Effects of Immunization Against Bone Morphogenetic Protein 15 and Growth Differentiation Factor 9 on Ovulation Rate, Fertilization, and Pregnancy in Ewes¹

Abstract

Immunization of ewes against growth differentiation factor 9 (GDF9) or bone morphogenetic protein 15 (BMP15) can lead to an increased ovulation rate; however, it is not known whether normal pregnancies occur following such treatments. The aims of the present study were to determine the effects of a short-term immunization regimen against BMP15 and GDF9 on ovulation rate, fertilization of released oocytes, the ability of fertilized oocytes to undergo normal fetal development, and the ability of immunized ewes to carry a pregnancy to term. Ewes were given a primary and booster immunization against keyhole limpet hemocyanin (KLH; control, n = 50), a GDF9-specific peptide conjugated to KLH (GDF9, n = 30), or a BMP15-specific peptide conjugated to KLH (BMP15, n = 30). The estrous cycles of all ewes were synchronized, and ewes were joined with fertile rams approximately 14 days after the booster immunization. The number of corpora lutea was determined by laparoscopy 3–4 days following mating. Subsequently, about one-half of the ewes in each group underwent an embryo transfer procedure 4–6 days following mating, with the embryos being transferred to synchronized, nonimmunized recipients. The remaining ewes were allowed to carry their pregnancies to term. Short-term immunization against either BMP15 or GDF9 peptides resulted in an increase in ovulation rate with no apparent detrimental affects on fertilization of released oocytes, the ability of fertilized oocytes to undergo normal fetal development, or the ability of the immunized ewes to carry a pregnancy to term. Therefore, regulation of BMP15, GDF9, or both is potentially a new technique to enhance fecundity in some mammals.

ovary, pregnancy, corpus luteum, follicle, growth factors

Introduction

Growth differentiation factor 9 (GDF9) and bone morphogenetic protein 15 (BMP15 also known as GDF9B) are related members of the transforming growth factor β (TGFβ) superfamily that are produced by the ovary and have profound effects on fertility [1, 2]. In mice, GDF9 [3] but not BMP15 [4] is essential for normal follicular development. However, in sheep, both GDF9 and BMP15 are essential for normal follicular growth [5], as repeated, long-term immunizations against either GDF9 or BMP15 results in cessation of follicular growth. Morphological examination of the ovaries of immunized ewes revealed that normal follicular growth past the type 2 or primary stage was rarely observed in ewes immunized against either GDF9 or BMP15 [5]. In particular, it was found that luteal function was often abnormal as assessed from plasma progesterone concentrations, and the oocytes were abnormal in appearance, being enlarged and surrounded by an irregular arrangement of granulosa cells in the cumulus complex. Abnormalities in oocytes have also been observed in GDF9 knockout mice [6], and while follicular growth appears normal in BMP15 knockout mice, these mice have smaller litters due to reduced fertilization [4]. In sheep, the function of the corpus luteum (CL), as measured by secretion of progesterone, was also abnormal following passive immunization of ewes against GDF9 [5].

In addition to their essential role in follicular growth, both of these oocyte-derived growth factors influence ovulation rate in sheep ([2, 5, 7–9] and J. Hanrahan and S. Galloway, personal communication). Indeed, we have recently shown an increase in ovulation rate in some ewes when immunized against BMP15 or GDF9 using a dextran-based adjuvant [2]. However, based on the observed abnormalities in long-term or passively immunized ewes, it was unclear whether neutralization of a portion of BMP15 or GDF9 would affect subsequent pregnancy outcomes. Furthermore, alterations in pregnancy outcomes potentially could be attributed to a change in oocyte function or insufficient luteal function. Therefore, the objectives of this experiment were to determine the effect of short-term immunization against BMP15 and GDF9 on ovulation rate, fertilization of released oocytes, ability of fertilized oocytes to undergo normal fetal development, and the ability of immunized ewes to carry a pregnancy to term.

Materials and Methods

All experiments were performed with the approval of the Animal Ethics Committee at Wallaceville Animal Research Centre in accordance with the 1999 Animal Welfare Act Regulations of New Zealand. All animals had ad libitum access to pasture and water. Except where indicated, laboratory chemicals were obtained from BDH Chemicals New Zealand Ltd (Palmerston North, New Zealand), Invitrogen (Auckland, New Zealand), or Roche Diagnostics N.Z. Ltd. (Auckland, New Zealand).

Immunization of Ewes Against GDF9 and BMP15

The peptides KKPLVPASVNLSEYFC (GDF9) and SEVPGPSREHDGPESC (BMP15) were synthesized and conjugated to keyhole limpet hemocyanin (KLH) through the C-terminal cysteine residue by Macromolecular Resources (Colorado State University, Fort Collins, CO). These peptides were chosen based on the dissimilarity of the peptide sequence with other members of the TGFβ superfamily as assessed by a basic local alignment search tool (BLAST) search for short, nearly exact, matches [10] and previous results indicating that immunization against these peptides influenced ovulation rate [2, 5]. Sexually mature Romney ewes (approximately 50 kg, 4–6 yr old) were given a primary immunization with 0.4 mg KLH (N = 50, KLH), KLH-GDF9 (n = 30, GDF9) or KLH-BMP15 (n = 30, BMP15) in 5% (w/v) DEAE-dextran adjuvant followed by a 0.2 mg booster injection 4 wk later. The estrous cycles of the immunized ewes were synchronized at the time of the booster by giving an intravaginal controlled internal drug release device (CIDR; Pharmacia & Upjohn Limited Company, Auckland, New Zealand) containing progesterone for 13 days, with Estrumate (125 μg; Schering-Plough Animal Health Limited, Upper Hutt, New Zealand) given on Day 12 to induce luteolysis of any remaining CL. To maintain progesterone levels, the initial CIDR was replaced with a new CIDR on Day 10. At the time the replacement CIDR was removed, intact sexually mature Poll Dorset rams (n = 2), fitted with mating harnesses and crayons for determination of mating, were introduced to the immunized ewes. In addition, the estrous cycles of nonimmunized sexually mature Romney ewes were synchronized at the same time using identical procedures. The number of CL was determined by laparoscopy in all immunized ewes 3 days after mating. Embryos from approximately one-half of the ewes in each treatment group (KLH, n = 26; GDF9, n = 15; BMP15, n = 16) were surgically transferred to synchronized recipients (n = 43) on Days 4–6 of the estrous cycle as previously described [11]. All ewes having four or more CL, most ewes with three CL, as well as a random selection of ewes with one or two CL were selected to undergo the embryo transfer procedure. Recipients received one to three embryos each. The number of CL in the ewes from which embryos were recovered was also determined during the subsequent estrous cycle that was induced by an injection of Estrumate given on approximately Day 10 of the estrous cycle. The ewes not subjected to embryo transfer were allowed to carry their pregnancies to term.

Determination of Antibody Titers and Cross-Reactivity

Sera were collected from all immunized ewes prior to immunization and 2 wk after the booster immunization for determination of reactivity to GDF9 and BMP15. Antibody titers and cross-reactivity were determined by ELISA as described [5] with the following modifications: Sera samples were diluted to 1:10 000 from the BMP15 and KLH immunized animals to measure antibody titers for BMP15, whereas sera samples were diluted to 1:1000 from the GDF9 and KLH immunized animals for measurement of antibody titers for GDF9. In addition, the amount of Escherichia coli-expressed mature region of ovine GDF9 coated onto each well was increased to 200 ng and the incubation time for the interaction of sera and antigen was reduced to 1 h.

Determination of Concentrations of Progesterone in Plasma

Plasma was collected from all immunized ewes at Days 1, 3 (administration of PGF_2α analogue), 4 (removal of CIDR), 5 (estrus), 7, and 9. Blood samples from those ewes that did not undergo the embryo flushing procedure were also collected at Days 14 and 19, and approximately every 10 days thereafter until Day 105 of pregnancy. Blood samples were also collected from recipient ewes beginning on the day they received an embryo and then with the same schedule as the immunized ewes carrying a pregnancy to term. The concentrations of progesterone in plasma were determined by RIA as previously described [12]. The sensitivity of the assay was 0.1 ng/ml, and the average interassay and intra-assay coefficients of variation were less then 10%.

Statistical Analysis

Antibody response was determined by comparing the optical density (OD) reading before immunization to those 2 wk following the booster immunization by paired t-test. One ewe in the BMP15 immunized group was excluded as she was shown to have nonspecific binding to BSA in her preimmune serum sample. Because no specific displacement was observed with the potentially cross-reacting protein for any animal, maximum cross-reactivity was calculated by expressing the concentration of specific protein required to displace 50% of maximum binding as a percentage of the maximum amount of cross-reacting protein tested in the assay. Animals in which the maximum binding was not determined were excluded from analyses.

Ovulation rate was calculated as the number of CL at a single observation (i.e., in ewes carrying pregnancy to term) or averaged for two observations (i.e., in ewes whose embryos were transferred to recipients). Four ewes in the BMP15 immunized group did not show estrus for approximately 30 days. For these animals, the ovulation rates were calculated by counting the number of CL after the first ovulation (two ewes) or following the first two ovulations (two ewes). Differences in ovulation rate were determined using chi-square analysis following grouping of the animals into three classes; namely, those with mean ovulation rates of 1 to <2, 2 to <3, or ≥3.

Differences in fertilization rates were determined using chi-square analysis by considering whether or not an embryo was recovered for every CL present on the ovary. In all groups, the data from ovaries in which flushing of the oviduct was noted as incomplete were excluded. Differences in embryo survival were determined using chi-square analysis after considering whether a lamb was born following transfer of an embryo. Three embryos from the BMP15 group were not transferred. Two of the embryos were from one of the ewes that was anestrous for approximately 30 days and were more advanced than anticipated at recovery and thus were not recovered whole, the third embryo was lost during the transfer procedure.

Differences in lambing percentages were determined by comparing the percentage of ewes carrying a pregnancy to term using chi-square analysis. One ewe in the KLH immunized group became ill and was killed shortly after breeding and was thus excluded from further analysis. Differences in the proportion of released oocytes (as assessed by number of CL) that resulted in the birth of a lamb in the immunized ewes were also determined by chi-square analysis. Effects of immunization on concentrations of progesterone in plasma were assessed using an ante-dependence analysis of order 1 in Genstat following natural log transformation of the data. Where progesterone values were at or below the sensitivity of the assay, a value of 0.1 ng/ml (i.e., the sensitivity of the assay) was assigned for the above analysis.

Results

Antibody Titers and Cross-Reactivities

Ewes immunized with KLH did not have measurable antibodies to either GDF9 or BMP15 (Table 1). Although a small amount of binding was observed to GDF9, this was deemed nonspecific, because binding levels did not increase following immunization. No binding was observed to BMP15 in sera collected from ewes prior to immunization or following immunization with KLH. Ewes immunized against GDF9 had measurable antibodies against GDF9 (P < 0.001) and those immunized against BMP15 had measurable antibodies against BMP15 (P < 0.001). Maximum cross-reactivity of antisera collected from BMP15 immunized ewes (n = 28) to GDF9 averaged 1.2% (range 0.1%–3.2%) and, where they were measurable (n = 12 ewes), maximum cross-reactivity of ewes immunized against GDF9 to BMP15 averaged 1.8% (range 0.8%–3.0%).

Table 1

Mean ± SEM optical density readings following ELISA using specified coating antigens of sera samples pre- and postimmunization.

Immunization treatment	Coating antigen
	GDF9		BMP15
	Preimmune	Postimmune	Preimmune	Postimmune
KLH	0.31 ± 0.01	0.33 ± 0.02	0.06 ± 0.004a	0.06 ± 0.004a
GDF9	0.31 ± 0.01	1.26 ± 0.12b	ND	ND
BMP15	ND	ND	0.07 ± 0.003a	1.02 ± 0.14b

^a

Not significantly different than background readings.

^b

Denotes difference between preimmune and postimmune samples (P < 0.001). ND, not determined.

Effect of Immunization on Ovulation Rate, Fertilization of Oocytes, Embryo Survival, and Lambing Percentages

Immunization against either GDF9 or BMP15 increased ovulation rate when compared to KLH immunized control ewes (P < 0.001; Fig. 1). The mean ovulation rates for the KLH, GDF9, and BMP15 immunized ewes were 1.8, 2.2, and 2.6, respectively. Neither the ability of ovulated oocytes to be fertilized nor embryo survival was affected by immunization against either GDF9 or BMP15 (Fig. 2). Furthermore, the ability of a ewe to carry her pregnancy to term was not affected by immunization against either GDF9 or BMP15 (Fig. 3).

Fig. 1

Distribution of ovulation rate in KLH (n = 50), GDF9 (n = 30), and BMP15 (n = 30) immunized ewes. Distribution of ovulation rate in GDF9 and BMP15 immunized ewes was different from that observed in KLH immunized ewes (P < 0.01)

Fig. 2

Percentages of embryos recovered based on the number of CL and percentages of lambs born per embryo transferred. Total numbers of CL were 48, 38, and 46 for KLH, GDF9, and BMP15 immunized ewes, respectively. Total numbers of embryos transferred were 32, 23, and 20 for the KLH, GDF9, and BMP15 immunized ewes, respectively. No significant differences were observed among treatments

Fig. 3

Percentages of ewes lambing and the percentages of lambs born based on the number of CL present at laparoscopy. Total numbers of ewes left to lamb were 23, 14, and 13 for the KLH, GDF9, and BMP15 immunized groups, respectively. Total numbers of CL present at laparoscopy were 36, 23, and 26 for the KLH, GDF9, and BMP15 immunized groups, respectively. No differences were observed among groups

Effects of Immunization on Secretion of Progesterone During Pregnancy

Immunization against BMP15 or GDF9 did not affect the pattern of secretion of progesterone during the first estrous cycle (Fig. 4). In addition, the concentrations of progesterone in pregnant immunized or recipient ewes were not different among the groups (Figs. 4 and 5).

Fig. 4

Mean concentrations of progesterone (P₄) in plasma of all immunized ewes not undergoing the embryo transfer procedure during the first 20 days of sampling. There were no significant differences among treatments

Fig. 5

Mean concentrations of progesterone in plasma of pregnant immunized ewes (top) and ewes carrying an embryo from an immunized ewe to term (bottom). Within a group (immunized or recipient), there were no significant differences among treatments

Discussion

A short-term immunization against either GDF9 or BMP15 using a DEAE dextran adjuvant increased ovulation rate in ewes. The increased ovulation rate following immunization was due to specific neutralization of GDF9 or BMP15, because very little if any cross-reaction was observed when tested against its most closely related family member (i.e., BMP15 for GDF9 and GDF9 for BMP15). These findings confirm our results from an earlier, preliminary study in which some ewes had increased ovulation rates following immunization with either an E. coli expressed mature region of ovine GDF9 or BMP15 or the peptides used in the present study [2]. In addition, this is in agreement with increased ovulation rates observed in ewes heterozygous for inactivating mutations in either BMP15 or GDF9 ([9] and J. Hanrahan and S. Galloway, personal communication).

It has been hypothesized that the mutation in Inverdale sheep alters follicular development by changing the processing of GDF9 and thus decreasing GDF9 secretion and that this reduction in GDF9 levels may be involved in the infertility observed in homozygous Inverdale ewes [13]. This hypothesis is based on the reduced expression of the mature form of GDF9 and BMP15 in cells coexpressing the mutant Inverdale form of BMP15 and normal GDF9. The present results are inconsistent with the effects of the Inverdale mutation being mediated through reduced secretion of GDF9, because it is unlikely that an antibody would be able to influence intracellular processing of BMP15 or GDF9. Based on the current study, we hypothesize that in the regulation of ovulation rate in sheep, either BMP15 and GDF9 homodimers have essential but similar roles, or that BMP15/GDF9 heterodimers have essential roles, or both.

In sheep, long-term immunization against either the BMP15 or GDF9 peptides used in these trials with an adjuvant capable of eliciting a strong immune response (i.e., Freunds adjuvant) resulted in abnormal follicular development [5]. These abnormalities were particularly evident in the oocyte itself as well as the cumulous granulosa cells surrounding the oocyte. Neutralization of GDF9 also appeared to affect the function of the CL. Oocyte function is altered in mice lacking functional GDF9 [6], and fertilization is reduced in BMP15 knockout mice [4]. This suggests that both GDF9 and BMP15 may play essential roles in not only regulating ovulation rate but also oocyte health and establishment of pregnancy. In the current study, neutralization of sufficient amounts of GDF9 or BMP15 to increase ovulation rate did not appear to dramatically affect oocyte health or the ability of the embryo to develop. Immunized ewes were also able to carry a pregnancy to term. Similar results are observed in ewes heterozygous for the Inverdale gene, in which ovulation rate increases lead to a predicable increase in litter size [14]. It could be argued that neutralization of a large portion of GDF9 or BMP15, which would be predicted to lead to a superovulation type response, might result in reduced quality of oocytes released or decreased ability of the immunized ewe to carry a pregnancy to term. Some of the ewes in the present study immunized with BMP15 peptide did undergo a short period of anovulation before returning to estrous with increased ovulation rates. These oocytes were apparently healthy, as they resulted in the birth of normal, healthy live lambs. In addition, the immunization regime used in the current experiment did not appear to affect luteal function as no differences were noted in the concentrations of progesterone in sera of immunized ewes. This study was not designed to examine the effects of GDF9 or BMP15 immunization on increasing lamb production. Such studies are underway and require much larger numbers of animals. However, because no untoward effects were noted on pregnancy outcomes, albeit with small number of ewes, there is a reasonable expectancy that immunization of sheep with GDF9 or BMP15 will lead to increased lamb production.

In summary, short-term immunization against either BMP15 or GDF9 peptides resulted in increased ovulation rate with no apparent detrimental effect on fertilization of released oocytes, ability of fertilized oocytes to undergo normal fetal development, or the ability of immunized ewes to carry a pregnancy to term. Therefore, regulation of BMP15, GDF9, or both is potentially a new technique for enhancing fecundity in some mammals.

Acknowledgments

The authors thank Dr. George Davis for his helpful comments regarding the manuscript. The authors also thank Doug Jensen for the management of animal care; Laurel Quirke, Peter Smith, and James Fitzwater for assistance with sample collection and embryo transfers; Karen Reader and Steve Lawrence for assistance with determination of antibody titers; and Lilian Morrison for assistance with statistical analysis.

Author notes

¹

Supported by the New Zealand Foundation for Research, Science and Technology, the Royal Society of New Zealand Marsden Fund, and Ovita Limited, New Zealand.