Sterility
1 eluted as described above. Protein levels were estimated by Coomassie blue binding to proteins. Procedures for Immunization and Determination of Ovarian Function
Figure 1 (A) and (B), Silver-stained protein pattern of porcine zona pellucida glycoproteins analyzed by high-resolution ZD-PAGE. First-dimensional isoelectric focusing: relative pH 3.8( +) to pH 8( -). Second-dimension separation by 10% SDSPAGE. (A), Silver-stained pattern of total porcine zona pellucida demonstrating three major glycoprotein families I, II, and III. (B), Purified zona pellucida glycoprotein I used to immunize group I baboons. M = molecular weight markers from high to low = 93, 81, 57,40, 35, 29, 17, and 6 K. Arrow: purified pig zona pellucida glycoprotein I exhibiting charge and molecular weight heterogeneity. *: silver stain in mercaptan artifacts (Mr = 68 K and 50 K). (C), Coomassie blue-stained pattern of deglycosylated porcine zonae pellucidae proteins separated by one-dimensional SDS-PAGE (10% acrylamide). The arrow indicates the 80 K pig zona pellucida protein III, which was electroeluted and used for immunization of the group II baboons. (D) and (E), Immunoblot pattern of antibody binding to zona pellucida proteins separated by high-resolution ZD-PAGE. Following electrophoretic transfer of proteins to nitrocellulose and incubation with baboon antisera, the transfer was incubated with rabbit antihuman IgG followed by 1251_protein A. (D), Immunoblot pattern obtained by all animals of Group I. Brackets indicate ZPIII. (E), Immunoblot pattern obtained by all animals of group II.
nucleonics, Oak Ridge, TN). The silver-based color stain outlined by DunbarI5 was used to detect zona pellucida proteins. The total zona pellucida proteins separated using 2D-PAGE give a protein pattern illustrated in Figure lA. The acidic end of the major zona pellucida glycoprotein I (2 cm of 14 cm length gel) was electroeluted using the methods and apparatus given by Dunbar. I5 The purity of this protein sample was evaluated by high-resolution 2D-PAGE and silverstaining. The zona pellucida protein obtained using this procedure is illustrated in Figure lB. Preparative one-dimensional SDS-PAGE was used to purify the 80K band of the deglycosylated zona pellucida protein III (Fig. lC), which was electroVol. 52, No.2, August 1989
Ten adult female baboons exhibiting ovulatory menstrual cycles were used for these studies. Three control menstrual cycles were monitored for cycle length, sex skin changes, and ovulation evidence by progesterone (P) measurements. During the second of these three cycles, frequent serum samples were collected for estradiol (E 2) and P level assessmentI 5 to establish normal ovarian cyclicity. During menstrual cycle days 1 to 5 of the fourth cycle, five group I females were immunized with 0.5 mg of the major acidic zona pellucida glycoprotein, together with 0.25 mg of muramyl dipeptide (MDP), a synthetic adjuvant compound (CGP11637) (Ciba-Geigy, Ltd., Basel, Switzerland) by emulsifying these materials with squalene and injecting the emulsion intramuscularly at two sites in a total volume of 1.0 mL Two additional immunizations were given at 28-day intervals and a fourth injection was given at approximately 6 months after the initial immunization. Serum samples were collected at frequent intervals and assayed for antibody levels. Serum samples for E2 measurement were collected during the third cycle after the first immunization. Menstrual cycle parameters were monitored for nine cycles (or time equivalents in those with disrupted cycles) after significant antibody levels were detected in serum. At the end of this period, ovaries from all of the females were surgically removed, preserved in buffered formalin, and processed by paraffin embedding for histologic analysis using a periodic acid-Schiff stain. Ovaries were sectioned so that an entire cross section of the tissue could be examined. The numbers of developing follicles were counted in ten different cross sections taken at different regions of the ovary. Control ovaries were obtained from baboons of comparable age for control histologic analysis. Group II females were immunized with approximately 0.1 mg of deglycosylated 80 K porcine protein (as estimated by Coomassie blue binding to protein in PAGE gels following electroelution. This immunogen was emulsified with the synthetic adjuvant using the same procedure as described above). Less of this immunogen was administered in this experiment, because of limited availability of this protein as compared with the major zona pellucida protein (deglycosylated molecular weight
Dunbar et aI. Immunization of baboons with ZP glycoproteins
313
of 35 to 38 K). These animals were boosted at 3, 9, and 25 weeks using an identical preparation as the initial immunization. Hormone levels were assessed in control cycle 2, but postimmunization levels were not measured until cycles 7 to 8 from the first immunization. A total of four immunizations were given at 3-, 9-, and 25-week intervals. Nine menstrual cycles were followed after the last booster injection. At 49 weeks, the ovaries of the animals were surgically removed, preserved in buffered formalin, and processed for histologic analysis. Antibody Analysis by Enzyme-Linked Immunoassay
The enzyme-linked immunoassay described1o.12 was used for these studies except that antihuman immunoglobulin (lgG) was used as the second antibody. The sera were titered to determine the point of 50% of maximum binding. Antibody Analysis by Immunoblotting
Antibody analysis by immunoblotting was carried out as previously described2.9 •15 to identify antibodies to specific zona pellucida glycoproteins. Unstained, unfixed 2D-PAGE gels were placed on the cathode side of the nitrocellulose paper (BioRad, Richmond, CA) and transferred for 2.5 hours at 1.3 A using the E-C Electroblot Apparatus (E.Y. Laboratories, St. Petersburg, FL) transfer unit. The nitrocellulose paper then was blocked overnight in 10 mmoljl Tris(hydroxymethyl) aminomethane(Tris)-saline, pH 7.2, with 3% (wt/vol) instant Carnation milk and 0.01 % sodium azide, followed by washing with two changes of Trissaline-azide. Polyclonal serum (1 to 5 ml), or monoclonal antibody supernatant (containing 100 JLg/ml IgG), was diluted in 50 ml Tris-saline-azide (containing milk) and incubated with the nitrocellulose transfer at 25°C with shaking overnight, again followed by washing with two changes of Tris-salineazide. The nitrocellulose transfer then was incubated with iodine 125 Protein A (Amersham, Arlington Heights, IL) goat antihuman IgG, followed by a total of 106 cpm Protein A in 50 ml Tris-saline-azide (containing milk), added to each transfer and incubated overnight with shaking. Extensive washing in Tris-saline-azide was carried out before air drying and exposure to Kodak XAR-5 X-ray film (Eastman Kodak Co., Rochester, NY) for autoradiography. 314
Dunbar et al.
RESULTS Characterization of Antibodies by 2D-P AGE Immunoblotting
The 2D-PAGE immunoblot patterns demonstrate that the antibodies in the sera of individuals of both groups recognize all isoforms of the major zona pellucida glycoprotein family (Fig. ID, E). The antibodies in the sera of all group I animals were recognized by zona pellucida glycoproteins of both zona pellucida glycoprotein families I and III, and gave patterns as shown in Figure ID. The antibodies in the sera of all Group II animals also detected the major glycoprotein as shown in Figure IE. Evaluation of Antibodies by Enzyme-Linked Immunosorbent Assay
The group I baboons immunized with the major zona pellucida protein I and MDP adjuvant all developed significant titers by 9 weeks postimmunization (Table 1). At 9 weeks, one of these animals (Donna) had a titer that was lower than the other four in the group. By 30 weeks, the titers of four animals dropped to minimal levels (data not shown), but were again increased following a third boost. The antibody titers of the group II animals were considerably lower than the group I animals, in that the maximum titer was 8 X 103 as compared with 6 X 104 • As with the group I animals, the antibody levels of group II dropped significantly between booster immunizations. Effects of Immunization on Ovarian Function and Morphology Group I Animals
Data regarding cycle and ovulations in control and 10 week postimmunization cycles are summarized in Figure 2 and Table 1. Two of the five females exhibited few effects of treatment on cycle length or ovulation; however, the remaining three had at least one cycle of extended duration and/or anovulation. Two of these three became amenorrheic, with no evidence of ovulation for several months. Data showing levels of serum E2 levels in a control and a IO-week postimmunization cycle for each female are presented in Figure 2, and these data clearly demonstrate reduced E2 levels in all animals of group I following immunization, although P levels were not significantly effected by this procedure (data not shown). Progesterone lev-
Immunization of baboons with ZP glycoproteins
Fertility and Sterility
Table 1
Summary of Effects of Immunization of Baboons with Zona Pellucida Immunizations on Ovarian Morphology Serum dilutions giving 50% maximum binding Ovulation status a (weeks postimmunization)
Maximum antibody titer Group Ie
9 weeks
33 weeks
Zerna Winnie Ariel Xandra Donna
1.2 X 105 6.8 X 10' 6.8 X 10' 6.8 X 10' 8 X 103
1.2 X 105 6.8 X 10' 6.8 X 10' 6.8 X 10' 6.8 X 10'
Group II
9 weeks
28 weeks
Laurel Ukena Effy Whitney Ina
8 2 6 2 2
X 103 X 103 X 103 X 103 X 103
8X 103 8 X 103 1 X 103 4 X 103 4 X 103
Ovarian histologyb (Average no. of antral follicles per cross section)
1.5 ± 11.9 ± 1.2 ± 12.5 ± 0.86 ±
Ovulation Ovulation Short follicular phase No ovulation (30) (amenorrheic) No ovulation (33) (amenorrheic)
1
2 1
.5 1
Od 7.4 ± 0.2e 4.0 ± 1.3 e
Ovulation Ovulation Ovulation Ovulation Ovulation
8.4 ±2.4e 6.7 ± 2 e
Control Bibbs Nola Noba
17.8 ± 3 18 ± 2 ±3 13
Ovulation determined by P levels and menses in baboons. Average number of antral follicles in a minimum of ten cross sections of different regions of the ovary. Ovaries from group 1 and control were frozen during shipment. Therefore, only late stages of follicular development could be evaluated in detail.
d No primordial follicle and developing follicles, but granulosa cell "clusters" without oocytes, as shown in Figure 3. e Some primordial and early stage follicle granulosa cells in majority of antral follicles had pyknotic nuclei that suggested increased atresia in follicles as compared with control animals.
els were reduced in only one (Xandra), at one cycle, but returned to normal P. This female became anovulatory in later cycles.
data is summarized in Table 1. The animals in group II, who had low antibody titers, exhibited normal oocytes and follicular development, as illustrated in Figure 3A, although the numbers of antral follicles were reduced (Ina and Whitney), while those having the greatest antibody titers (Laurel, Ukena, and Effy) had numerous abnormalities. Typical morphology of these ovaries is illustrated in Figures 3B to D. No developing follicles were observed in ovarian sections from Laurel, but numerous follicle cell clusters typical of those found in zona pellucida immunized rabbits 2 were observed (Fig. 3B). The ovaries of Ukena and Effy contained numerous atretic follicles, which were characterized by irregular zonae pellucidae that were always separated from the inner layer of follicle cells. These abnormal granulosa cell "clusters" or atretic follicles were not observed in control baboon ovaries.
a
b
C
Group II Animals
Menstrual cycle data in 80 K deglycosylated zona pellucida protein (ZPIII) immunized animals are shown in Table 1. Fewer effects on cycle length and ovulation were found within 25 weeks of the immunization period in these females than those receiving major zona pellucida glycoprotein (ZPI) immunization. Two animals had two and one other had one anovulatory cycle, but no significant cycle length extensions were observed. While these alterations may have been related to immunizations, one or two anovulatory cycles during a period of 16 to 18 months is not uncommon in baboons receiving no treatment. Effects of Immunization on Ovarian Morphology
Histologic analysis of a minimum of ten cross sections (5 ~m) taken from different regions of each ovary were examined for follicle morphology. The numbers of antral follicles in an average cross section of ovary was calculated for each ovary. This Vol. 52, No.2, August 1989
DISCUSSION
These studies demonstrated that antibodies are developed by nonhuman primates immunized with denatured or deglycosylated zona pellucida pro-
Dunbar et at.
Immunization of baboons with ZP glycoproteins
315
240
Ariel o Pre-immunization
180
• Post-immunization
120
240
.
Donna
180
c 120
i
ID
g g 240
240
Winnie
Xendra
180 120
/
80
240
Zerna
14 21 28 7 Days 01 Menstrual Cycle
the protein purification and deglycosylation methods require disruption of the zona pellucida matrix and solubilization of the proteins; therefore, many conformational determinants that may be involved in immunogenicity as well as antigenicity may be destroyed. The methods of protein quantitation following staining and electroelution also can be inaccurate and give higher estimations of proteins. 19 Thirdly, the synthetic adjuvant used in the study instead of complete Freund's adjuvant may account for reduced titers. The antibodies from group I baboons injected with the ZPI immunogen recognize antigenic determinants on both proteins ZPI and ZPIII. (Note that, as described by Sacco et al.,20 porcine ZP1, ZP2 and ZP3 correspond to porcine zona pellucida glycoprotein families III, II, and I in this study.) For recent review of zona pellucida protein nomenclature, see Timmons and Dunbar. 3 Antibodies from animals immunized with the purified deglycosylated 80 K protein ZPIII also recognized determinants on protein I, but not ZPIII by immunoblot analysis. These results strongly suggest that zona pellucida protein families I and III may share some antigenic determinants. Because the procedure used to deglycosylate and purify the zona pellucida proteins is different from that used to separate the proteins for immunoblot analysis, it is possible that not all antigenic determinants are recognized by antibodies using this procedure, and the antigenic
35
Figure 2 Profile of serum E2 levels of group I baboons taken throughout the menstrual cycle. Preimmunization: levels determined prior to immunization. Postimmunization: levels determined at 9 to 10 weeks postimmunization.
teins using a synthetic peptide adjuvant. The use of baboons in these studies facilitated the evaluation of ovarian cyclicity because this primate exhibits regular menses in captivity, as compared with other nonhuman primates excluding apes. Furthermore, the use of adjuvants does not interfere with normal ovarian function in this speciesI7 or in other nonhuman primates. 18 The antibody titers obtained with the purified zona pellucida proteins were lower than those previously obtained with heat-solubilized porcine zonae pellucidae, which have been measured in previous animal studies using similar immunization methods. 2,9 This could be attributed to several factors. First, 316
Dunbar et al.
Figure 3 Histology of sections representative of morphology of ovaries obtained from baboons immunized with deglycosylated 80 K zona pellucida glycoproteins. (A), Section from ovary of baboon Whitney showing a normal secondary follicle. 0 = oocyte; n = oocyte nucleus; Z = zona pellucida; G = granulosa cell clusters. (B), Section from ovary of Laurel showing granulosa cell, which were extremely abundant in cell section clusters (arrows). (C and D), Section of ovaries from Ukena and Effy showing examples of prevalent degenerate oocytes surrounded by irregular zona pellucida and granulosa cells with pyknotic nuclei.
Immunization of baboons with ZP glycoproteins
Fertility and Sterility
determinant recognized by sera of group II is preferentially exposed and/or more abundant in ZPI than in ZPIII. It is possible that either purified ZPI and ZPIII could reciprocally contaminate the respective purified protein preparations. However, the fact that analysis of the purified protein samples prepared in this way shows no contamination, nor are antibodies developed that recognize ZPII in either group, would argue against this. Because ZPII migrates between proteins ZPI and ZPIII and is more abundant than ZPIII, it would be likely that equal or greater antibody levels would be generated against ZPII because it would provide a greater extent of contamination. These observations therefore would suggest that there are one or more antigenic determinants that are shared among the different zona pellucida glycoproteins. These findings are consistent with previous reports using monoclonal antibodies ,11 as well as recent studies in which it is apparent that different zona pellucida protein antigens expressed by cDNA clones share antigenic determinants. 2 These studies also demonstrate that nonhuman primates immunized with total purified and deglycosylated porcine zona pellucida proteins will develop antibodies to zona pellucida antigens that can alter ovarian structure and cyclicity. Group I baboons had higher titers of antibodies, which correlated with the onset of amenorrhea in two of the five animals. Group II baboons had lower antibody titers than group I and did not exhibit any visible changes in cyclicity, although the ovarian histology following longer term immunization suggests that the early development of primary follicles are being altered. The mechanisms by which this alteration is effecting the ovarian granulosa cell differentiation and, therefore, steroid production remains to be determined and longer term studies will be essential to pinpoint the precise stage at which follicular differentiation is altered. The antibodies that appear to be responsible for these are directed against determinants on ZPI, and the extent ofthis effect correlates with antibody titers in the group II animals. These studies further demonstrate the complexity of the zona pellucida macromolecules as immunogens and antigens and emphasizes the importance of determining which antigenic determinants of individual zona pellucida molecules may be targeted to elicit antibodies that will inhibit fertilization but that will not alter ovarian function, which is a prerequisite for a safe immunocontraceptive Vol. 52, No.2, August 1989
method. Because antibodies to the porcine protein (ZPI) appear to be responsible for the alteration in ovarian morphology and therefore may effect normal ovarian function, it is apparent that attention should be focused on dissecting individual antigenic determinants of the individual zonapellucida proteins. This can be carried out using the recombinant DNA technology or with the production of synthetic peptides. Zonae pellucidae proteins appear to be important molecules in the fertilization process, as they are susceptible to proteolysis by the sperm acrosomal enzyme, acrosin, and other proteases,22,23 as well as in sperm binding. It is apparent, therefore, that future studies need to be directed toward the use of synthetic peptides or recombinant DNA technology for the generation of sufficient quantities of purified proteins, as well as to select specific antigenic determinants of different zona pellucida proteins that will elicit antibodies that will effectively inhibit fertilization without altering ovarian follicular development for use as a safe, effective immunocontraceptive. It has now been possible to isolate cDNAs that express proteins of mouse zonae pellucidae and to use strategies to prepare cDNAs from rabbit zonae pellucidae, which express zona pellucida antigens that are shared among the zonae pellucidae of different mammalian species22 ; the potential use of recombinant DNA technology now makes it practical to carry out such studies.
Acknowledgments. We wish to thank Ms. Donna Bundman for her technical expertise in protein purification and antibody analysis as well as Ms. Suzanne Mascola for her secretarial assistance. REFERENCES 1. Aiken J: Contraceptive potential of antibodies to the zona pellucida. J Reprod FertiI83:325, 1988 2. Skinner SM, Mills T, Kirchick HJ, Dunbar BS: Immunization with zona pellucida proteins results in abnormal ovarian follicular differentiation and inhibition of gonadotropin-induced steroid secretion. Endocrinology 115:2,418, 1984 3. Timmons T, Dunbar BS: Antigens of the mammalian zona pellucida. In Perspectives in Immunoreproduction: Conception and Contraception, Edited by S Mathew, CM Fredericks. New York, Hemisphere Publishing Corp, 1988, p 242 4. Mahi-Brown CA, Yanagimachi R, Hoffman JC, Huang TTF Jr: Fertility control in the bitch by active immunization with porcine zonae pellucidae: use of different adjuvants and patterns of estradiol and progesterone levels in estrous cycles. BioI Reprod 32:761, 1985 5. Mahi-Brown CA, HuangTTF Jr, Yanagimachi R: Infertil-
Dunbar et al.
Immunization of baboons with ZP glycoproteins
317
6.
7.
8.
9.
10.
11.
12. 13.
ity in bitches induced by active immunization with porcine zonae pellucidae. J Exp ZooI222:89, 1982 Sacco AG, Pierce DL, Subramanian MG, Yurewicz EC, Dukelow WR: Ovaries remain functional in squirrel monkeys (Saimeri sciureus) immunized with porcine zona pellucida 55,000 macromolecule. BioI Reprod 36:481, 1987 Sacco AG, Subramanian MG, Yurewicz EC, DeMayo FJ, Dukelow WR: Heteroimmunization of squirrel monkey (Saimeri sciureus) with a purified porcine zona antigen (PPZA): immune response and biologic activity of antiserum. Fertil Steril 39:350, 1983 Hodgen GD, Tullner WW, Vaitukaitus JL, Ward DN: Specific radioimmunoassay of chorionic gonadotropin during implantation in rhesus monkey. J Clin Endocrinol Metab 39:457,1974 Wood D, Liu C, Dunbar BS: The effect of alloimmunization and heteroimmunization with zonae pellucidae on fertility in rabbits. BioI Reprod 25:439, 1981 Drell DW, Dunbar BS: Monoclonal antibodies to rabbit and pig zona pellucida distinguish species specific and shared antigenic determinants. BioI Reprod 30:445, 1984 Timmons TM, Maresh GA, Bundman DS, Dunbar BS: Use of specific monoclonal and polyclonal antibodies to define distinct antigens of porcine zona pellucida. BioI Reprod 36: 1,275,1987 Maresh GA, Dunbar BS: Antigenic comparison of fine species of mammalian zonae pellucidae. J Exp Zool 244:299, 1987 Sacco AG, Yurewicz EC, Subramanian MG: Carbohydrate influences the immunogenic and antigenic characteristics of the ZP3 macromolecule (Mr 55,000) of the pig zona pellucida. J Reprod Fertil 76:575, 1986
318
Dunbar et al.
14. Powell JE, Stevens VE: Simple radioimmunoassay of five unconjugated ovarian steroids in a single sample of serum or plasma. Clin Chem 19:210, 1973 15. Dunbar BS: Two-Dimensional Gel Electrophoresis and Immunological Techniques, New York, Plenum Press, 1987, 372pp 16. Dunbar BS, Liu C, Sammons DW: Identification of the three major proteins of porcine and rabbit zonae pellucidae by two-dimensional gel electrophoresis: comparison with follicular fluid, sera and ovarian cell proteins. BioI Reprod 24:1,111,1981 17. Stevens VC, Powell JE, Lee AC, Griffin D: Anti-fertility effects ofimmunization offemale baboons with C-terminal peptides of the tJ-subunit of human chorionic gonadotrophin. Fertil Steril 36:98, 1981 18. Thau RB, Sundaram K, Thornton YS, Seidman LS: Effects of immunization with the tJ-subunit of ovine luteinizing hormone on corpus luteum function in the rhesus monkey. Fertil Steril 31:200, 1979 19. Timmons T, Dunbar BS: Unpublished data 20. Sacco AG, Yurewicz EC, Subramanian MG, DeMayo FJ: Zona pellucida composition: species crossreactivity and contraceptive potential of antiserum to a purified pig zona antigen (PPZA). BioI Reprod 25:977,1981 21. Dunbar B, Timmons T: Unpublished data 22. Dunbar BS, Bundman DS: Evidence for a structural role of the major glycoprotein of the pig zona pellucida. J Reprod Fertil81:363, 1987 23. Dunbar BS, Dudkewicz A, Bundman DS: Proteolysis of specific porcine zona pellucida glycoproteins by boar acrosin. BioI Reprod 32:619, 1985
Immunization of baboons with ZP glycoproteins
Fertility and Sterility
FERTILITY AND STERILITY
Printed in U.S.A.
Copyright <> 1989 The American Fertility Society
Use of a synthetic peptide adjuvant for the immunization of baboons with denatured and deglycosylated pig zona pellucida glycoproteins*
Bonnie S. Dunbar, Ph.D.t:j: Claire Lo, B.A. t John Powell, B.Sc§ Vernon C. Stevens, Ph.D.§ Baylor College of Medicine, Houston, Texas, and The Ohio State University Hospital, Columbus, Ohio
Baboons were immunized using a synthetic peptide adjuvant with two purified pig zona pellucida glycoproteins. The major zona pellucida glycoprotein (ZPI) was purified by preparative isoelectric focusing, and the 80 K deglycosylated zona pellucida protein (ZPIII) was purified by preparative sodium dodecyl sulfate polyacrylamide gel electrophoresis. The immunogenicity as well as the a;ntigenicity of these proteins were evaluated by characterizing antibodies using the enzyme-linked immunoassay and by immunoblotting of zona pellucida proteins separated by high-resolution two-dimensional polyacrylamide gel electrophoresis. Both groups of animals developed antibodies that recognize the major zona pellucida glycoprotein, (ZPI) and immunoblotting procedures provide evidence that two of the major porcine zona pellucida glycoprotein families (ZPI and ZPIII) contain shared antigenic determinants. The animals immunized with ZPI showed decreased levels of estrogen throughout their menstrual cycles, and two of the animals ceased ovulation. All animals in the group immunized with ZPIII had a significant reduction in the numbers of antral follicles as compared with control animals. Although ovarian cyclicity was not altered significantly within a few months after immunization, two of the five animals in this group became amenorrheic by 8 months. Histologic analysis of ovarian tissue shows that follicles were absent or frequently abnormal in animals of both groups following longterm immunization. These studies demonstrate that the synthetic adjuvant is effective in inducing antibodies (to purified zona pellucida glycoproteins) that recognize antigenic determinants to either denatured or deglycosylated zona pellucida glycoproteins, and that some of these antibodies may interfere with normal ovarian function. Fertil Steril 52: 311, 1989
It has been clearly established that active heteroimmunization of a variety of mammalian species with zona pellucida glycoproteins will cause the Received September 13, 1988; revised and accepted March 6, 1989. • This project was funded by contract no. 359 from Program Applied Research Fertility Regulation to B.S.D. and V.C.S. t Department of Cell Biology, Baylor College of Medicine. t Reprint requests: Bonnie S. Dunbar, Ph.D., Department of Cell Biology, Baylor College of Medicine, Houston, Texas 77030. § Department of Obstetrics and Gynecology, The Ohio State University Hospital. Vol. 52, No.2, August 1989
production of antibodies that inhibit fertilization by preventing the spermatozoa from binding to the surface of the zona pellucida in vitro. 1- 3 These studies have led to hopes that an effective i_m munologic contraceptive method could be developed using zona pellucida proteins as an immunogen. However, studies have been carried out that demonstrate that porcine zona pellucida proteins can result in altered cell types and ovarian function in the rabbit, bonnet monkeys, and dogs.2•4- 8 The studies carried out with rabbits by Skinner and colleagues2 have demonstrated that small primordial and primary ovarian follicles are damaged at the
Dunbar et al.
Immunization of baboons with ZP glycoproteins
311
stage at which the zona pellucida extracellular matrix is forming as the oocyte is growing and follicular cells are proliferating. Therefore, subsequent differentiation of steroid-producing granulosa cells does not differentiate in these animals. In studies by Sacco et al.,7 the major porcine zona pellucida glycoprotein was purified by chromatographic methods, and it was demonstrated that this form of the zona pellucida antigen also resulted in alteration of bonnet monkeys' ovarian function, in that fewer oocytes could be obtained by laparoscopy. These effects have more recently been reported to be reversible in squirrel monkeys.6 The use of monoclonal as well as polyclonal antibodies have shown that the antigenicity and immunogenicity of the mammalian zona pellucida is complex,9-13 and other biochemical studies have demonstrated that there are marked variations in the constituent glycoproteins of different mammalian zonae pellucidae, a factor that has made biochemical nomenclature of zona pellucida proteins complicated. 3 Multiple antigens are associated with the zona pellucida, and both shared and unique antigenic determinants are associated with the zonae pellucidae of different mammalian species. 9- 12 Further studies by Sacco et alY have shown that the carbohydrate moieties of the porcine zona pellucida can effect the immunogenicity and antigenicity of zonae pellucidae. If a practical and effective contraceptive method using a zona pellucida vaccine is to be developed for use in humans, it will most likely be through the use of proteins produced by recombinant deoxyribonucleic acid (DNA) technology or by synthetically produced peptides and synthetic adjuvants. Because these types of peptides will not contain the carbohydrate structures, nor the secondary or tertiary structures that are associated with the native zona pellucida matrix, it is important to evaluate the immunogenicity and antigenicity of the denatured and deglycosylated forms of these proteins. Furthermore, the use of acceptable adjuvants will be required for these purposes. One of these adjuvants (CGP-ll,637) has been reported to be effective in stimulating antibody production to both peptides as well as to human chorionic gonadotropin/ 4 and therefore provides an attractive candidate as an adjuvant for immunization with zona pellucida proteins. The present studies were therefore designed to use the denatured and deglycosylated forms of porcine zona pellucida proteins for the following reasons: 312
Dunbar et al.
1. To determine the immunogenicity of dena-
tured and deglycosylated zona pellucida proteins in nonhuman primate, the baboon; 2. To determine the efficacy of the use of a synthetic adjuvant in primates; 3. To characterize antigens using the established enzyme-linked immunoassay, and high-resolution two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) immunoblotting methods; and 4. To determine if purified zona pellucida proteins cause alteration in ovarian function, as do the native proteins, using a species that can be easily evaluated for ovarian cyclicity. MATERIALS AND METHODS Isolation of Zona Pellucida Proteins Used for Immunization
Porcine zonae pellucidae were isolated from frozen ovaries obtained from Health Products, Inc. (South Haven, MI), using methods described previously.s Zonae pellucidae were solubilized and deglycosylated using trifluoromethane sulfonic acid. 10 Preparative electrophoresis and electroelution methods were used to purify specific zona pellucida glycoproteins. The porcine zona pellucida glycoprotein terminology first described by Dunbar et al. 3 is used to describe the purified proteins used in this study. Sodium Dodecyl Sulfate (SDS)-P AGE and High-Resolution 2D-P AGE
One-dimensional electrophoresis of zona pellucida protein was carried out as previously described. 15,16 The methods used for high-resolution 2D-PAGE were those previously described. 15 ,16 Isoelectric focusing using 3.5 to 10 pH ampholines (LKB) (first-dimension electrophoresis) was carried out at 25°C for 16 hours at 750 V for a total of 12,000 volt-hours after solubilization at pH 9.5 (95°C, 10 minutes) in isoelectric focusing solubilization buffer containing 2% SDS, 2% /3-mercaptoethanol, 1 % cyclohexaminoethyl sulfonic acid, and 10% glycerol. Second-dimension gradient slab gels of 10% to 20% polyacrylamide (including bis-crosslinker) were prepared using the Pace Linear Gradient Maker (Isolabs, Akron, OH) and the MEGADALT gel casting system (Health Products, Rockford, IL); electrophoresis was carried out using the Anderson DALT electrophoresis tank (Electro-
Immunization of baboons with ZP glycoproteins
Fertility and Sterility
1 eluted as described above. Protein levels were estimated by Coomassie blue binding to proteins. Procedures for Immunization and Determination of Ovarian Function
Figure 1 (A) and (B), Silver-stained protein pattern of porcine zona pellucida glycoproteins analyzed by high-resolution ZD-PAGE. First-dimensional isoelectric focusing: relative pH 3.8( +) to pH 8( -). Second-dimension separation by 10% SDSPAGE. (A), Silver-stained pattern of total porcine zona pellucida demonstrating three major glycoprotein families I, II, and III. (B), Purified zona pellucida glycoprotein I used to immunize group I baboons. M = molecular weight markers from high to low = 93, 81, 57,40, 35, 29, 17, and 6 K. Arrow: purified pig zona pellucida glycoprotein I exhibiting charge and molecular weight heterogeneity. *: silver stain in mercaptan artifacts (Mr = 68 K and 50 K). (C), Coomassie blue-stained pattern of deglycosylated porcine zonae pellucidae proteins separated by one-dimensional SDS-PAGE (10% acrylamide). The arrow indicates the 80 K pig zona pellucida protein III, which was electroeluted and used for immunization of the group II baboons. (D) and (E), Immunoblot pattern of antibody binding to zona pellucida proteins separated by high-resolution ZD-PAGE. Following electrophoretic transfer of proteins to nitrocellulose and incubation with baboon antisera, the transfer was incubated with rabbit antihuman IgG followed by 1251_protein A. (D), Immunoblot pattern obtained by all animals of Group I. Brackets indicate ZPIII. (E), Immunoblot pattern obtained by all animals of group II.
nucleonics, Oak Ridge, TN). The silver-based color stain outlined by DunbarI5 was used to detect zona pellucida proteins. The total zona pellucida proteins separated using 2D-PAGE give a protein pattern illustrated in Figure lA. The acidic end of the major zona pellucida glycoprotein I (2 cm of 14 cm length gel) was electroeluted using the methods and apparatus given by Dunbar. I5 The purity of this protein sample was evaluated by high-resolution 2D-PAGE and silverstaining. The zona pellucida protein obtained using this procedure is illustrated in Figure lB. Preparative one-dimensional SDS-PAGE was used to purify the 80K band of the deglycosylated zona pellucida protein III (Fig. lC), which was electroVol. 52, No.2, August 1989
Ten adult female baboons exhibiting ovulatory menstrual cycles were used for these studies. Three control menstrual cycles were monitored for cycle length, sex skin changes, and ovulation evidence by progesterone (P) measurements. During the second of these three cycles, frequent serum samples were collected for estradiol (E 2) and P level assessmentI 5 to establish normal ovarian cyclicity. During menstrual cycle days 1 to 5 of the fourth cycle, five group I females were immunized with 0.5 mg of the major acidic zona pellucida glycoprotein, together with 0.25 mg of muramyl dipeptide (MDP), a synthetic adjuvant compound (CGP11637) (Ciba-Geigy, Ltd., Basel, Switzerland) by emulsifying these materials with squalene and injecting the emulsion intramuscularly at two sites in a total volume of 1.0 mL Two additional immunizations were given at 28-day intervals and a fourth injection was given at approximately 6 months after the initial immunization. Serum samples were collected at frequent intervals and assayed for antibody levels. Serum samples for E2 measurement were collected during the third cycle after the first immunization. Menstrual cycle parameters were monitored for nine cycles (or time equivalents in those with disrupted cycles) after significant antibody levels were detected in serum. At the end of this period, ovaries from all of the females were surgically removed, preserved in buffered formalin, and processed by paraffin embedding for histologic analysis using a periodic acid-Schiff stain. Ovaries were sectioned so that an entire cross section of the tissue could be examined. The numbers of developing follicles were counted in ten different cross sections taken at different regions of the ovary. Control ovaries were obtained from baboons of comparable age for control histologic analysis. Group II females were immunized with approximately 0.1 mg of deglycosylated 80 K porcine protein (as estimated by Coomassie blue binding to protein in PAGE gels following electroelution. This immunogen was emulsified with the synthetic adjuvant using the same procedure as described above). Less of this immunogen was administered in this experiment, because of limited availability of this protein as compared with the major zona pellucida protein (deglycosylated molecular weight
Dunbar et aI. Immunization of baboons with ZP glycoproteins
313
of 35 to 38 K). These animals were boosted at 3, 9, and 25 weeks using an identical preparation as the initial immunization. Hormone levels were assessed in control cycle 2, but postimmunization levels were not measured until cycles 7 to 8 from the first immunization. A total of four immunizations were given at 3-, 9-, and 25-week intervals. Nine menstrual cycles were followed after the last booster injection. At 49 weeks, the ovaries of the animals were surgically removed, preserved in buffered formalin, and processed for histologic analysis. Antibody Analysis by Enzyme-Linked Immunoassay
The enzyme-linked immunoassay described1o.12 was used for these studies except that antihuman immunoglobulin (lgG) was used as the second antibody. The sera were titered to determine the point of 50% of maximum binding. Antibody Analysis by Immunoblotting
Antibody analysis by immunoblotting was carried out as previously described2.9 •15 to identify antibodies to specific zona pellucida glycoproteins. Unstained, unfixed 2D-PAGE gels were placed on the cathode side of the nitrocellulose paper (BioRad, Richmond, CA) and transferred for 2.5 hours at 1.3 A using the E-C Electroblot Apparatus (E.Y. Laboratories, St. Petersburg, FL) transfer unit. The nitrocellulose paper then was blocked overnight in 10 mmoljl Tris(hydroxymethyl) aminomethane(Tris)-saline, pH 7.2, with 3% (wt/vol) instant Carnation milk and 0.01 % sodium azide, followed by washing with two changes of Trissaline-azide. Polyclonal serum (1 to 5 ml), or monoclonal antibody supernatant (containing 100 JLg/ml IgG), was diluted in 50 ml Tris-saline-azide (containing milk) and incubated with the nitrocellulose transfer at 25°C with shaking overnight, again followed by washing with two changes of Tris-salineazide. The nitrocellulose transfer then was incubated with iodine 125 Protein A (Amersham, Arlington Heights, IL) goat antihuman IgG, followed by a total of 106 cpm Protein A in 50 ml Tris-saline-azide (containing milk), added to each transfer and incubated overnight with shaking. Extensive washing in Tris-saline-azide was carried out before air drying and exposure to Kodak XAR-5 X-ray film (Eastman Kodak Co., Rochester, NY) for autoradiography. 314
Dunbar et al.
RESULTS Characterization of Antibodies by 2D-P AGE Immunoblotting
The 2D-PAGE immunoblot patterns demonstrate that the antibodies in the sera of individuals of both groups recognize all isoforms of the major zona pellucida glycoprotein family (Fig. ID, E). The antibodies in the sera of all group I animals were recognized by zona pellucida glycoproteins of both zona pellucida glycoprotein families I and III, and gave patterns as shown in Figure ID. The antibodies in the sera of all Group II animals also detected the major glycoprotein as shown in Figure IE. Evaluation of Antibodies by Enzyme-Linked Immunosorbent Assay
The group I baboons immunized with the major zona pellucida protein I and MDP adjuvant all developed significant titers by 9 weeks postimmunization (Table 1). At 9 weeks, one of these animals (Donna) had a titer that was lower than the other four in the group. By 30 weeks, the titers of four animals dropped to minimal levels (data not shown), but were again increased following a third boost. The antibody titers of the group II animals were considerably lower than the group I animals, in that the maximum titer was 8 X 103 as compared with 6 X 104 • As with the group I animals, the antibody levels of group II dropped significantly between booster immunizations. Effects of Immunization on Ovarian Function and Morphology Group I Animals
Data regarding cycle and ovulations in control and 10 week postimmunization cycles are summarized in Figure 2 and Table 1. Two of the five females exhibited few effects of treatment on cycle length or ovulation; however, the remaining three had at least one cycle of extended duration and/or anovulation. Two of these three became amenorrheic, with no evidence of ovulation for several months. Data showing levels of serum E2 levels in a control and a IO-week postimmunization cycle for each female are presented in Figure 2, and these data clearly demonstrate reduced E2 levels in all animals of group I following immunization, although P levels were not significantly effected by this procedure (data not shown). Progesterone lev-
Immunization of baboons with ZP glycoproteins
Fertility and Sterility
Table 1
Summary of Effects of Immunization of Baboons with Zona Pellucida Immunizations on Ovarian Morphology Serum dilutions giving 50% maximum binding Ovulation status a (weeks postimmunization)
Maximum antibody titer Group Ie
9 weeks
33 weeks
Zerna Winnie Ariel Xandra Donna
1.2 X 105 6.8 X 10' 6.8 X 10' 6.8 X 10' 8 X 103
1.2 X 105 6.8 X 10' 6.8 X 10' 6.8 X 10' 6.8 X 10'
Group II
9 weeks
28 weeks
Laurel Ukena Effy Whitney Ina
8 2 6 2 2
X 103 X 103 X 103 X 103 X 103
8X 103 8 X 103 1 X 103 4 X 103 4 X 103
Ovarian histologyb (Average no. of antral follicles per cross section)
1.5 ± 11.9 ± 1.2 ± 12.5 ± 0.86 ±
Ovulation Ovulation Short follicular phase No ovulation (30) (amenorrheic) No ovulation (33) (amenorrheic)
1
2 1
.5 1
Od 7.4 ± 0.2e 4.0 ± 1.3 e
Ovulation Ovulation Ovulation Ovulation Ovulation
8.4 ±2.4e 6.7 ± 2 e
Control Bibbs Nola Noba
17.8 ± 3 18 ± 2 ±3 13
Ovulation determined by P levels and menses in baboons. Average number of antral follicles in a minimum of ten cross sections of different regions of the ovary. Ovaries from group 1 and control were frozen during shipment. Therefore, only late stages of follicular development could be evaluated in detail.
d No primordial follicle and developing follicles, but granulosa cell "clusters" without oocytes, as shown in Figure 3. e Some primordial and early stage follicle granulosa cells in majority of antral follicles had pyknotic nuclei that suggested increased atresia in follicles as compared with control animals.
els were reduced in only one (Xandra), at one cycle, but returned to normal P. This female became anovulatory in later cycles.
data is summarized in Table 1. The animals in group II, who had low antibody titers, exhibited normal oocytes and follicular development, as illustrated in Figure 3A, although the numbers of antral follicles were reduced (Ina and Whitney), while those having the greatest antibody titers (Laurel, Ukena, and Effy) had numerous abnormalities. Typical morphology of these ovaries is illustrated in Figures 3B to D. No developing follicles were observed in ovarian sections from Laurel, but numerous follicle cell clusters typical of those found in zona pellucida immunized rabbits 2 were observed (Fig. 3B). The ovaries of Ukena and Effy contained numerous atretic follicles, which were characterized by irregular zonae pellucidae that were always separated from the inner layer of follicle cells. These abnormal granulosa cell "clusters" or atretic follicles were not observed in control baboon ovaries.
a
b
C
Group II Animals
Menstrual cycle data in 80 K deglycosylated zona pellucida protein (ZPIII) immunized animals are shown in Table 1. Fewer effects on cycle length and ovulation were found within 25 weeks of the immunization period in these females than those receiving major zona pellucida glycoprotein (ZPI) immunization. Two animals had two and one other had one anovulatory cycle, but no significant cycle length extensions were observed. While these alterations may have been related to immunizations, one or two anovulatory cycles during a period of 16 to 18 months is not uncommon in baboons receiving no treatment. Effects of Immunization on Ovarian Morphology
Histologic analysis of a minimum of ten cross sections (5 ~m) taken from different regions of each ovary were examined for follicle morphology. The numbers of antral follicles in an average cross section of ovary was calculated for each ovary. This Vol. 52, No.2, August 1989
DISCUSSION
These studies demonstrated that antibodies are developed by nonhuman primates immunized with denatured or deglycosylated zona pellucida pro-
Dunbar et at.
Immunization of baboons with ZP glycoproteins
315
240
Ariel o Pre-immunization
180
• Post-immunization
120
240
.
Donna
180
c 120
i
ID
g g 240
240
Winnie
Xendra
180 120
/
80
240
Zerna
14 21 28 7 Days 01 Menstrual Cycle
the protein purification and deglycosylation methods require disruption of the zona pellucida matrix and solubilization of the proteins; therefore, many conformational determinants that may be involved in immunogenicity as well as antigenicity may be destroyed. The methods of protein quantitation following staining and electroelution also can be inaccurate and give higher estimations of proteins. 19 Thirdly, the synthetic adjuvant used in the study instead of complete Freund's adjuvant may account for reduced titers. The antibodies from group I baboons injected with the ZPI immunogen recognize antigenic determinants on both proteins ZPI and ZPIII. (Note that, as described by Sacco et al.,20 porcine ZP1, ZP2 and ZP3 correspond to porcine zona pellucida glycoprotein families III, II, and I in this study.) For recent review of zona pellucida protein nomenclature, see Timmons and Dunbar. 3 Antibodies from animals immunized with the purified deglycosylated 80 K protein ZPIII also recognized determinants on protein I, but not ZPIII by immunoblot analysis. These results strongly suggest that zona pellucida protein families I and III may share some antigenic determinants. Because the procedure used to deglycosylate and purify the zona pellucida proteins is different from that used to separate the proteins for immunoblot analysis, it is possible that not all antigenic determinants are recognized by antibodies using this procedure, and the antigenic
35
Figure 2 Profile of serum E2 levels of group I baboons taken throughout the menstrual cycle. Preimmunization: levels determined prior to immunization. Postimmunization: levels determined at 9 to 10 weeks postimmunization.
teins using a synthetic peptide adjuvant. The use of baboons in these studies facilitated the evaluation of ovarian cyclicity because this primate exhibits regular menses in captivity, as compared with other nonhuman primates excluding apes. Furthermore, the use of adjuvants does not interfere with normal ovarian function in this speciesI7 or in other nonhuman primates. 18 The antibody titers obtained with the purified zona pellucida proteins were lower than those previously obtained with heat-solubilized porcine zonae pellucidae, which have been measured in previous animal studies using similar immunization methods. 2,9 This could be attributed to several factors. First, 316
Dunbar et al.
Figure 3 Histology of sections representative of morphology of ovaries obtained from baboons immunized with deglycosylated 80 K zona pellucida glycoproteins. (A), Section from ovary of baboon Whitney showing a normal secondary follicle. 0 = oocyte; n = oocyte nucleus; Z = zona pellucida; G = granulosa cell clusters. (B), Section from ovary of Laurel showing granulosa cell, which were extremely abundant in cell section clusters (arrows). (C and D), Section of ovaries from Ukena and Effy showing examples of prevalent degenerate oocytes surrounded by irregular zona pellucida and granulosa cells with pyknotic nuclei.
Immunization of baboons with ZP glycoproteins
Fertility and Sterility
determinant recognized by sera of group II is preferentially exposed and/or more abundant in ZPI than in ZPIII. It is possible that either purified ZPI and ZPIII could reciprocally contaminate the respective purified protein preparations. However, the fact that analysis of the purified protein samples prepared in this way shows no contamination, nor are antibodies developed that recognize ZPII in either group, would argue against this. Because ZPII migrates between proteins ZPI and ZPIII and is more abundant than ZPIII, it would be likely that equal or greater antibody levels would be generated against ZPII because it would provide a greater extent of contamination. These observations therefore would suggest that there are one or more antigenic determinants that are shared among the different zona pellucida glycoproteins. These findings are consistent with previous reports using monoclonal antibodies ,11 as well as recent studies in which it is apparent that different zona pellucida protein antigens expressed by cDNA clones share antigenic determinants. 2 These studies also demonstrate that nonhuman primates immunized with total purified and deglycosylated porcine zona pellucida proteins will develop antibodies to zona pellucida antigens that can alter ovarian structure and cyclicity. Group I baboons had higher titers of antibodies, which correlated with the onset of amenorrhea in two of the five animals. Group II baboons had lower antibody titers than group I and did not exhibit any visible changes in cyclicity, although the ovarian histology following longer term immunization suggests that the early development of primary follicles are being altered. The mechanisms by which this alteration is effecting the ovarian granulosa cell differentiation and, therefore, steroid production remains to be determined and longer term studies will be essential to pinpoint the precise stage at which follicular differentiation is altered. The antibodies that appear to be responsible for these are directed against determinants on ZPI, and the extent ofthis effect correlates with antibody titers in the group II animals. These studies further demonstrate the complexity of the zona pellucida macromolecules as immunogens and antigens and emphasizes the importance of determining which antigenic determinants of individual zona pellucida molecules may be targeted to elicit antibodies that will inhibit fertilization but that will not alter ovarian function, which is a prerequisite for a safe immunocontraceptive Vol. 52, No.2, August 1989
method. Because antibodies to the porcine protein (ZPI) appear to be responsible for the alteration in ovarian morphology and therefore may effect normal ovarian function, it is apparent that attention should be focused on dissecting individual antigenic determinants of the individual zonapellucida proteins. This can be carried out using the recombinant DNA technology or with the production of synthetic peptides. Zonae pellucidae proteins appear to be important molecules in the fertilization process, as they are susceptible to proteolysis by the sperm acrosomal enzyme, acrosin, and other proteases,22,23 as well as in sperm binding. It is apparent, therefore, that future studies need to be directed toward the use of synthetic peptides or recombinant DNA technology for the generation of sufficient quantities of purified proteins, as well as to select specific antigenic determinants of different zona pellucida proteins that will elicit antibodies that will effectively inhibit fertilization without altering ovarian follicular development for use as a safe, effective immunocontraceptive. It has now been possible to isolate cDNAs that express proteins of mouse zonae pellucidae and to use strategies to prepare cDNAs from rabbit zonae pellucidae, which express zona pellucida antigens that are shared among the zonae pellucidae of different mammalian species22 ; the potential use of recombinant DNA technology now makes it practical to carry out such studies.
Acknowledgments. We wish to thank Ms. Donna Bundman for her technical expertise in protein purification and antibody analysis as well as Ms. Suzanne Mascola for her secretarial assistance. REFERENCES 1. Aiken J: Contraceptive potential of antibodies to the zona pellucida. J Reprod FertiI83:325, 1988 2. Skinner SM, Mills T, Kirchick HJ, Dunbar BS: Immunization with zona pellucida proteins results in abnormal ovarian follicular differentiation and inhibition of gonadotropin-induced steroid secretion. Endocrinology 115:2,418, 1984 3. Timmons T, Dunbar BS: Antigens of the mammalian zona pellucida. In Perspectives in Immunoreproduction: Conception and Contraception, Edited by S Mathew, CM Fredericks. New York, Hemisphere Publishing Corp, 1988, p 242 4. Mahi-Brown CA, Yanagimachi R, Hoffman JC, Huang TTF Jr: Fertility control in the bitch by active immunization with porcine zonae pellucidae: use of different adjuvants and patterns of estradiol and progesterone levels in estrous cycles. BioI Reprod 32:761, 1985 5. Mahi-Brown CA, HuangTTF Jr, Yanagimachi R: Infertil-
Dunbar et al.
Immunization of baboons with ZP glycoproteins
317
6.
7.
8.
9.
10.
11.
12. 13.
ity in bitches induced by active immunization with porcine zonae pellucidae. J Exp ZooI222:89, 1982 Sacco AG, Pierce DL, Subramanian MG, Yurewicz EC, Dukelow WR: Ovaries remain functional in squirrel monkeys (Saimeri sciureus) immunized with porcine zona pellucida 55,000 macromolecule. BioI Reprod 36:481, 1987 Sacco AG, Subramanian MG, Yurewicz EC, DeMayo FJ, Dukelow WR: Heteroimmunization of squirrel monkey (Saimeri sciureus) with a purified porcine zona antigen (PPZA): immune response and biologic activity of antiserum. Fertil Steril 39:350, 1983 Hodgen GD, Tullner WW, Vaitukaitus JL, Ward DN: Specific radioimmunoassay of chorionic gonadotropin during implantation in rhesus monkey. J Clin Endocrinol Metab 39:457,1974 Wood D, Liu C, Dunbar BS: The effect of alloimmunization and heteroimmunization with zonae pellucidae on fertility in rabbits. BioI Reprod 25:439, 1981 Drell DW, Dunbar BS: Monoclonal antibodies to rabbit and pig zona pellucida distinguish species specific and shared antigenic determinants. BioI Reprod 30:445, 1984 Timmons TM, Maresh GA, Bundman DS, Dunbar BS: Use of specific monoclonal and polyclonal antibodies to define distinct antigens of porcine zona pellucida. BioI Reprod 36: 1,275,1987 Maresh GA, Dunbar BS: Antigenic comparison of fine species of mammalian zonae pellucidae. J Exp Zool 244:299, 1987 Sacco AG, Yurewicz EC, Subramanian MG: Carbohydrate influences the immunogenic and antigenic characteristics of the ZP3 macromolecule (Mr 55,000) of the pig zona pellucida. J Reprod Fertil 76:575, 1986
318
Dunbar et al.
14. Powell JE, Stevens VE: Simple radioimmunoassay of five unconjugated ovarian steroids in a single sample of serum or plasma. Clin Chem 19:210, 1973 15. Dunbar BS: Two-Dimensional Gel Electrophoresis and Immunological Techniques, New York, Plenum Press, 1987, 372pp 16. Dunbar BS, Liu C, Sammons DW: Identification of the three major proteins of porcine and rabbit zonae pellucidae by two-dimensional gel electrophoresis: comparison with follicular fluid, sera and ovarian cell proteins. BioI Reprod 24:1,111,1981 17. Stevens VC, Powell JE, Lee AC, Griffin D: Anti-fertility effects ofimmunization offemale baboons with C-terminal peptides of the tJ-subunit of human chorionic gonadotrophin. Fertil Steril 36:98, 1981 18. Thau RB, Sundaram K, Thornton YS, Seidman LS: Effects of immunization with the tJ-subunit of ovine luteinizing hormone on corpus luteum function in the rhesus monkey. Fertil Steril 31:200, 1979 19. Timmons T, Dunbar BS: Unpublished data 20. Sacco AG, Yurewicz EC, Subramanian MG, DeMayo FJ: Zona pellucida composition: species crossreactivity and contraceptive potential of antiserum to a purified pig zona antigen (PPZA). BioI Reprod 25:977,1981 21. Dunbar B, Timmons T: Unpublished data 22. Dunbar BS, Bundman DS: Evidence for a structural role of the major glycoprotein of the pig zona pellucida. J Reprod Fertil81:363, 1987 23. Dunbar BS, Dudkewicz A, Bundman DS: Proteolysis of specific porcine zona pellucida glycoproteins by boar acrosin. BioI Reprod 32:619, 1985
Immunization of baboons with ZP glycoproteins
Fertility and Sterility