Biochemical and biological properties of urinary follicle-stimulating hormone (FSH) from the rhesus monkey (Macaca mulatta).код для вставкиСкачать
American Journal of Primatology 27:205-213 (1992) Biochemical and Biological Properties of Urinary Follicle-Stimulating Hormone (FSH) From the Rhesus Monkey (Macaca mulatta) ROBERT L. MATTERI, PRADEEP K. WARIKOO, AND BARRY D. BAVISTER Wisconsin Regional Primate Research Center, University of Wisconsin, Madison Follicle-stimulating hormone (FSH) is routinely used for the induction of superovulation in women. Homologous gonadotropin preparations that could be used for reproductive studies in macaques would have valuable research applications. Accordingly, we set out to characterize the physical and biological characteristics of urinary FSH (UFSH) in the ovariectomized rhesus monkey. In urine from 7 monkeys, concentrations of bioactive FSH ranged from 16 to 57 Fgll, relative to cynFSH-RP1 (NIDDK). UFSH was contrasted to pituitary FSH (PFSH) by non-reducing SDSpolyacrylamide gel electrophoresis (PAGE), native disc PAGE, and FPLC chromatofocusing. The apparent molecular weights of UFSH and PFSH are similar (approximately 35 kD); however, UFSH is more negatively charged and demonstrates a lower overall isoelectric (pl) range than PFSH. The bioactivity of UFSH was assessed by the stimulation of aromatase activity in cultured Sertoli cells and by induction of follicular maturation in hamsters. Two fractions of pituitary FSH, which differed in isoelectric properties, were obtained by chromatofocusing. The in vivo biological activity of FSH-A (acidic, pl3.8-4.6) and UFSH (pl3.5-4.5) were similar, but greater than FSH-B (basic, pl4.6-5.5). These results support the hypothesis that heavily sialylated, low pl FSH expresses high in vivo bioactivity. This may reflect the well-known effect of sialic acid in prolonging the circulating half-life of glycoproteins. Thus, the quality and quantity of FSH present in ovariectomized rhesus monkey urine indicates that this may be a useful source for the preparation of enriched hormone preparations. o 1992 Wiley-Liss, Inc. Key words: superovulation, macaque, bioactivity INTRODUCTION There is a need for macaque follicle-stimulating hormone (FSH) that can be used in studies in those species which require long-term or repetitive hormone treatment. In macaques, the use of FSH from other species is known to induce the formation of antibodies which neutralize the hormone [Platia et al., 1984;Bavister h i v e d for publication April 8, 1991;revision accepted June 11, 1991. Address reprint requests to Dr. Robert L. Matteri, Wisconsin Regional Primate Research Center, University of Wisconsin, 1223 Capitol Court, Madison, WI 53715-1299. 0 1992 Wiley-Liss, Inc. 206 I Matteri et al. et al., 1986; Rotten et al., 1986; Wolf et al., 19901. Although 2 to 3 exogenous stimulation procedures may be carried out in single monkeys, animals become refractory after this point [Wolf et al., 19901. Therefore, there are serious concerns in performing studies that require the repeated administration of heterologous FSH for the induction of superovulation or the study of ovarian physiology in macaques. Such studies would require a prohibitive number of animals to attain statistically relevant results, produce results that could be influenced by immunological neutralization of hormone treatment, and have unknown effects on the future reproductive status of valuable animals following experimentation. In order to circumvent this problem, it is necessary t o identify and evaluate potential sources of FSH. Based upon such information, hormone enrichment/ purification protocols can subsequently be developed. Since urinary gonadotropins are routinely used for superovulation of women prior to artificial insemination or in vitro fertilization, we were interested in evaluating urine from the rhesus monkey as a potential source of these hormones. As a first step in this process, we have characterized the biochemical and biological properties of urinary FSH (UFSH) in the ovariectomized rhesus monkey. METHODS Animals Seven ovariectomized rhesus monkeys were housed at the Wisconsin Regional Primate Research Center (WRPRC). The animals were kept a t 25-28°C under controlled lighting conditions (14L:10D),and received standard pellet chow twice daily with fruit supplement and tap water ad libitum. Overnight urine samples were collected on ice from animals housed in metabolic cages. Concentration of Urinary Proteins Urinary protein was precipitated by the addition of solid ammonium sulfate until a 75% saturated solution was achieved. This solution was kept at 4°C overnight. The precipitate was collected by centrifugation, redissolved in 0.05 M ammonium bicarbonate, desalted by Sephadex G-25 chromatography, and lyophilized. A single monkey (rhesus 1662) was selected on the basis of having the highest level of excreted bioactive FSH. One liter of urine was collected from this animal and the protein harvested as above. This preparation is designated as UFSH. A second concentration (UFSH-PoolC) was prepared from 1 liter of pooled urine from 5 animals. Biochemical Analyses Pituitary and urinary proteins were analyzed by FPLC chromatofocusing [Matteri and Papkoff, 19881. A Mono-P column (Mono-P HR 5/20, Pharmacia) was used in conjunction with an ISCO model 2350 HPLC pump (0.75 ml/min) and model 2360 gradient programmer. One milligram of a pituitary protein extract [Matteri and Ziegler, in press] was dissolved in 1 ml of 0.025 M ethanolamine, pH 9.5. The column was equilibrated with 0.025 M ethanolamine. The sample was injected and 5 frictions were collected prior to beginning elution with 1:12 polybuffer 96 (Pharmacia), pH 5.5. After 20 min, the elution buffer was changed to a 1:lO solution of polybuffer 74 (Pharmacia), pH 3.5. Both polybuffer solutions contained 0.1% glycerol. When the pH gradient had stabilized, 1 ml 2 M sodium acetate was injected to remove all material still adherent to the column (apparent pl < 3.5). Fractions were collected into 12 X 75 glass tubes containing 0.75 ml PBS-0.5% BSA, pH 7.5. The FSH content of each fraction was determined by RIA (see below). Two fractions corresponding to the pl ranges of 4.6-5.5 (FSH-B) and Urinary FSH in the Rhesus Monkey I 207 3.8-4.6 (FSH-A) were pooled, dialyzed, and lyophilized for subsequent biological comparison to UFSH. Electrophoretic evaluations of molecular weight, relative electric charge, and isoelectric properties were performed in polyacrylamide gels, followed by immunological detection of FSH [Matteri & Ziegler, 19921. Hormone Assays In vitro FSH bioactivity was assessed as described previously [Matteri et al., 19901. Reference preparation cynFSH-RP1 was obtained from the National Hormone and Pituitary Program (NHPP), NIDDK. Within- and between-assay variabilities were 5.6 and 8.1%, respectively. The FSH radioimmunoassay (RIA) has been described previously [Hodgen et al., 19761, with reagents provided by the NHPP. As with the bioassay, cynFSHRP1 was used as the reference preparation. Analysis of assay data was performed by data reduction programs on the WRPRC VAX computer. Within- and betweenassay coefficients of variation were 4.2 and 9.6, respectively. Iodinations were performed by the iodogen method [Matteri et al., 19871. Stimulation of Follicular Development in Hamsters In vivo FSH biopotency was determined in hamsters, using an endpoint of the number of mature follicles produced [Bavister et al., 1986; Bavister, 19891. In short, each female hamster weighing between 90 and 120 gm was injected intraperitoneally on the morning of post-estrous discharge (day 1) with a single injection of FSH. Doses of FSH, as determined by the in vitro bioassay, ranged from approximately 4 to 700 ng (see Fig. 5, Results). At least 2 hamsters were used at each dose level. Fifty-three to 55 h after hormone injection (day 3), animals were sacrificed by cervical dislocation. Their ovaries were removed and the number of stimulated follicles was counted. Statistics In vivo bioactivity was plotted vs. in vitro potency for each FSH preparation. Differences in the slopes of the regression lines, which represent in vivo vs. in vitro biopotency, were detected by t-test [Glantz, 19811. RESULTS In the in vitro FSH bioassay, 1 vial (75 I.U.) of Pergonal was found to be as potent as 20.3 rfr 3.1 pg cynFSH-RP1 (n = 3 assays). In order to determine the quantity of FSH present in rhesus monkey urine, samples were obtained from seven ovariectomized monkeys and analyzed for in vitro FSH bioactivity. Relative to cynFSH-RP1 (NHPP, NIDDK), bioactive FSH concentrations ranged from 16 to 57 pg/l. In contrast, average bioactive FSH levels in urine from the follicular and luteal phases of the rhesus monkey are less than 5 pg/l [Matteri et al., 19901. Figure 1 demonstrates the FSH bioactivity present in the urine of three of the animals. The isoelectric properties of UFSH and PFSH were determined by chromatofocusing (Fig. 2). The UFSH demonstrated a more acidic isoelectric profile relative to pituitary FSH. The hormone which remained bound to the column a t a pl of 3.5 was eluted with 2 M sodium acetate. This amounted to 3.4 and 8.7% of the total hormone recovered from the chromatofocusing of PFSH and UFSH, respectively. As predicted from their relative isoelectric characteristics, UFSH possessed more material with high mobility than PFSH when analyzed by native disc gel electrophoresis (Fig. 3). Similar data were obtained with UFSH-Pool C. As in the 208 I Matted et al. 10000 - cynFSH-RPI --%Y--- 1662 I u-91 1 AF-02 (u W m n 100 : .01 = - -......1 - I ' '.""I 1 - - '...'I 10 ' ..-I 100 ng or ~IllWell Fig. 1. Bioactive FSH concentrations in urine samples from 3 of the 7 ovariectomized monkeys utilized in the present study. The reference preparation utilized was cynFSH-RP1 (NHPP, NIDDK estimated to be 15-20% pure). Animal identification codes are indicated as U-91,AF-02,and 1662. 100 - 1 I PFSH UFSH Y 0 Q) P 3 4 5 7 6 8 9 10 PI Fig. 2. Isoelectric resolution of immunoreactive urinary and pituitary FSH by FPLC chromatofocusing over a pH gradient of 9.5to 3.5.Fractions between the pH values of 4.6 and 5.5 and 3.5 and 4.6were pooled to create a basic and an acidic preparation of FSH (FSH-B and FSH-A, respectively). chromatofocusing data, similar forms were present in both samples. However, the relative proportions of these forms varied. Differences in isoelectric characteristics between hormone preparations were not reflected in large differences in molecular weight (Fig. 4). The FSH present in urine appears to be fully bioactive in vitro, since the hormone content of UFSH was determined as 0.248 and 0.246 pglmg by in vitro - Urinary FSH in the Rhesus Monkey / 209 100 80 V UFSH UFSH-PooIC 60 40 20 0 0.0 0.2 0.4 0.6 1 .o 0.8 Rf Fig. 3. Native disc gel electrophoresis of pituitary and urinary FSH. While the overall charge properties vary, similar forms appear in each preparation. Mobility is presented as migration relative to the dye front (Rf). - 100 80 I_f_ FSH UFSH-PooIC 60 40 20 0 0.0 0.2 0.4 0.6 0.8 1 .o Rf Fig. 4. Non-reducingSDS-PAGE [Matteri et al., 19901 of pituitary [PFSH] and urinary (UFSH, UFSH-Pool C) FSH from the rhesus monkey. The FSH in all preparations was similar (approximately 35 M)). Mobility is presented as migration relative to the dye front (FW. bioassay and immunoassay, respectively (relative to cynFSH-RP1 reference). Considerable in vivo biopotency was also observed (Fig. 5). The relative in vivo vs. in vitro activities of UFSH and UFSH-Pool C were similar to FSH-A, but greater than FSH-B (P< 0.05;Fig. 5). The isoelectric properties of FSH-A and FSH-B were evaluated by IEF-PAGE (Fig. 6 ) and found to correspond well to the chromatofocusing fractions from which they were derived. 210 I Matteri et al. - 60 50 - FSH-A FSH-B UFSH Pool'C' 40 TT/ 30 20 1 10 100 1000 ng FSH (In vltro potency rel. to cynFSH-RP1) Fig. 5. The number of mature follicles resulting from the injection of various doses of urinary FSH (UFSH, UFSH-Pool C), FSH-A (acidic PFSH), and FSH-B (basic PFSH) are plotted vs. in vitro biopotencies. The slopes of the resulting regression lines were contrasted in order to assess relative in vivo and in vitro bioactivities. DISCUSSION Urinary FSH is produced commercially for the clinical induction of superovulation in women. Comparable gonadotropin preparations do not exist for macaques. The ability to stimulate ovarian function repeatedly in these species would permit the development of models for the study of human fertility and of propagation of endangered nonhuman primates. Accordingly, the goal of the present study was to evaluate the quantity, biochemical characteristics, and biological potency of urinary FSH in the rhesus monkey. The ovariectomized rhesus monkey appears to excrete FSH in quantities which could be utilized for ovarian stimulation. In the course of identifying animals with high levels of FSH excretion, we have found that up to 57 pg FSH/1 may be present (relative to cynFSH-RP1). These levels significantly exceed those found in the follicular and luteal phases of the rhesus monkey [Matteri et al., 19901. We have determined 20 pg to be equivalent to one vial of 75 IU Pergonal (Serono Labs) in the in vitro bioassay. It is known that the administration of 540 IU of Pergonal over a 9 day period will induce follicular maturation in the rhesus monkey [Bavister, 19881. Depending on the animal used, this is the equivalent of approximately 3-6 1of urine with high FSH concentrations. While rhesus urinary FSH is highly active in hamsters (Fig. 5), subsequent studies are needed to confirm ovarian stimulation in macaques. The present results indicate that UFSH possesses a high apparent negative charge and low isoelectric profile. There is little doubt that this reflects an abun- Urinary FSH in the Rhesus Monkey / 211 - 100 . I FSH-B 80 FSH-A 60 40 20 0 3 4 5 6 7 PI Fig. 6. Isoelectric focusing (IEF) PAGE of FSH-A and FSH-B. Isoelectric characteristics of both preparations correspond with the chromatofocusing fractions from which they were derived (Fig. 2). dance of isoforms/isohormonesof FSH which contain a relatively large amount of sialic acid [Chappel et al., 1983; Matteri & Papkoff, 19881. Isoforms of FSH with high isoelectric points (pl) are known to be highly bioactive in terms of in vitro biopotency and/or receptor binding activity [Chappel et al., 19831. While high pl FSH isoforms are likely to possess a high degree of activity a t the target cell, these forms are relatively deficient in sialic acid [Blum et al., 19851. Since sialic acid prolongs the circulating half-life of glycoproteins [Morell et al., 1971; Ashwell & Morell, 19741, it is not surprising that heavily sialylated, low pl forms of FSH are metabolized relatively slowly [Blum & Gupta, 1985; Galway et al., 19901.This may translate into a high degree of bioactivity in vivo which would not necessarily be observed in vitro. Our present data, and those of Galway et al. [19901, support this hypothesis (Fig. 4). The biological quality of gonadotropins, as related to isohormone profile, is influenced by a variety of physiological conditions. The qualitative properties of secreted and/or intra-pituitary gonadotropins are influenced by GnRH [Dufau et al., 1976; Resko et al., 19821, sex steroids [Peckham & Knobil, 1976a; Marut et al., 1981; Chappel et al., 19841, gonadectomy [Peckham & Knobil, 1976a,bl, stage of sexual development [Reader et al., 1983; Stroll0 et al., 1981; Wide & Hobson, 19831, and phase of the reproductive cycle [Adams et al., 1987; Cameron & Chappel, 1985; Robertson et al., 19791. It is expected, therefore, that the properties of excreted gonadotropins also will be influenced by individual differences in endocrine status. The removal of steroid feedback produced by ovariectomy increases the amount of gonadotropin secretion and the relative abundance of highly sialylated isohormones [Khan et al., 19851.In terms of quantity and quality, urine from the ovariectomized rhesus monkey may offer a viable source of highly potent FSH. Such gonadotropin preparations would have immediate research applications and would serve as standards for the future production of recombinant FSH. CONCLUSIONS 1. Ovariectomized rhesus monkeys secrete relatively high levels of bioactive FSH. 212 I Matteri et al. 2. Urinary FSH from the ovariectomized monkey is highly charged and possesses a low pl, which is likely due to a relatively high sialic acid content. 3. The chemical and biological characteristics of urinary FSH produce further evidence to support the concept that natural or recombinant FSH preparations to be used for superovulation should be selected for a sialic acid content. ACKNOWLEDGMENTS We are grateful to Elaine Anderson for preparing this manuscript for publication. We wish to thank Frederick Wegner, Guenther Scheffler, and Daniel Wittwer for their expert technical assistance. We also thank NIDDK and the National Hormone and Pituitary Program (University of Maryland School of Medicine) for their gifts of hormones and reagents for assay purposes. This is Wisconsin Regional Primate Research Center publication 31-002. This work was supported by NIH grant RR00167 to the Wisconsin Regional Primate Research Center. This study was conducted in accordance with the Guide for the Care and Use of Laboratory Animals, NIH publication 85-23 and Public law 89-544, “The Animal Welfare Act,” August 24, 1966, and its amendments. REFERENCES Adams, T.E.; Adams, B.M.; Watson, J.G. Secretory dynamics of bioactive and immunoactive LH during the oestrous cycle of the sheep. JOURNAL OF REPRODUCTION AND FERTILITY 79555463,1987. Ashwell, G.; Morell, A.G. The role of carbohydrates in the hepatic recognition and transport of circulating glycoproteins. ADVANCES IN ENZYMOLOGY AND RELATED AREAS OF MOLECULAR BIOLOGY 41:99-128,1974. Bavister, B.D. Oocyte maturation and in vitro fertilization in the rhesus monkey. Pp. 119-137 in THE PRIMATE OVARY. R.L. Stouffer, ed. New York, Plenum Press, 1988. Bavister, B.D. A consistently successful procedure for in vitro fertilization of golden hamster eggs. GAMETE RESEARCH 23: 139-158,1989. Bavister, B.D.; Dees, H.C.; Schultz, R.D. Refractoriness of rhesus monkeys to repeated gonadotropin stimulation is due to formation of non-precipitating antibodies. AMERICAN JOURNAL OF REPRODUCTIVE IMMUNOLOGY AND MICROBIOLOGY 11:ll-16, 1986. Blum, W.F.P.; Gupta, D. Heterogeneity of rat FSH by chromatofocusing: Studies on serum FSH, hormone released in vitro and metabolic clearance rates of its various forms. JOURNAL OF ENDOCRINOLOGY 105~29-37,1985. Blum, W.F.P.; Riegelbauer, G.; Gupta, D. Heterogeneity of rat FSH by chromatofocusing: Studies on in vitro bioactivity of pituitary FSH forms and effect of neuraminidase treatment. JOURNAL OF ENDOCRINOLOGY 105:17-27,1985. Cameron, J.L.; Chappel, S.C. Follicle-stimu- lating hormone within and secreted from anterior pituitaries of female golden hamsters during the estrous cycle and after ovariectomy. BIOLOGY OF REPRODUCTION 33:132-139, 1985. Chappel, S.C.; Bethea, C.L.; Spies, H.G. Existence of multiple forms of follicle stimulating hormone within the anterior pituitaries of cynomolgus monkeys. ENDOCRINOLOGY 115452-461,1984. Chappel, S.C.: Ulloa-Aguirre, A.; Coutifaris, C. Biosynthesis and secretion of folliclestimulating hormone. ENDOCRINE REVIEWS 4179-211,1983. Dufau, M.L.; Beitins, I.Z.; McArthur, J.W.; Catt, K.H. Effects of luteinizing hormone releasing hormone (LHRH) upon bioactive and immunoactive serum LH levels in normal subjects. JOURNAL OF CLINICAL ENDOCRINOLOGY AND METABOLISM 43:658-667, 1976. Galway, A.B.; Hsueh, A.J.W.; Keene, J.L.; Yamoto, M.; Fauser, B.C.J.M.; Boime, I. In vitro and in vivo bioactivity of recombinant human follicle-stimulating hormone and partially deglycosylated variants secreted from transfected eukaryotic cell lines. ENDOCRINOLOGY 127:93-100,1990. Glantz, S.A. How to test for trends. Pp. 180228 in PRIMER OF BIOSTATISTICS. S.A. Glantz, ed. San Francisco, McGraw-Hill Book Company, 1981. Hodgen, G.D.; Wilks, J.W.; Vaitukaitis, J.L.; Chen, H.C.; Papkoff, H.; Ross, G.T. A new radioimmunoassay for follicle stimulating hormone in macaques: Ovulatory menstrual cycles. ENDOCRINOLOGY 99:137145, 1976. Khan, S.A.; Syed, V.; Froysa, B.; Lindberg, M.; Diczfalusy, E. Influence of gonadec- Urinary FSH in the Rhesus Monkey / 213 tomy on isoelectrofocusing profiles of pituitary gonadotropins in rhesus monkeys. JOURNAL OF MEDICAL PRIMATOLOGY 14:177-194,1985. Marut, E.L.; Williams, R.F.; Cowan, B.D.; Lynch, A.; Lerner, S.; Hodgen, G.D. Pulsatile pituitary gonadotropin secretion during maturation of the dominant follicle in monkeys: Estrogen positive feedback enhances the biological activity of LH. ENDOCRINOLOGY 109~2270-2272, 1981. Matteri, R.L.; Papkoff, H. Microheterogeneity of equine follicle-stimulating hormone. BIOLOGY OF REPRODUCTION 38324331,1988. Matteri, R.L.; Durning, M.; Dierschke, D.J.; Handrow, R.R. Determination of bioactive FSH in rhesus monkeys (Mucucu muluttu). AMERICAN JOURNAL OF PRIMATOLOGY 21~295-305,1990. Matteri, R.L.; Roser, J.F.; Baldwin, D.M.; Lipovetsky, V.; Papkoff, H. Characterization of a monoclonal antibody which detects luteinizing hormone from diverse mammalian species. DOMESTIC ANIMAL ENDOCRINOLOGY 41157-165, 1987. Matteri, R.L.; Ziegler, T.E. Detection of nonhuman primate gonadotropins in polyacrylamide gels: An alternative to the Western blot. AMERICAN JOURNAL OF PRIMATOLOGY 26:155-166, 1992. Morell, A.G.: Gregoriadis, G.; Scheinberg, I.H.; Hickman, J., Ashwell, G. The role of sialic acid in determining the survival of glycoproteins in circulation. JOURNAL OF BIOLOGICAL CHEMISTRY 246: 1461-1467,1971. Peckham W.D.; Knobil, E. The effects of ovariectomy, estrogen replacement and neuraminidase treatment on the properties of the adenohypophyseal glycoprotein hormones in the rhesus monkey. ENDOCRINOLOGY 98:1054-1060, 1976a. Peckham, W.D.; Knobil, E. Qualitative changes in the pituitary gonadotropins of the male rhesus monkey following castration. ENDOCRINOLOGY 98:1061-1064, 1976b. Platia, M.P.; Bloomquist, G.; Williams, R.F.; Hodgen, G.D. Refractoriness to gonadotro- pin therapy: How to distinguish ovarian failure versus pseudoovarian resistance caused by neutralizing antibodies. FERTILITY AND STERILITY 42:143-145, 1984. Reader, S.C.J.; Robertson, W.R.; Diczfalusy, E. Microheterogeneity of luteinizing hormone in pituitaw glands from women of pre- and post-menopausal age. CLINICAL ENDOCRINOLOGY 19~355-363.1983. Resko, J.A.; Belanger, A.; Labrie, F. Effects of chronic treatment with a potent luteinizing hormone-releasing hormone agonist on serum luteinizing hormone and steroid levels in the male rhesus monkey. BIOLOGY OF REPRODUCTION 26:378-384, 1982. Robertson, D.M.; Puri, V.; Lindberg, M.; Diczfalusy, E. Biologically active luteinizing hormone in plasma. V. An analysis of the differences in the ratio of biological to immunological LH activities during the menstrual cycle. ACTA ENDOCRINOLOGICA (COPENHAGEN) 92:615-626,1979. Rotten, D.; Laherty, R.F.; Monroe, S.E.; Jaffe, R.B. Induction of ovulation in the post-partum rhesus monkey: Factors determining success in obtaining primate luteal tissue. FERTILITY AND STERILITY 45: 859466.1986. Strollo, F.; Hariin, J.; Hernandez-Montes, H.; Robertson, D.M.; Zaidi, A.A.; Diczfalusy, E. Qualitative and quantitative differences in the isoelectric focusing profile of biologically active lutropin in the blood of normally menstruating and post-menopausal women. ACTA ENDOCRINOLOG ICA (COPENHAGEN) 97~166-175,1981. Wide, L.; Hobson, B.M. Qualitative differences in follicle-stimulating hormone activity in the pituitaries of young women compared to that of men and elderly women. JOURNAL OF CLINICAL ENDOCRINOLOGY AND METABOLISM 56: 371-375,1983. Wolf, D.P.; Thompson, J.A.; ZelinskiWooten, M.B.; Stouffer, R.L. In vitro fertilization-embryo transfer in nonhuman primates: The technique and its applications. MOLECULAR REPRODUCTION AND DEVELOPMENT 27:261-280,1990.