In Vitro Growth and Ovulation of Follicles from Ovaries of Estrogen Receptor (ER) and ER? Null Mice Indicate a Role for ER? in Follicular Ma
http://www.100md.com
内分泌学杂志 2005年第6期
Receptor Biology Section (J.M.A.E., J.F.C., M.M.Y., K.S.K.), Laboratory of Reproductive and Developmental Toxicology, Laboratory of Experimental Pathology (S.A.E.), and Biostatistics Branch (G.E.K.), National Institute of Environmental Health Sciences/National Institutes of Health, Research Triangle Park, North Carolina 27709
Address all correspondence and requests for reprints to: Dr. Kenneth S. Korach, Receptor Biology Section, Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences/National Institutes of Health, MD B3-02, P.O. Box 12233, Research Triangle Park, North Carolina 27709. E-mail: korach@niehs.nih.gov.
Abstract
Both estrogen receptor (ER) and ? are expressed within the ovary and lack of either of these receptors affects ovarian function. In this study, the role of ER and ER? in folliculogenesis and ovulation was further analyzed. Evaluation of ovarian follicle populations in wild-type and ER? knockout (?ERKO) ovaries revealed reduced late antral growth and ovulatory capacity of ?ERKO follicles, indicated by reduced numbers of large antral follicles and corpora lutea and increased atresia of large antral follicles. An in vitro culture system was used to study growth, rupture, and luteinization of wild-type, ER knockout (ERKO) and ?ERKO ovarian follicles. ERKO follicles exhibited wild-type-like growth and ovulation rates but an increased capacity to synthesize estradiol. In contrast, ?ERKO follicles showed a significant lack of progression from early antral to large antral stage, decreased estradiol production, and reduced ovulation. Expression patterns of several genes involved in follicle maturation and ovulation were analyzed in follicles grown in vitro. Ar, Pgr, and Has2 mRNA expression levels were the same among the three genotypes. However, ?ERKO follicles showed reduced expression of Cyp19 mRNA during follicle maturation and reduced Lhcgr and Ptgs2 mRNA expression after human chorionic gonadotropin stimulus. Luteinization occurs normally in ERKO and ?ERKO follicles, shown by increased progesterone secretion and increased cdkn1b mRNA expression after human chorionic gonadotropin. Collectively, these data indicate that ER?, but not ER, plays a direct role in folliculogenesis. ER? appears to facilitate follicle maturation from the early antral to the preovulatory stage.
Introduction
ESTROGENS PLAY A KEY role in female reproduction, with the ovary serving as both the primary site of estrogen synthesis and a major estrogen target. The importance of estrogen actions in the ovary have recently been illustrated by the phenotypes of mice lacking functional estrogen receptors (ER) and/or ? [ER knockout (ERKO)] and those lacking the capacity to produce estradiol, the Cyp19 [aromatase knockout (ArKO)] null mice (1, 2). Adult ERKO and ArKO females are infertile and exhibit ovaries that possess large, hemorrhagic and cystic follicles and no evidence of spontaneous ovulation, whereas adult ?ERKO females are subfertile and possess ovaries that are best characterized as exhibiting a paucity of growing follicles and corpora lutea (2, 3, 4, 5). These findings, along with the expression pattern of each nuclear ER form within the ovary suggest a critical role for intraovarian estrogen signaling in ovarian function (6, 7, 8).
In the rodent ovary, estrogen is primarily produced by preovulatory follicles under the influence of FSH (9). The endocrine actions of estrogen in the female reproductive tract and hypothalamic-pituitary axis are well documented and critical to reproduction (1). However, estrogen is also thought to fulfill a number of intraovarian roles during folliculogenesis. For example, in granulosa cells of maturing follicles, estradiol has little effect alone but is required for maximum FSH stimulation of 1) Cyp19 expression and estradiol synthesis (10, 11, 12, 13); 2) LH receptor (Lhcgr) expression and LH responsiveness (14, 15, 16); 3) antrum formation (17, 18, 19); 4) gap-junction formation (20); and 5) prevention of atresia (21). In thecal cells of growing follicles, estradiol modulates androgen synthesis by negatively regulating Cyp17 expression in thecal cells (22, 23).
Estrogens signal via two nuclear estrogen receptors, ER and ER?, and both are present in the ovary but in different cell types. Granulosa cells of follicles from the primary to large antral stage abundantly express ER?-encoding transcripts and protein, whereas expression is low or absent in interstitial and luteal cells (7, 24, 25). In contrast, ER-encoding transcripts and protein are predominantly localized to thecal and interstitial cells with little or no expression in granulosa cells (24, 26, 27, 28). Whether oocytes express ER remains controversial; however, because oocytes from both ER null mice and ArKO mice are able to undergo fertilization and yield viable embryos, neither ER nor estrogen appears critical to oocyte integrity (4, 29, 30). The divergent pattern of expression of ER and ER? in the ovary and the distinct ovarian phenotypes exhibited by ERKO and ?ERKO females indicate that the two ER forms have distinct roles in ovarian function. Although ERKO ovaries contain normal follicles up to the small antral stage, the predominant features are multiple, large hemorrhagic follicles and the absence of corpora lutea, indicating an anovulatory phenotype (5, 31). We have previously shown that this dramatic phenotype in the ERKO ovary is secondary to chronically elevated levels of plasma LH that result from the loss of ER-mediated negative feedback in the hypothalamus (23, 29). In contrast, ?ERKO female mice exhibit normal levels of circulating gonadotropins and ovaries that possess all stages of folliculogenesis, including corpora lutea (4, 23). However, when bred, ?ERKO female mice yield significantly fewer litters and offspring compared with wild-type counterparts (4). Furthermore, the oocyte yield from ?ERKO female mice after gonadotropin-induced ovulation is approximately 15% that of wild-type females (4, 32). These data suggest that ER? plays a more direct role in ovarian function, although the exact mechanism has yet to be determined.
Detailed studies of the intraovarian events during follicular development and ovulation are difficult to carry out in vivo because of the various endocrine factors known to influence ovarian function, including gonadotropins, prolactin, steroids, and various growth factors. This is especially true for the ERKO mouse because ovaries of these mice exist in a severely abnormal hormonal environment (23). To gain further insight into the role(s) of ER and ER? in ovarian function we have established an in vitro culture system using follicles from ERKO and ?ERKO ovaries. This system allows us to compare growth, steroidogenesis, and ovulation rates between wild-type, ERKO, and ?ERKO follicles in a controlled hormonal environment. In addition, expression patterns of several genes known to be involved in follicular maturation, ovulation, or luteinization were evaluated in follicles grown in vitro. Using in vitro culture of follicles, we have shown that ER?, and not ER, is most critical to follicle growth and maturation leading up ovulation.
Materials and Methods
Animals
All animal procedures were approved by the National Institute of Environmental Health Sciences Animal Care and Use Committee. Generation of ERKO and ?ERKO null mice has been previously described by Lubahn et al. (3) and Krege et al. (4), respectively. Mice were generated from ERKO and ?ERKO heterozygous breeding pairs on a background of C57BL6 obtained from Taconic Farms (Germantown, NY). Offspring were genotyped by PCR analysis of tail-snip genomic DNA as previously described (23).
Ovarian follicle counts in vivo
Ovaries were collected from immature (28–30 d) wild-type and ?ERKO female mice 48 h after a single injection of 2.2 IU of pregnant mare serum gonadotropin (PMSG) (Sigma Chemical Co., St. Louis, MO) or vehicle (0.85% saline). In addition, ovaries were collected from adult (109 d) wild-type and ?ERKO female mice that were untreated. The ovaries were fixed in 10% neutral buffered formalin, paraffin embedded, sectioned (5 μm), and stained with hematoxylin and eosin. One ovary per animal was used for follicle counts. Follicles were counted in every 10th section, beginning with the fifth section, through the entire ovary. Only follicles with a visible nucleus were counted to avoid double counting. Follicles were categorized as primordial if they contained an intact oocyte surrounded by a single layer of predominantly squamoid, flattened epithelial cells, lacking a connective tissue or thecal investment. Follicles were categorized as primary if they had an enlarged oocyte surrounded by a single layer of predominantly cuboidal granulosa cells. Small preantral follicles were identified as having an oocyte surrounded by two layers of granulosa cells, whereas large preantral follicles were identified as having an oocyte surrounded by a multilayered (more than two), solid mantle of granulosa cells. Follicles were categorized as small antral if they had a central oocyte surrounded by layered granulosa cells and containing fluid-filled areas within the granulosa cell layer. Large antral follicles were identified as having a single cavity containing follicle fluid surrounded by a granulosa cell layer of even thickness and a cumulus-cell enclosed oocyte. In addition, the total number of corpora lutea was counted. Atretic follicles, defined as follicles containing one or more apoptotic cells with pyknotic nuclei, were not included in the above categories but counted and categorized separately. The presence of apoptotic cells was verified using the terminal deoxynucleotidyl transferase biotin-dUTP nick end labeling assay (Upstate Biotechnology, Charlottesville, VA) according to the manufacturer’s protocol. For each ovary, total number of follicles (per follicle class) present in the scored sections was normalized to the total number of scored sections from that ovary to obtain a mean follicle count per section. The average number of scored sections per ovary was the same between ?ERKO animals and the respective wild-type controls.
No data on follicle counts were collected from ERKO ovaries, because these ovaries exhibit extensive formation of hemorrhagic cystic follicles that distort the overall morphology and make it difficult to obtain representative follicle counts.
In vitro culture of follicles
Follicles were isolated from whole ovaries of 21- to 25-d-old female mice (wild-type, ERKO, and ?ERKO) according to the method of Nayudu and Osborne (33). In brief, mice were killed by CO2 asphyxiation, and the ovaries were removed and placed in Leibovitz’s L-15 medium (Life Technologies, Inc., Invitrogen, Carlsbad, CA) supplemented with insulin (5 μg/ml; Life Technologies), transferrin (10 μg/ml; Sigma), selenium (2 ng/ml; Sigma), ascorbic acid (50 μg/ml; Sigma), and 0.3% BSA (Sigma) at 37 C. Large preantral follicles (200 ± 20 μm in diameter) were then mechanically dissected from the ovaries using 25-gauge syringe needles. Isolated follicles were transferred to four-well multidishes (Nalge, Nunc International, Rochester, NY) containing -MEM (Life Technologies) supplemented with insulin, transferrin, selenium, and ascorbic acid as above as well as penicillin (100 IU/ml; Life Technologies), streptomycin (100 μg/ml; Life Technologies), 5% fetal bovine serum (Sigma), and 100 mIU recombinant human FSH (kindly donated by Ares Serono, Geneva, Switzerland). The culture dishes were placed in a humidified incubator at 37 C in 5% CO2 in air until a sufficient number of follicles had been collected.
The harvested preantral follicles were evaluated, and those exhibiting an intact basement membrane, a high density of granulosa cells, a centrally located oocyte, and some thecal cells attached were selected for further culture. The culture method used for in vitro evaluation of follicle growth and androstenedione, testosterone, and estradiol secretion was based on that of Nayudu and Osborne (33) with modifications according to Rose et al. (34). Follicles were individually cultured in Millipore CM culture plate inserts (Millipore, Bedford, MA) filled with 250 μl -MEM supplemented as described above, at 37 C in 5% CO2 in air. Medium was changed on culture d 1, 3, and 5. Spent medium collected on culture d 5 was stored at –80 C until assayed. Follicular growth was monitored daily by measuring follicle diameter using a precalibrated ocular micrometer. Follicles that did not survive the 5-d culture period, those with obvious darkening of the granulosa cell mass and/or signs of disintegration or that had spontaneously extruded their oocyte, were not included in the analysis.
For in vitro ovulation and progesterone production assays, follicles were cultured in Costar Ultra Low flat-bottom 96-well plates (35). After isolation and selection, follicles were transferred individually into wells containing 40 μl -MEM, supplemented as described above. The culture dishes were placed in a humidified incubator at 37 C in 5% CO2 in air. Medium was changed every 48 h. When follicles reached a diameter of at least 340 μm, after approximately 4 or 5 d of culture, the medium was replaced by medium containing human chorionic gonadotropin (hCG) (2–32 IU/ml) (kindly donated by Ares Serono, Geneva, Switzerland) to induce ovulation. Follicles were checked for ovulation at 16 h after hCG stimulus and were considered to have ovulated when the follicle was visibly ruptured and the cumulus oocyte complex (COC) extruded.
The size at which large antral follicles were exposed to hCG was determined as follows. Large antral wild-type follicles were divided into three groups: follicles that had reached a diameter of 320–340 μm, 340–360 μm, or more than 360 μm. After treatment with 2 IU/ml hCG, 35% of follicles with a diameter of 320–340 μm, 60% of follicles with a diameter between 340 and 360 μm, and 70% of follicles with a size greater than 360 μm ovulated. Initial follicle growth studies indicated a reduction in the number of ?ERKO follicles reaching a size of 340 μm or more. To collect sufficient late antral ?ERKO follicles for the ovulation assays, these assays were performed with follicles that had reached a diameter of at least 340 μm.
In some of the cultures, the effect of prostaglandin on ovulation was evaluated by in vitro ovulation assays and by histology. Prostaglandin E2 (PGE2) (500 ng/ml) (36) dissolved in ethanol was added to the medium 4 h after the hCG pulse, and follicles were checked for ovulation 16 h later. For histological analysis, follicles were collected before hCG or 16 h after incubation with hCG alone or a combination of hCG and PGE2, fixed in 10% neutral buffered formalin, embedded in paraffin, sectioned (5 μm), and stained with hematoxylin and eosin. At least five follicles per experimental group per genotype were studied, and representative photomicrographs are presented for histology.
RIA
Androstenedione, testosterone, and estradiol were measured in spent medium collected on culture d 5, after incubation for 48 h. Androstenedione and testosterone were assayed in a single assay using direct RIA kits from Diagnostics Systems Laboratories (Webster, TX), according to the manufacturer’s protocol. The intraassay coefficients of variation were 4.3% for androstenedione and 8.5% for testosterone, and the assay sensitivities were 0.02 ng/ml for androstenedione and 0.08 ng/ml for testosterone. Estradiol was measured using a double-antibody RIA kit (Diagnostics Systems Laboratories) with inter- and intraassay variations of 8.5 and 4.6%, respectively, and a sensitivity of 4.7 pg/ml. Progesterone production was measured in spent medium collected just before or 24 h after hCG using a direct RIA kit (Diagnostics Systems Laboratories). The inter- and intraassay coefficients of variation were 11.7 and 6.6%, respectively, and the assay sensitivity was 0.12 ng/ml. For each hormone, the level in medium that had not been in contact with follicles was below the level of detection.
Analysis of mRNA expression
Follicles that reached a diameter of at least 340 μm were collected either immediately (no hCG) or after incubation with 2 IU/ml hCG for 3 or 24 h. Total RNA was extracted from individual follicles using PicoPure RNA isolation kit (Arcturus, Mountain View, CA) according to the manufacturer’s protocol. To remove any potential genomic DNA contamination, samples were incubated with 2 μl RNase-free DNase I (2 U/μl) (Ambion Inc., Austin, TX) for 15 min at room temperature. Semiquantitative RT-PCR was performed as described previously (23) with some modifications. Total RNA (0.1 μg) was reverse transcribed in a total volume of 25 μl using the Superscript first-strand synthesis system (Invitrogen Life Technologies Inc.). PCR amplification was carried out with 1 μl of the RT reaction in a 15-μl reaction at 95 C for 30 sec, 58 C for 45 sec, and 72 C for 30 sec for 28 cycles. All samples were electrophoresed on agarose gels containing ethidium bromide and electroblotted to BrightStar nylon membrane (Ambion). The membranes were hybridized with 33P-labeled probes and washed, and results were quantified by using a PhosphorImager and ImageQuant software (Molecular Dynamics, Inc., Sunnyvale, CA). Per experiment, the intensity values for each transcript were normalized to the respective expression level of Actb, a housekeeping gene that is nondifferentially expressed, and then expressed as fold induction over the wild-type no-hCG value by dividing each value by the wild-type no-hCG mean value. The primers and probes used for the semiquantitative RT-PCR are listed in Table 1.
TABLE 1. Primer pairs and probes used for semiquantitative RT-PCR
Statistical analysis
Within the follicle count studies, differences between the mean numbers of follicles per section were evaluated for each follicle class per genotype using the nonparametric Mann-Whitney U test. Data of ERKO and ?ERKO follicles cultured in vitro obtained in separate experiments were analyzed as separate groups. Each experiment always included follicles derived from wild-type ovaries as controls. Follicle growth in vitro was analyzed by comparison of daily follicular sizes between genotypes using repeated-measures ANOVA followed by two-sample t tests comparing mean size at each day. Differences between genotypes in reaching a large antral size or ovulation rates were assessed by 2 tests. Differences in estradiol secretion among genotypes were identified by nonparametric Kruskal-Wallis ANOVA followed by Mann-Whitney U tests. Comparisons of the secreted amount of progesterone and of the gene expression data between the genotypes were done using the Mann-Whitney U test. All results are presented as mean ± SE of the mean (SEM). Differences were considered significant at P < 0.05.
Results
?ERKO ovaries exhibit a reduced number of large antral follicles and corpora lutea
The follicle distribution in wild-type and ?ERKO ovaries is shown in Fig. 1. Ovaries from untreated 28-d-old wild-type and ?ERKO mice revealed similar numbers of primordial and primary follicles per section (Fig. 1A). Among the pool of growing follicles in immature mice, only the number of large preantral follicles per section in ?ERKO ovaries vs. wild-type ovaries was significantly increased (P < 0.05). Although the total number of atretic follicles per section did not significantly differ between immature wild-type and ?ERKO ovaries (Fig. 1A), there were fewer atretic small antral follicles per section in ?ERKO ovaries compared with wild-type ovaries (P < 0.05).
FIG. 1. Distribution of the follicle populations in wild-type and ?ERKO ovaries. A, Follicle population (healthy and atretic) in ovaries of immature wild-type (WT) and ?ERKO mice 28–30 d of age after treatment with vehicle for 48 h (n = 4 per group). B, Follicle population (healthy and atretic) in ovaries of immatureWT and ?ERKO mice 28–30 d of age after treatment with PMSG for 48 h (n = 7 wild type; n = 9 ?ERKO). C, Follicle population (healthy and atretic) in ovaries of adult WT and ?ERKO approximately 109 d of age (n = 6 per group). Shown is the mean number of follicles per section for each follicle class per genotype (± SEM). *, P < 0.05 vs. respective WT value; **, P < 0.01 vs. respective WT value.
PMSG treatment for 48 h had no marked effect on the pool of primordial or primary follicles in immature ovaries of either genotype (Fig. 1B). There were also no differences in the number of preantral follicles per section between ?ERKO and wild-type PMSG-treated ovaries. Treated ?ERKO ovaries did exhibit an increased number of small antral follicles per section compared with treated wild-type ovaries (P < 0.05). As expected, PMSG treatment led to the presence of large antral follicles that were not observed in ovaries of untreated animals (compare Fig. 1, A with B; P < 0.01, untreated vs. treated wild-type ovaries; P < 0.05, untreated vs. treated ?ERKO ovaries). Interestingly, PMSG-treated ?ERKO females exhibited significantly fewer large antral follicles per section (P < 0.01) compared with treated wild type. No differences were observed in the number of atretic follicles per section, total or per category, between PMSG-treated wild-type and PMSG-treated ?ERKO ovaries (Fig. 1B).
A comparison of follicle counts per section between adult wild-type and ?ERKO ovaries revealed that ?ERKO ovaries contain significantly more primordial (P < 0.01) but significantly fewer primary follicles per section (P < 0.05; Fig. 1C). ?ERKO ovaries also exhibited a significantly reduced number of large preantral follicles and corpora lutea per section (P < 0.05). Although no difference was observed in the total number of atretic follicles between adult wild-type and ?ERKO ovaries, ?ERKO ovaries contained fewer atretic large preantral follicles and more atretic large antral follicles per section than wild-type ovaries.
?ERKO follicles fail to reach maximum size in vitro
Figure 2 shows the growth profiles of in vitro cultured follicles from ERKO (Fig. 2A) and ?ERKO (Fig. 2B) ovaries compared with follicles from wild-type ovaries. At the start of culture (d 0), preantral follicles from wild-type, ?ERKO, and ERKO ovaries exhibited a similar mean diameter, and each continued to grow and exhibit visible antrum formation by d 2. The overall growth profile of ERKO follicles did not differ from their respective wild-type controls (P = 0.08; Fig. 2A), but a significant difference in size was observed between wild-type and ERKO follicles on culture d 1 (P < 0.001) and d 5 (P < 0.01). Nonetheless, the percentage of ERKO follicles (77%) that reached 340 μm did not differ from wild type (91%; Table 2). In contrast, there was a significant difference between the overall growth profiles of wild-type and ?ERKO follicles (P < 0.0001; Fig. 2B) such that ?ERKO follicles exhibited retarded growth that first became apparent on d 2 (P < 0.01). On subsequent days, the difference in follicle diameter between wild-type and ?ERKO follicles became more pronounced (P < 0.001 on d 3, 4, and 5). As a result, the percentage of ?ERKO follicles reaching a large antral size (designated as a diameter 340 μm) after 5 d in culture was significantly lower relative to wild-type, 60 vs. 92%, respectively (P < 0.0001; Table 2).
FIG. 2. In vitro growth of follicles from wild-type (WT), ?ERKO, and ERKO ovaries. Preantral follicles were grown in vitro in the presence of 100 mIU/ml recombinant human FSH. The follicle diameter was measured each culture day. Follicles that exceeded a diameter of 340 μm (dotted line) were classified large antral follicles. A, Growth profiles of WT and ERKO follicles (n = 7–46 follicles per data point). The growth of ERKO follicles was similar to that of WT follicles (F1,218 = 3.1; P = 0.08). There was a significant difference between the size of WT and ERKO follicles on d 1 and 5 of culture. B, Growth profiles of WT and ?ERKO follicles (n = 11–133 follicles per data point). The growth profile of ?ERKO follicles was different from WT follicles (F1,630 = 181.1; P < 0.001). From culture d 2 onward, WT follicles were significantly bigger in diameter than ?ERKO follicles at each culture day. Shown is the mean follicle diameter (μm) ± SEM for each culture day per genotype. *, P < 0.01 vs. WT; **, P < 0.001 vs. WT.
TABLE 2. Percentage of wild-type, ERKO, and ?ERKO follicles that reached a large antral size
The steroidogenic capacity of follicles in culture was assessed by measuring the level of androstenedione, testosterone and estradiol that accumulated in the culture medium over a 48 h interval, between days 3 and 5 of culture. Androstenedione and testosterone in the medium were undetectable. Although we observed a considerable degree of variation in estradiol release among follicles within each genotype, a significant difference among the genotypes was evident when comparing the mean levels (Table 3). Estradiol secretion by ERKO follicles significantly exceeded that of wild-type follicles by at least 4-fold (P < 0.05), whereas ?ERKO follicles exhibited a significantly lower rate of secretion compared with wild-type follicles (P < 0.05).
TABLE 3. Estradiol secretion by follicles in vitro
?ERKO follicles do not fully respond to hCG
Wild-type, ERKO, and ?ERKO follicles that had reached a large antral size of at least 340 μm were induced to ovulate by treatment with hCG and checked for ovulation at 16 h after the hCG stimulus (Table 4). Initially, a dose of 2 IU/ml hCG was used, because this was shown to be the optimal dose to induce ovulation in vitro by Rose et al. (34). After exposure to 2 IU/ml hCG, a similar percentage of wild-type (47%) and ERKO (66%) follicles exhibited successful rupture. However, only 10% of ?ERKO follicles that had reached a diameter of at least 340 μm ovulated in response to hCG, a significant reduction (P < 0.0001) relative to their wild-type (53%) counterparts. Increased hCG doses ranging from 4–32 IU/ml did not improve the percentage of ovulating follicles among wild-type, ERKO, or ?ERKO follicles. Regardless of genotype, released oocytes were consistently surrounded by expanded, mucified cumulus cells and appeared to resume meiosis as evident by germinal vesicle breakdown and occasionally polar body extrusion. No additional evaluation of oocyte competence was performed. Luteinization of the follicle after hCG exposure was monitored by assessing the release of progesterone into the culture medium 24 h after hCG treatment. Before the hCG stimulus, progesterone secretion by large antral follicles was low in all genotypes but rose rapidly by 24 h after the induction of ovulation (Table 5).
TABLE 4. Percentage of follicles ovulating in vitro
TABLE 5. Progesterone secretion of follicles in vitro upon hCG stimulus
To extend the above findings, we selected genes known to show coordinated expression during antral growth and ovulation, including Lhcgr, Cyp19, Ar, Timp1, Cdkn1b, Pgr, Ptgs2, and Has2 for evaluation of their expression patterns in wild-type, ERKO, and ?ERKO follicles grown in vitro. Semiquantitative RT-PCR analysis for these genes was performed on total mRNA from large antral follicles that reached a diameter of at least 340 μm and were collected just before or 3 or 24 h after incubation with hCG (2 IU/ml). Because some of selected genes show a rapid but transient in response to the LH surge (37), collecting follicles at the indicated time points enabled us to determine whether this rapid changes also occur in follicles of all three genotypes.
Representative blots for each transcript and bar graphs of those transcripts that showed differences between genotypes are shown in Fig. 3. Gene expression assays from ERKO follicles indicate no significant differences in the temporal pattern or level of expression for Lhcgr, Cyp19, Ar, Cdkn1b, Ptgs2, and Has2 when compared with similarly treated wild-type follicles (Fig. 3A). ERKO follicles did have reduced levels of Pgr expression just before and 24 h after hCG (P < 0.01); however, the expected hCG induction of Pgr mRNA expression occurred normally in ERKO follicles. Moreover, ERKO follicles showed a reduction in Timp1 expression before the hCG pulse (P < 0.05).
FIG. 3. Analysis of Lhcgr, Cyp19, Ar, Timp1, Cdkn1b, Pgr, Ptgs2, and Has2 mRNA expression patterns in follicles from wild-type (WT), ERKO, and ?ERKO ovaries. Expression was measured in follicles grown in vitro that exceeded a diameter of 340 μm and were collected immediately (no hCG) or 3 or 24 h after incubation with hCG (2 IU/ml). A, PhosphorImage analysis of representative RT-PCR assays for Lhcgr, Cyp19, Ar, Timp1, Cdkn1b, Pgr, Ptgs2, and Has2 mRNA in WT, ?ERKO, and ERKO follicles at indicated time points (no hCG, 3 h hCG, and 24 h hCG). Actb mRNA analysis was included as control. B, Bar graphs for Lhcgr, Cyp19, Timp1, and Ptgs2 mRNA levels in WT, ERKO, and ?ERKO follicles. Expression levels have been normalized to Actb expression levels and expressed as fold induction over the WT no-hCG value. Data are expressed as the mean ± SEM (three to six follicles per data point in WT vs. ERKO graphs; three to13 follicles per data point in WT vs. ?ERKO graphs). Asterisk denotes statistical significance between genotypes at a time point: *, P < 0.05 vs. respective WT value; **, P < 0.01 vs. respective WT value. Letters denote statistical significance between time points within a genotype; a, P < 0.05 vs. no-hCG value; b, P < 0.01 vs. no-hCG value; c, P < 0.05 vs. 3-h hCG value; d, P < 0.01 vs. 3-h hCG value.
Gene expression assays in ?ERKO follicles indicated no significant differences in the temporal expression pattern or level of expression for Ar, Cdkn1b, Pgr,and Has2 when compared with similarly treated wild-type follicles (Fig. 3A). In contrast, Cyp19 mRNA levels were reduced relative to wild-type follicles of similar size just before the hCG pulse (P < 0.01). Lhcgr mRNA levels in ?ERKO follicles differed from wild-type levels 3 and 24 h after hCG (P < 0.05). The dramatic and transient induction of Ptgs2 by hCG in wild-type follicles (>200-fold) was significantly reduced in ?ERKO follicles (P < 0.05). Moreover, ?ERKO follicles showed a reduction in Timp1 expression before and 3 h after the hCG pulse (P < 0.05).
Exogenous PGE2 does not restore ovulation in ?ERKO follicles
Davis et al. (38) showed that Ptgs2-null mice have a reduced ovulation rate that could be restored to normal by treatment with exogenous PGE2. Therefore, the effect of PGE2 on ovulation rates of wild-type and ?ERKO follicles in vitro was examined. For this, large antral follicles (diameter 340 μm) were incubated with PGE2 (500 ng/ml) 4 h after hCG administration. As shown in Fig. 4, PGE2 had no effect on the ovulation rates of either wild-type follicles or ?ERKO follicles, but histological analysis of unruptured follicles treated with a combination of hCG and PGE2 indicated the presence of an expanded COC and thinning of the follicle wall. The majority of wild-type follicles showed this thinning and expansion after treatment with hCG only. ?ERKO follicles, however, often failed to undergo this thinning and expansion in response to hCG only but did show it in response to hCG combined with PGE2.
FIG. 4. Effect of PGE2 on in vitro ovulation of wild-type and ?ERKO follicles. A, Percentage of follicles with a diameter of at least 340 μm that ovulated within 16 h in response to treatment with hCG (2 IU/ml) alone or hCG plus PGE2 (500 ng/ml PGE2 added 4 h after hCG). *, P < 0.001 vs. respective wild-type value. The number of follicles used for each group is shown in parentheses. B–G, Histological analysis of follicles cultured for 5 d. B, Wild-type follicle; C, ?ERKO follicle; D, nonovulated wild-type follicle 16 h after stimulation with hCG (2 IU/ml) (note the expanded COC and thinning of the follicle wall); E, nonovulated ?ERKO follicle 16 h after hCG (2 IU/ml) (less than robust COC expansion and thinning of the wall compared with wild-type follicle); F, nonovulated wild-type follicle treated with hCG plus PGE2, 16 h after hCG pulse; G, nonovulated ?ERKO follicle treated with hCG plus PGE2, 16 h after hCG pulse showing similar COC expansion and thinning of the follicle wall. Scale bar, 100 μm. These data are representative of results obtained from analyzing at least five follicles per experimental group per genotype.
Discussion
In the present study, we have used the method of in vitro follicle culture to gain further insight into the intraovarian role of ER and ER? in follicle growth, ovulation, and luteinization. ?ERKO females exhibited an abnormal growth pattern, decreased estradiol production, and an attenuated response to an ovulatory dose of hCG, whereas follicles from ERKO ovaries exhibited wild-type-like growth and ovulatory efficiency but an elevated capacity to synthesize estradiol. These data confirm as well as advance our previous descriptions of ovarian phenotypes in ERKO mice that were derived from in vivo studies. The capacity of ERKO follicles to behave normally when placed in culture, removed from the chronically high levels of LH, estradiol, and androgens that are characteristic of this animal, supports our previous hypothesis that the more dramatic phenotypes in the ERKO ovary are secondary to the loss of ER-mediated negative feedback in the hypothalamic-pituitary axis (23, 29). The reduced efficiency of ?ERKO follicles to ovulate in vitro after hCG exposure corroborates our earlier in vivo superovulation data (4) and confirms that reduced fertility in these animals is primarily a result of the loss of ER? actions in the ovary.
A sizeable percentage of ?ERKO follicles, both in vivo and in vitro, fail to reach maximal size after gonadotropin exposure. The reduced size of mature ?ERKO follicles is not a result of a dramatic decrease in the granulosa cell population, although no quantitative assays for cell number were carried out. Instead, the in vitro growth curve demonstrates that ?ERKO follicles exhibit wild-type-like increases in diameter up to the small antral stage, when granulosa cell proliferation is most rapid and the primary contributor to increased follicle size (19). However, during the later stages of follicular growth when granulosa cell proliferation is slowed but follicle size continues to increase because of antrum formation (19), ?ERKO follicles begin to exhibit retarded growth. The reduced number of large antral follicles in immature ?ERKO ovaries after gonadotropin stimulation observed in vivo corroborates the in vitro findings. Dupont et al. (32) also described follicles with an underdeveloped antrum after gonadotropin treatment in the ovaries of a separate strain of ER?-null mice. Granulosa cell proliferation and antrum development are predominantly under the positive regulation of FSH and estradiol (19, 39). ER? is localized to the granulosa cells of growing follicles and expression peaks during the progression from small to large antral size (24, 25, 28, 40). Therefore, our combined in vitro and in vivo data indicate that ER? may contribute little to granulosa cell proliferation but is important to antrum formation during follicle maturation. This is consistent with the findings of Burns et al. (41) showing that ER? is not a determining factor in the growth of granulosa cell tumors in mice. Because some but not all ?ERKO follicles do progress through the large antral stage, ER? is not absolutely required for antrum formation.
In addition to gonadotropin-stimulated growth of follicles in vitro, we also compared the ability of mature follicles from all three genotypes to respond to a surge of hCG and successfully expel the oocyte in vitro as well as demonstrate the expected changes in gene expression and steroid production. Ovulation is a tightly coordinated process involving multiple factors and signaling pathways (37). In vivo, ovulation is triggered by a surge of LH upon the ovary, which simultaneously stimulates follicular somatic cells to cease proliferation and terminally differentiate toward a luteinized cell type as well as the processes necessary to rupture the follicle and release the oocyte into the extraovarian environment. ?ERKO follicles that reach a large antral size are less likely to ovulate in response to hCG than wild-type follicles, whereas ERKO follicles showed no deficit. Increasing the dose of hCG could not improve the poor ovulation rate of ?ERKO follicles. Although most ?ERKO follicles failed to rupture in response to hCG, the somatic cells of the follicles did luteinize as shown by the rapid increase in progesterone secretion regardless of ovulatory success and increased expression of Cdkn1b, a gene that is known to be increased in luteinized granulosa cells (42). And although initially decreased in ?ERKO and ERKO follicles, TIMP-1, another gene that is likely to play a role in corpus luteum formation (43), is properly up-regulated 24 h after hCG in ERKO follicles. These findings are consistent with earlier in vivo studies demonstrating that hCG-stimulated ?ERKO ovaries possessed corpora lutea and luteinized follicles that still contained trapped oocytes (4, 32). Therefore, we conclude that ER? facilitates follicle rupture but not luteinization.
Ovulation is preceded by rapid and transient increases in gene expression in the follicle, including several genes whose products are obligatory to follicle rupture via the study of gene-targeted mice, including Pgr (44) and Ptgs2 (38). HAS2 (Has2), the enzyme necessary for synthesis of glycosaminoglycan hyaluronan, the main component of the expanded COC, is also rapidly induced by LH just before ovulation (45). To determine whether the abnormal failure rate among ?ERKO follicles to ovulate was because of a lack of necessary gene expression, we evaluated the transient expression of Pgr, Ptgs2, and Has2 mRNA in wild-type and ERKO follicles in vitro. Whereas Pgr and Has2 induction by hCG was similar in wild-type, ERKO, and ?ERKO follicles, hCG-induced Ptgs2 levels among ?ERKO follicles were significantly reduced. Ptgs2 encodes the enzyme prostaglandin-endoperoxide synthase 2 necessary for synthesis of prostaglandins, including PGE2, which are critical to COC expansion and follicle rupture (38). The less than robust COC expansion observed herein among ?ERKO follicles in vitro correlates with a lack of sufficient Ptgs2 induction. In addition, we and others have previously reported impaired mucification of the COC among ?ERKO oocytes in vivo after superovulation of ER?-null females (4, 32). Therefore, the reduced induction of Ptgs2 after hCG exposure in ?ERKO follicles may be a causal factor in reduced ovulation rates observed both in vitro and in vivo. The ovulatory defect in Ptgs2-null mice can be restored by treatment with PGE2 during superovulation (38). Herein, similar treatment of ?ERKO follicles in vitro improved expansion of the COC and thinning of the follicle wall, but no improvement of follicle rupture was observed. This suggests that ?ERKO follicles may have reduced levels of PGE2 but intact prostaglandin-receptor signaling pathways. However, no quantitative assays for PGE2 content or PGE2 target genes were carried out to elucidate this.
Although Ptgs2 was the only hCG-regulated gene evaluated herein to show a reduced response, we did observe that preovulatory ?ERKO follicles in vitro exhibit deficits in granulosa cell differentiation, as indicated by a significant reduction in preovulatory level of Cyp19, as was estradiol output. Cyp19 expression in granulosa cells of large antral follicles is primarily dependent on FSH, but this requires estradiol (11). A similar role for estradiol has been shown to occur in the FSH regulation of Lhcgr mRNA expression (14, 15, 16), which was slightly reduced in preovulatory ?ERKO follicles in vitro but within normal range. These data suggest that the long-known synergistic effect of estradiol on FSH actions may be mediated by ER?. Additional support for this hypothesis comes from Boa et al. (40) and their report of increased ER? expression just before initiation of Cyp19 and Lhcgr mRNA expression in granulosa cells of preovulatory follicles. Another indication of a failure of ER?-null granulosa cells to fully differentiate is reports of elevated androgen receptor expression in antral ?ERKO follicles (46), which is normally reduced in differentiated granulosa cells (10). However, we were unable to verify this abnormality in individual ?ERKO follicles in vitro.
Interestingly, in vitro cultured ER-null follicles behaved similarly to wild-type follicles among all of the parameters evaluated with the exception of estradiol synthesis. Thus, ER has a minimal intrafollicular role in terms of regulating growth and ovulatory capacity of the individual follicle. However, the increased estradiol output by individual ERKO follicles when placed in a controlled hormonal environment was surprising and in contrast to our earlier hypotheses (23). We have previously shown that by reducing the chronically high plasma LH levels in the ERKO female via a GnRH antagonist, several aspects of the ERKO ovarian phenotype are prevented, including the increased Cyp19 expression and estradiol synthesis (23, 29). Although the present in vitro data further support the in vivo data in terms of rescuing ERKO follicle growth when placed in a more physiological hormonal milieu, the continued elevated level of estradiol secretion was puzzling. No increase in level of Cyp19 mRNA was observed between wild-type and ERKO follicles. However, the availability of androgen precursors from the theca is considered limiting to estradiol synthesis by granulosa cells in preovulatory follicles (9). We have previously shown that ERKO ovaries have increased Cyp17 mRNA expression, and this may persist in vitro (23). Alternatively, the explanation for this observation may be that the follicles were prepared from 21- to 25-d-old ovaries, at which time ERKO females already show elevated levels of serum LH (29). Exposure of ERKO follicles to higher endogenous LH before removal and culture may enhance the steroidogenic capacity of these follicles when compared with wild type.
In summary, our combined use of in vitro follicle culture methods with ERKO mice has proved to be a powerful tool to expand our knowledge of the role of ER and ER? in follicular development and ovulation. Our in vitro data suggest that folliculogenesis and ovulation can normally occur in the absence of ER but not in the absence of ER?. Several indicators of follicle maturity, e.g. antrum formation as well as acquisition of Cyp19 expression, fail to fully develop in ER?-null follicles. These findings suggest that ER? may not be directly mediating the ovulatory response to the LH surge but instead is most important to propagating the necessary differentiation of the follicle that leads up to the LH surge and successful ovulation.
Acknowledgments
We thank Vickie Walker and Linwood Koonce for management of the ERKO colonies and Wendy Jefferson and Dr. David Schomberg for advice and comments on the manuscript.
References
Couse JF, Korach KS 1999 Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev 20:358–417
Britt KL, Drummond AE, Dyson M, Wreford NG, Jones ME, Simpson ER, Findlay JK 2001 The ovarian phenotype of the aromatase knockout (ArKO) mouse. J Steroid Biochem Mol Biol 79:181–185
Lubahn DB, Moyer JS, Golding TS, Couse JF, Korach KS, Smithies O 1993 Alteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor gene. Proc Natl Acad Sci USA 90:11162–11166
Krege JH, Hodgin JB, Couse JF, Enmark E, Warner M, Mahler JF, Sar M, Korach KS, Gustafsson JA, Smithies O 1998 Generation and reproductive phenotypes of mice lacking estrogen receptor ?. Proc Natl Acad Sci USA 95:15677–15682
Schomberg DW, Couse JF, Mukherjee A, Lubahn DB, Sar M, Mayo KE, Korach KS 1999 Targeted disruption of the estrogen receptor- gene in female mice: characterization of ovarian responses and phenotype in the adult. Endocrinology 140:2733–2744
Green S, Walter P, Kumar V, Krust A, Bornert JM, Argos P, Chambon P 1986 Human oestrogen receptor cDNA: sequence, expression and homology to v-erb-A. Nature 320:134–139
Kuiper GG, Enmark E, Pelto-Huikko M, Nilsson S, Gustafsson JA 1996 Cloning of a novel receptor expressed in rat prostate and ovary. Proc Natl Acad Sci USA 93:5925–5930
Mosselman S, Polman J, Dijkema R 1996 ER?: identification and characterization of a novel human estrogen receptor. FEBS Lett 392:49–53
Gore-Langton RE, Amstrong DT 1994 Follicular steroidogenesis and its control. In: Knobil E, Neil JD, eds. The physiology of reproduction. 2nd ed. New York: Raven Press; 571–617
Tetsuka M, Hillier SG 1996 Androgen receptor gene expression in rat granulosa cells: the role of follicle-stimulating hormone and steroid hormones. Endocrinology 137:4392–4397
Fitzpatrick SL, Richards JS 1991 Regulation of cytochrome P450 aromatase messenger ribonucleic acid and activity by steroids and gonadotropins in rat granulosa cells. Endocrinology 129:1452–1462
Zhuang LZ, Adashi EY, Hsuch AJ 1982 Direct enhancement of gonadotropin-stimulated ovarian estrogen biosynthesis by estrogen and clomiphene citrate. Endocrinology 110:2219–2221
Adashi EY, Hsueh AJ 1982 Estrogens augment the stimulation of ovarian aromatase activity by follicle-stimulating hormone in cultured rat granulosa cells. J Biol Chem 257:6077–6083
Segaloff DL, Wang HY, Richards JS 1990 Hormonal regulation of luteinizing hormone/chorionic gonadotropin receptor mRNA in rat ovarian cells during follicular development and luteinization. Mol Endocrinol 4:1856–1865
Kessel B, Liu YX, Jia XC, Hsueh AJ 1985 Autocrine role of estrogens in the augmentation of luteinizing hormone receptor formation in cultured rat granulosa cells. Biol Reprod 32:1038–1050
Knecht M, Brodie AM, Catt KJ 1985 Aromatase inhibitors prevent granulosa cell differentiation: an obligatory role for estrogens in luteinizing hormone receptor expression. Endocrinology 117:1156–1161
Goldenberg RL, Vaitukaitis JL, Ross GT 1972 Estrogen and follicle stimulation hormone interactions on follicle growth in rats. Endocrinology 90:1492–1498
Wang XN, Greenwald GS 1993 Synergistic effects of steroids with FSH on folliculogenesis, steroidogenesis and FSH- and hCG-receptors in hypophysectomized mice. J Reprod Fertil 99:403–413
Hirshfield AN 1991 Development of follicles in the mammalian ovary. Int Rev Cytol 124:43–101
Burghardt RC, Anderson E 1981 Hormonal modulation of gap junctions in rat ovarian follicles. Cell Tissue Res 214:181–193
Billig H, Furuta I, Hsueh AJ 1993 Estrogens inhibit and androgens enhance ovarian granulosa cell apoptosis. Endocrinology 133:2204–2212
Banks PK, Meyer K, Brodie AM 1991 Regulation of ovarian steroid biosynthesis by estrogen during proestrus in the rat. Endocrinology 129:1295–12304
Couse JF, Yates MM, Walker VR, Korach KS 2003 Characterization of the hypothalamic-pituitary-gonadal axis in estrogen receptor (ER) null mice reveals hypergonadism and endocrine sex reversal in females lacking ER but not ER?. Mol Endocrinol 17:1039–1053
Byers M, Kuiper GG, Gustafsson JA, Park-Sarge OK 1997 Estrogen receptor-? mRNA expression in rat ovary: down-regulation by gonadotropins. Mol Endocrinol 11:172–182
Sharma SC, Clemens JW, Pisarska MD, Richards JS 1999 Expression and function of estrogen receptor subtypes in granulosa cells: regulation by estradiol and forskolin. Endocrinology 140:4320–4334
Fitzpatrick SL, Funkhouser JM, Sindoni DM, Stevis PE, Deecher DC, Bapat AR, Merchenthaler I, Frail DE 1999 Expression of estrogen receptor-? protein in rodent ovary. Endocrinology 140:2581–2591
Tetsuka M, Milne M, Hillier SG 1998 Expression of oestrogen receptor isoforms in relation to enzymes of oestrogen synthesis in rat ovary. Mol Cell Endocrinol 141:29–35
Sar M, Welsch F 1999 Differential expression of estrogen receptor-? and estrogen receptor- in the rat ovary. Endocrinology 140:963–971
Couse JF, Bunch DO, Lindzey J, Schomberg DW, Korach KS 1999 Prevention of the polycystic ovarian phenotype and characterization of ovulatory capacity in the estrogen receptor- knockout mouse. Endocrinology 140:5855–5865
Huynh K, Jones G, Thouas G, Britt KL, Simpson ER, Jones ME 2004 Estrogen is not directly required for oocyte developmental competence. Biol Reprod 70:1263–1269
Rosenfeld CS, Murray AA, Simmer G, Hufford MG, Smith MF, Spears N, Lubahn DB 2000 Gonadotropin induction of ovulation and corpus luteum formation in young estrogen receptor- knockout mice. Biol Reprod 62:599–605
Dupont S, Krust A, Gansmuller A, Dierich A, Chambon P, Mark M 2000 Effect of single and compound knockouts of estrogen receptors (ER) and ? (ER?) on mouse reproductive phenotypes. Development 127:4277–4291
Nayudu PL, Osborn SM 1992 Factors influencing the rate of preantral and antral growth of mouse ovarian follicles in vitro. J Reprod Fertil 95:349–362
Rose UM, Hanssen RG, Kloosterboer HJ 1999 Development and characterization of an in vitro ovulation model using mouse ovarian follicles. Biol Reprod 61:503–511
Wu J, Nayudu PL, Kiesel PS, Michelmann HW 2000 Luteinizing hormone has a stage-limited effect on preantral follicle development in vitro. Biol Reprod 63:320–327
Ochsner SA, Day AJ, Rugg MS, Breyer RM, Gomer RH, Richards JS 2003 Disrupted function of tumor necrosis factor--stimulated gene 6 blocks cumulus cell-oocyte complex expansion. Endocrinology 144:4376–4384
Richards JS, Russell DL, Ochsner S, Hsieh M, Doyle KH, Falender AE, Lo YK, Sharma SC 2002 Novel signaling pathways that control ovarian follicular development, ovulation, and luteinization. Recent Prog Horm Res 57:195–220
Davis BJ, Lennard DE, Lee CA, Tiano HF, Morham SG, Wetsel WC, Langenbach R 1999 Anovulation in cyclooxygenase-2-deficient mice is restored by prostaglandin E2 and interleukin-1?. Endocrinology 140:2685–2695
Rao MC, Midgley Jr AR, Richards JS 1978 Hormonal regulation of ovarian cellular proliferation. Cell 14:71–78
Bao B, Kumar N, Karp RM, Garverick HA, Sundaram K 2000 Estrogen receptor-? expression in relation to the expression of luteinizing hormone receptor and cytochrome P450 enzymes in rat ovarian follicles. Biol Reprod 63:1747–1755
Burns KH, Agno JE, Chen L, Haupt B, Ogbonna SC, Korach KS, Matzuk MM 2003 Sexually dimorphic roles of steroid hormone receptor signaling in gonadal tumorigenesis. Mol Endocrinol 17:2039–2052
Robker RL, Richards JS 1998 Hormone-induced proliferation and differentiation of granulosa cells: a coordinated balance of the cell cycle regulators cyclin D2 and p27Kip1. Mol Endocrinol 12:924–940
Curry Jr TE, Osteen KG 2001 Cyclic changes in the matrix metalloproteinase system in the ovary and uterus. Biol Reprod 64:1285–1296
Lydon JP, DeMayo FJ, Funk CR, Mani SK, Hughes AR, Montgomery CA, Shyamala G, Conneely OM, O’Malley BW 1995 Mice lacking progesterone receptor exhibit pleiotropic reproductive abnormalities. Genes Dev 9:2266–2278
Ochsner SA, Russell DL, Day AJ, Breyer RM, Richards JS 2003 Decreased expression of tumor necrosis factor--stimulated gene 6 in cumulus cells of the cyclooxygenase-2 and EP2 null mice. Endocrinology 144:1008–1019
Cheng G, Weihua Z, Makinen S, Makela S, Saji S, Warner M, Gustafsson JA, Hovatta O 2002 A role for the androgen receptor in follicular atresia of estrogen receptor ? knockout mouse ovary. Biol Reprod 66:77–84(Judith M. A. Emmen, John )
Address all correspondence and requests for reprints to: Dr. Kenneth S. Korach, Receptor Biology Section, Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences/National Institutes of Health, MD B3-02, P.O. Box 12233, Research Triangle Park, North Carolina 27709. E-mail: korach@niehs.nih.gov.
Abstract
Both estrogen receptor (ER) and ? are expressed within the ovary and lack of either of these receptors affects ovarian function. In this study, the role of ER and ER? in folliculogenesis and ovulation was further analyzed. Evaluation of ovarian follicle populations in wild-type and ER? knockout (?ERKO) ovaries revealed reduced late antral growth and ovulatory capacity of ?ERKO follicles, indicated by reduced numbers of large antral follicles and corpora lutea and increased atresia of large antral follicles. An in vitro culture system was used to study growth, rupture, and luteinization of wild-type, ER knockout (ERKO) and ?ERKO ovarian follicles. ERKO follicles exhibited wild-type-like growth and ovulation rates but an increased capacity to synthesize estradiol. In contrast, ?ERKO follicles showed a significant lack of progression from early antral to large antral stage, decreased estradiol production, and reduced ovulation. Expression patterns of several genes involved in follicle maturation and ovulation were analyzed in follicles grown in vitro. Ar, Pgr, and Has2 mRNA expression levels were the same among the three genotypes. However, ?ERKO follicles showed reduced expression of Cyp19 mRNA during follicle maturation and reduced Lhcgr and Ptgs2 mRNA expression after human chorionic gonadotropin stimulus. Luteinization occurs normally in ERKO and ?ERKO follicles, shown by increased progesterone secretion and increased cdkn1b mRNA expression after human chorionic gonadotropin. Collectively, these data indicate that ER?, but not ER, plays a direct role in folliculogenesis. ER? appears to facilitate follicle maturation from the early antral to the preovulatory stage.
Introduction
ESTROGENS PLAY A KEY role in female reproduction, with the ovary serving as both the primary site of estrogen synthesis and a major estrogen target. The importance of estrogen actions in the ovary have recently been illustrated by the phenotypes of mice lacking functional estrogen receptors (ER) and/or ? [ER knockout (ERKO)] and those lacking the capacity to produce estradiol, the Cyp19 [aromatase knockout (ArKO)] null mice (1, 2). Adult ERKO and ArKO females are infertile and exhibit ovaries that possess large, hemorrhagic and cystic follicles and no evidence of spontaneous ovulation, whereas adult ?ERKO females are subfertile and possess ovaries that are best characterized as exhibiting a paucity of growing follicles and corpora lutea (2, 3, 4, 5). These findings, along with the expression pattern of each nuclear ER form within the ovary suggest a critical role for intraovarian estrogen signaling in ovarian function (6, 7, 8).
In the rodent ovary, estrogen is primarily produced by preovulatory follicles under the influence of FSH (9). The endocrine actions of estrogen in the female reproductive tract and hypothalamic-pituitary axis are well documented and critical to reproduction (1). However, estrogen is also thought to fulfill a number of intraovarian roles during folliculogenesis. For example, in granulosa cells of maturing follicles, estradiol has little effect alone but is required for maximum FSH stimulation of 1) Cyp19 expression and estradiol synthesis (10, 11, 12, 13); 2) LH receptor (Lhcgr) expression and LH responsiveness (14, 15, 16); 3) antrum formation (17, 18, 19); 4) gap-junction formation (20); and 5) prevention of atresia (21). In thecal cells of growing follicles, estradiol modulates androgen synthesis by negatively regulating Cyp17 expression in thecal cells (22, 23).
Estrogens signal via two nuclear estrogen receptors, ER and ER?, and both are present in the ovary but in different cell types. Granulosa cells of follicles from the primary to large antral stage abundantly express ER?-encoding transcripts and protein, whereas expression is low or absent in interstitial and luteal cells (7, 24, 25). In contrast, ER-encoding transcripts and protein are predominantly localized to thecal and interstitial cells with little or no expression in granulosa cells (24, 26, 27, 28). Whether oocytes express ER remains controversial; however, because oocytes from both ER null mice and ArKO mice are able to undergo fertilization and yield viable embryos, neither ER nor estrogen appears critical to oocyte integrity (4, 29, 30). The divergent pattern of expression of ER and ER? in the ovary and the distinct ovarian phenotypes exhibited by ERKO and ?ERKO females indicate that the two ER forms have distinct roles in ovarian function. Although ERKO ovaries contain normal follicles up to the small antral stage, the predominant features are multiple, large hemorrhagic follicles and the absence of corpora lutea, indicating an anovulatory phenotype (5, 31). We have previously shown that this dramatic phenotype in the ERKO ovary is secondary to chronically elevated levels of plasma LH that result from the loss of ER-mediated negative feedback in the hypothalamus (23, 29). In contrast, ?ERKO female mice exhibit normal levels of circulating gonadotropins and ovaries that possess all stages of folliculogenesis, including corpora lutea (4, 23). However, when bred, ?ERKO female mice yield significantly fewer litters and offspring compared with wild-type counterparts (4). Furthermore, the oocyte yield from ?ERKO female mice after gonadotropin-induced ovulation is approximately 15% that of wild-type females (4, 32). These data suggest that ER? plays a more direct role in ovarian function, although the exact mechanism has yet to be determined.
Detailed studies of the intraovarian events during follicular development and ovulation are difficult to carry out in vivo because of the various endocrine factors known to influence ovarian function, including gonadotropins, prolactin, steroids, and various growth factors. This is especially true for the ERKO mouse because ovaries of these mice exist in a severely abnormal hormonal environment (23). To gain further insight into the role(s) of ER and ER? in ovarian function we have established an in vitro culture system using follicles from ERKO and ?ERKO ovaries. This system allows us to compare growth, steroidogenesis, and ovulation rates between wild-type, ERKO, and ?ERKO follicles in a controlled hormonal environment. In addition, expression patterns of several genes known to be involved in follicular maturation, ovulation, or luteinization were evaluated in follicles grown in vitro. Using in vitro culture of follicles, we have shown that ER?, and not ER, is most critical to follicle growth and maturation leading up ovulation.
Materials and Methods
Animals
All animal procedures were approved by the National Institute of Environmental Health Sciences Animal Care and Use Committee. Generation of ERKO and ?ERKO null mice has been previously described by Lubahn et al. (3) and Krege et al. (4), respectively. Mice were generated from ERKO and ?ERKO heterozygous breeding pairs on a background of C57BL6 obtained from Taconic Farms (Germantown, NY). Offspring were genotyped by PCR analysis of tail-snip genomic DNA as previously described (23).
Ovarian follicle counts in vivo
Ovaries were collected from immature (28–30 d) wild-type and ?ERKO female mice 48 h after a single injection of 2.2 IU of pregnant mare serum gonadotropin (PMSG) (Sigma Chemical Co., St. Louis, MO) or vehicle (0.85% saline). In addition, ovaries were collected from adult (109 d) wild-type and ?ERKO female mice that were untreated. The ovaries were fixed in 10% neutral buffered formalin, paraffin embedded, sectioned (5 μm), and stained with hematoxylin and eosin. One ovary per animal was used for follicle counts. Follicles were counted in every 10th section, beginning with the fifth section, through the entire ovary. Only follicles with a visible nucleus were counted to avoid double counting. Follicles were categorized as primordial if they contained an intact oocyte surrounded by a single layer of predominantly squamoid, flattened epithelial cells, lacking a connective tissue or thecal investment. Follicles were categorized as primary if they had an enlarged oocyte surrounded by a single layer of predominantly cuboidal granulosa cells. Small preantral follicles were identified as having an oocyte surrounded by two layers of granulosa cells, whereas large preantral follicles were identified as having an oocyte surrounded by a multilayered (more than two), solid mantle of granulosa cells. Follicles were categorized as small antral if they had a central oocyte surrounded by layered granulosa cells and containing fluid-filled areas within the granulosa cell layer. Large antral follicles were identified as having a single cavity containing follicle fluid surrounded by a granulosa cell layer of even thickness and a cumulus-cell enclosed oocyte. In addition, the total number of corpora lutea was counted. Atretic follicles, defined as follicles containing one or more apoptotic cells with pyknotic nuclei, were not included in the above categories but counted and categorized separately. The presence of apoptotic cells was verified using the terminal deoxynucleotidyl transferase biotin-dUTP nick end labeling assay (Upstate Biotechnology, Charlottesville, VA) according to the manufacturer’s protocol. For each ovary, total number of follicles (per follicle class) present in the scored sections was normalized to the total number of scored sections from that ovary to obtain a mean follicle count per section. The average number of scored sections per ovary was the same between ?ERKO animals and the respective wild-type controls.
No data on follicle counts were collected from ERKO ovaries, because these ovaries exhibit extensive formation of hemorrhagic cystic follicles that distort the overall morphology and make it difficult to obtain representative follicle counts.
In vitro culture of follicles
Follicles were isolated from whole ovaries of 21- to 25-d-old female mice (wild-type, ERKO, and ?ERKO) according to the method of Nayudu and Osborne (33). In brief, mice were killed by CO2 asphyxiation, and the ovaries were removed and placed in Leibovitz’s L-15 medium (Life Technologies, Inc., Invitrogen, Carlsbad, CA) supplemented with insulin (5 μg/ml; Life Technologies), transferrin (10 μg/ml; Sigma), selenium (2 ng/ml; Sigma), ascorbic acid (50 μg/ml; Sigma), and 0.3% BSA (Sigma) at 37 C. Large preantral follicles (200 ± 20 μm in diameter) were then mechanically dissected from the ovaries using 25-gauge syringe needles. Isolated follicles were transferred to four-well multidishes (Nalge, Nunc International, Rochester, NY) containing -MEM (Life Technologies) supplemented with insulin, transferrin, selenium, and ascorbic acid as above as well as penicillin (100 IU/ml; Life Technologies), streptomycin (100 μg/ml; Life Technologies), 5% fetal bovine serum (Sigma), and 100 mIU recombinant human FSH (kindly donated by Ares Serono, Geneva, Switzerland). The culture dishes were placed in a humidified incubator at 37 C in 5% CO2 in air until a sufficient number of follicles had been collected.
The harvested preantral follicles were evaluated, and those exhibiting an intact basement membrane, a high density of granulosa cells, a centrally located oocyte, and some thecal cells attached were selected for further culture. The culture method used for in vitro evaluation of follicle growth and androstenedione, testosterone, and estradiol secretion was based on that of Nayudu and Osborne (33) with modifications according to Rose et al. (34). Follicles were individually cultured in Millipore CM culture plate inserts (Millipore, Bedford, MA) filled with 250 μl -MEM supplemented as described above, at 37 C in 5% CO2 in air. Medium was changed on culture d 1, 3, and 5. Spent medium collected on culture d 5 was stored at –80 C until assayed. Follicular growth was monitored daily by measuring follicle diameter using a precalibrated ocular micrometer. Follicles that did not survive the 5-d culture period, those with obvious darkening of the granulosa cell mass and/or signs of disintegration or that had spontaneously extruded their oocyte, were not included in the analysis.
For in vitro ovulation and progesterone production assays, follicles were cultured in Costar Ultra Low flat-bottom 96-well plates (35). After isolation and selection, follicles were transferred individually into wells containing 40 μl -MEM, supplemented as described above. The culture dishes were placed in a humidified incubator at 37 C in 5% CO2 in air. Medium was changed every 48 h. When follicles reached a diameter of at least 340 μm, after approximately 4 or 5 d of culture, the medium was replaced by medium containing human chorionic gonadotropin (hCG) (2–32 IU/ml) (kindly donated by Ares Serono, Geneva, Switzerland) to induce ovulation. Follicles were checked for ovulation at 16 h after hCG stimulus and were considered to have ovulated when the follicle was visibly ruptured and the cumulus oocyte complex (COC) extruded.
The size at which large antral follicles were exposed to hCG was determined as follows. Large antral wild-type follicles were divided into three groups: follicles that had reached a diameter of 320–340 μm, 340–360 μm, or more than 360 μm. After treatment with 2 IU/ml hCG, 35% of follicles with a diameter of 320–340 μm, 60% of follicles with a diameter between 340 and 360 μm, and 70% of follicles with a size greater than 360 μm ovulated. Initial follicle growth studies indicated a reduction in the number of ?ERKO follicles reaching a size of 340 μm or more. To collect sufficient late antral ?ERKO follicles for the ovulation assays, these assays were performed with follicles that had reached a diameter of at least 340 μm.
In some of the cultures, the effect of prostaglandin on ovulation was evaluated by in vitro ovulation assays and by histology. Prostaglandin E2 (PGE2) (500 ng/ml) (36) dissolved in ethanol was added to the medium 4 h after the hCG pulse, and follicles were checked for ovulation 16 h later. For histological analysis, follicles were collected before hCG or 16 h after incubation with hCG alone or a combination of hCG and PGE2, fixed in 10% neutral buffered formalin, embedded in paraffin, sectioned (5 μm), and stained with hematoxylin and eosin. At least five follicles per experimental group per genotype were studied, and representative photomicrographs are presented for histology.
RIA
Androstenedione, testosterone, and estradiol were measured in spent medium collected on culture d 5, after incubation for 48 h. Androstenedione and testosterone were assayed in a single assay using direct RIA kits from Diagnostics Systems Laboratories (Webster, TX), according to the manufacturer’s protocol. The intraassay coefficients of variation were 4.3% for androstenedione and 8.5% for testosterone, and the assay sensitivities were 0.02 ng/ml for androstenedione and 0.08 ng/ml for testosterone. Estradiol was measured using a double-antibody RIA kit (Diagnostics Systems Laboratories) with inter- and intraassay variations of 8.5 and 4.6%, respectively, and a sensitivity of 4.7 pg/ml. Progesterone production was measured in spent medium collected just before or 24 h after hCG using a direct RIA kit (Diagnostics Systems Laboratories). The inter- and intraassay coefficients of variation were 11.7 and 6.6%, respectively, and the assay sensitivity was 0.12 ng/ml. For each hormone, the level in medium that had not been in contact with follicles was below the level of detection.
Analysis of mRNA expression
Follicles that reached a diameter of at least 340 μm were collected either immediately (no hCG) or after incubation with 2 IU/ml hCG for 3 or 24 h. Total RNA was extracted from individual follicles using PicoPure RNA isolation kit (Arcturus, Mountain View, CA) according to the manufacturer’s protocol. To remove any potential genomic DNA contamination, samples were incubated with 2 μl RNase-free DNase I (2 U/μl) (Ambion Inc., Austin, TX) for 15 min at room temperature. Semiquantitative RT-PCR was performed as described previously (23) with some modifications. Total RNA (0.1 μg) was reverse transcribed in a total volume of 25 μl using the Superscript first-strand synthesis system (Invitrogen Life Technologies Inc.). PCR amplification was carried out with 1 μl of the RT reaction in a 15-μl reaction at 95 C for 30 sec, 58 C for 45 sec, and 72 C for 30 sec for 28 cycles. All samples were electrophoresed on agarose gels containing ethidium bromide and electroblotted to BrightStar nylon membrane (Ambion). The membranes were hybridized with 33P-labeled probes and washed, and results were quantified by using a PhosphorImager and ImageQuant software (Molecular Dynamics, Inc., Sunnyvale, CA). Per experiment, the intensity values for each transcript were normalized to the respective expression level of Actb, a housekeeping gene that is nondifferentially expressed, and then expressed as fold induction over the wild-type no-hCG value by dividing each value by the wild-type no-hCG mean value. The primers and probes used for the semiquantitative RT-PCR are listed in Table 1.
TABLE 1. Primer pairs and probes used for semiquantitative RT-PCR
Statistical analysis
Within the follicle count studies, differences between the mean numbers of follicles per section were evaluated for each follicle class per genotype using the nonparametric Mann-Whitney U test. Data of ERKO and ?ERKO follicles cultured in vitro obtained in separate experiments were analyzed as separate groups. Each experiment always included follicles derived from wild-type ovaries as controls. Follicle growth in vitro was analyzed by comparison of daily follicular sizes between genotypes using repeated-measures ANOVA followed by two-sample t tests comparing mean size at each day. Differences between genotypes in reaching a large antral size or ovulation rates were assessed by 2 tests. Differences in estradiol secretion among genotypes were identified by nonparametric Kruskal-Wallis ANOVA followed by Mann-Whitney U tests. Comparisons of the secreted amount of progesterone and of the gene expression data between the genotypes were done using the Mann-Whitney U test. All results are presented as mean ± SE of the mean (SEM). Differences were considered significant at P < 0.05.
Results
?ERKO ovaries exhibit a reduced number of large antral follicles and corpora lutea
The follicle distribution in wild-type and ?ERKO ovaries is shown in Fig. 1. Ovaries from untreated 28-d-old wild-type and ?ERKO mice revealed similar numbers of primordial and primary follicles per section (Fig. 1A). Among the pool of growing follicles in immature mice, only the number of large preantral follicles per section in ?ERKO ovaries vs. wild-type ovaries was significantly increased (P < 0.05). Although the total number of atretic follicles per section did not significantly differ between immature wild-type and ?ERKO ovaries (Fig. 1A), there were fewer atretic small antral follicles per section in ?ERKO ovaries compared with wild-type ovaries (P < 0.05).
FIG. 1. Distribution of the follicle populations in wild-type and ?ERKO ovaries. A, Follicle population (healthy and atretic) in ovaries of immature wild-type (WT) and ?ERKO mice 28–30 d of age after treatment with vehicle for 48 h (n = 4 per group). B, Follicle population (healthy and atretic) in ovaries of immatureWT and ?ERKO mice 28–30 d of age after treatment with PMSG for 48 h (n = 7 wild type; n = 9 ?ERKO). C, Follicle population (healthy and atretic) in ovaries of adult WT and ?ERKO approximately 109 d of age (n = 6 per group). Shown is the mean number of follicles per section for each follicle class per genotype (± SEM). *, P < 0.05 vs. respective WT value; **, P < 0.01 vs. respective WT value.
PMSG treatment for 48 h had no marked effect on the pool of primordial or primary follicles in immature ovaries of either genotype (Fig. 1B). There were also no differences in the number of preantral follicles per section between ?ERKO and wild-type PMSG-treated ovaries. Treated ?ERKO ovaries did exhibit an increased number of small antral follicles per section compared with treated wild-type ovaries (P < 0.05). As expected, PMSG treatment led to the presence of large antral follicles that were not observed in ovaries of untreated animals (compare Fig. 1, A with B; P < 0.01, untreated vs. treated wild-type ovaries; P < 0.05, untreated vs. treated ?ERKO ovaries). Interestingly, PMSG-treated ?ERKO females exhibited significantly fewer large antral follicles per section (P < 0.01) compared with treated wild type. No differences were observed in the number of atretic follicles per section, total or per category, between PMSG-treated wild-type and PMSG-treated ?ERKO ovaries (Fig. 1B).
A comparison of follicle counts per section between adult wild-type and ?ERKO ovaries revealed that ?ERKO ovaries contain significantly more primordial (P < 0.01) but significantly fewer primary follicles per section (P < 0.05; Fig. 1C). ?ERKO ovaries also exhibited a significantly reduced number of large preantral follicles and corpora lutea per section (P < 0.05). Although no difference was observed in the total number of atretic follicles between adult wild-type and ?ERKO ovaries, ?ERKO ovaries contained fewer atretic large preantral follicles and more atretic large antral follicles per section than wild-type ovaries.
?ERKO follicles fail to reach maximum size in vitro
Figure 2 shows the growth profiles of in vitro cultured follicles from ERKO (Fig. 2A) and ?ERKO (Fig. 2B) ovaries compared with follicles from wild-type ovaries. At the start of culture (d 0), preantral follicles from wild-type, ?ERKO, and ERKO ovaries exhibited a similar mean diameter, and each continued to grow and exhibit visible antrum formation by d 2. The overall growth profile of ERKO follicles did not differ from their respective wild-type controls (P = 0.08; Fig. 2A), but a significant difference in size was observed between wild-type and ERKO follicles on culture d 1 (P < 0.001) and d 5 (P < 0.01). Nonetheless, the percentage of ERKO follicles (77%) that reached 340 μm did not differ from wild type (91%; Table 2). In contrast, there was a significant difference between the overall growth profiles of wild-type and ?ERKO follicles (P < 0.0001; Fig. 2B) such that ?ERKO follicles exhibited retarded growth that first became apparent on d 2 (P < 0.01). On subsequent days, the difference in follicle diameter between wild-type and ?ERKO follicles became more pronounced (P < 0.001 on d 3, 4, and 5). As a result, the percentage of ?ERKO follicles reaching a large antral size (designated as a diameter 340 μm) after 5 d in culture was significantly lower relative to wild-type, 60 vs. 92%, respectively (P < 0.0001; Table 2).
FIG. 2. In vitro growth of follicles from wild-type (WT), ?ERKO, and ERKO ovaries. Preantral follicles were grown in vitro in the presence of 100 mIU/ml recombinant human FSH. The follicle diameter was measured each culture day. Follicles that exceeded a diameter of 340 μm (dotted line) were classified large antral follicles. A, Growth profiles of WT and ERKO follicles (n = 7–46 follicles per data point). The growth of ERKO follicles was similar to that of WT follicles (F1,218 = 3.1; P = 0.08). There was a significant difference between the size of WT and ERKO follicles on d 1 and 5 of culture. B, Growth profiles of WT and ?ERKO follicles (n = 11–133 follicles per data point). The growth profile of ?ERKO follicles was different from WT follicles (F1,630 = 181.1; P < 0.001). From culture d 2 onward, WT follicles were significantly bigger in diameter than ?ERKO follicles at each culture day. Shown is the mean follicle diameter (μm) ± SEM for each culture day per genotype. *, P < 0.01 vs. WT; **, P < 0.001 vs. WT.
TABLE 2. Percentage of wild-type, ERKO, and ?ERKO follicles that reached a large antral size
The steroidogenic capacity of follicles in culture was assessed by measuring the level of androstenedione, testosterone and estradiol that accumulated in the culture medium over a 48 h interval, between days 3 and 5 of culture. Androstenedione and testosterone in the medium were undetectable. Although we observed a considerable degree of variation in estradiol release among follicles within each genotype, a significant difference among the genotypes was evident when comparing the mean levels (Table 3). Estradiol secretion by ERKO follicles significantly exceeded that of wild-type follicles by at least 4-fold (P < 0.05), whereas ?ERKO follicles exhibited a significantly lower rate of secretion compared with wild-type follicles (P < 0.05).
TABLE 3. Estradiol secretion by follicles in vitro
?ERKO follicles do not fully respond to hCG
Wild-type, ERKO, and ?ERKO follicles that had reached a large antral size of at least 340 μm were induced to ovulate by treatment with hCG and checked for ovulation at 16 h after the hCG stimulus (Table 4). Initially, a dose of 2 IU/ml hCG was used, because this was shown to be the optimal dose to induce ovulation in vitro by Rose et al. (34). After exposure to 2 IU/ml hCG, a similar percentage of wild-type (47%) and ERKO (66%) follicles exhibited successful rupture. However, only 10% of ?ERKO follicles that had reached a diameter of at least 340 μm ovulated in response to hCG, a significant reduction (P < 0.0001) relative to their wild-type (53%) counterparts. Increased hCG doses ranging from 4–32 IU/ml did not improve the percentage of ovulating follicles among wild-type, ERKO, or ?ERKO follicles. Regardless of genotype, released oocytes were consistently surrounded by expanded, mucified cumulus cells and appeared to resume meiosis as evident by germinal vesicle breakdown and occasionally polar body extrusion. No additional evaluation of oocyte competence was performed. Luteinization of the follicle after hCG exposure was monitored by assessing the release of progesterone into the culture medium 24 h after hCG treatment. Before the hCG stimulus, progesterone secretion by large antral follicles was low in all genotypes but rose rapidly by 24 h after the induction of ovulation (Table 5).
TABLE 4. Percentage of follicles ovulating in vitro
TABLE 5. Progesterone secretion of follicles in vitro upon hCG stimulus
To extend the above findings, we selected genes known to show coordinated expression during antral growth and ovulation, including Lhcgr, Cyp19, Ar, Timp1, Cdkn1b, Pgr, Ptgs2, and Has2 for evaluation of their expression patterns in wild-type, ERKO, and ?ERKO follicles grown in vitro. Semiquantitative RT-PCR analysis for these genes was performed on total mRNA from large antral follicles that reached a diameter of at least 340 μm and were collected just before or 3 or 24 h after incubation with hCG (2 IU/ml). Because some of selected genes show a rapid but transient in response to the LH surge (37), collecting follicles at the indicated time points enabled us to determine whether this rapid changes also occur in follicles of all three genotypes.
Representative blots for each transcript and bar graphs of those transcripts that showed differences between genotypes are shown in Fig. 3. Gene expression assays from ERKO follicles indicate no significant differences in the temporal pattern or level of expression for Lhcgr, Cyp19, Ar, Cdkn1b, Ptgs2, and Has2 when compared with similarly treated wild-type follicles (Fig. 3A). ERKO follicles did have reduced levels of Pgr expression just before and 24 h after hCG (P < 0.01); however, the expected hCG induction of Pgr mRNA expression occurred normally in ERKO follicles. Moreover, ERKO follicles showed a reduction in Timp1 expression before the hCG pulse (P < 0.05).
FIG. 3. Analysis of Lhcgr, Cyp19, Ar, Timp1, Cdkn1b, Pgr, Ptgs2, and Has2 mRNA expression patterns in follicles from wild-type (WT), ERKO, and ?ERKO ovaries. Expression was measured in follicles grown in vitro that exceeded a diameter of 340 μm and were collected immediately (no hCG) or 3 or 24 h after incubation with hCG (2 IU/ml). A, PhosphorImage analysis of representative RT-PCR assays for Lhcgr, Cyp19, Ar, Timp1, Cdkn1b, Pgr, Ptgs2, and Has2 mRNA in WT, ?ERKO, and ERKO follicles at indicated time points (no hCG, 3 h hCG, and 24 h hCG). Actb mRNA analysis was included as control. B, Bar graphs for Lhcgr, Cyp19, Timp1, and Ptgs2 mRNA levels in WT, ERKO, and ?ERKO follicles. Expression levels have been normalized to Actb expression levels and expressed as fold induction over the WT no-hCG value. Data are expressed as the mean ± SEM (three to six follicles per data point in WT vs. ERKO graphs; three to13 follicles per data point in WT vs. ?ERKO graphs). Asterisk denotes statistical significance between genotypes at a time point: *, P < 0.05 vs. respective WT value; **, P < 0.01 vs. respective WT value. Letters denote statistical significance between time points within a genotype; a, P < 0.05 vs. no-hCG value; b, P < 0.01 vs. no-hCG value; c, P < 0.05 vs. 3-h hCG value; d, P < 0.01 vs. 3-h hCG value.
Gene expression assays in ?ERKO follicles indicated no significant differences in the temporal expression pattern or level of expression for Ar, Cdkn1b, Pgr,and Has2 when compared with similarly treated wild-type follicles (Fig. 3A). In contrast, Cyp19 mRNA levels were reduced relative to wild-type follicles of similar size just before the hCG pulse (P < 0.01). Lhcgr mRNA levels in ?ERKO follicles differed from wild-type levels 3 and 24 h after hCG (P < 0.05). The dramatic and transient induction of Ptgs2 by hCG in wild-type follicles (>200-fold) was significantly reduced in ?ERKO follicles (P < 0.05). Moreover, ?ERKO follicles showed a reduction in Timp1 expression before and 3 h after the hCG pulse (P < 0.05).
Exogenous PGE2 does not restore ovulation in ?ERKO follicles
Davis et al. (38) showed that Ptgs2-null mice have a reduced ovulation rate that could be restored to normal by treatment with exogenous PGE2. Therefore, the effect of PGE2 on ovulation rates of wild-type and ?ERKO follicles in vitro was examined. For this, large antral follicles (diameter 340 μm) were incubated with PGE2 (500 ng/ml) 4 h after hCG administration. As shown in Fig. 4, PGE2 had no effect on the ovulation rates of either wild-type follicles or ?ERKO follicles, but histological analysis of unruptured follicles treated with a combination of hCG and PGE2 indicated the presence of an expanded COC and thinning of the follicle wall. The majority of wild-type follicles showed this thinning and expansion after treatment with hCG only. ?ERKO follicles, however, often failed to undergo this thinning and expansion in response to hCG only but did show it in response to hCG combined with PGE2.
FIG. 4. Effect of PGE2 on in vitro ovulation of wild-type and ?ERKO follicles. A, Percentage of follicles with a diameter of at least 340 μm that ovulated within 16 h in response to treatment with hCG (2 IU/ml) alone or hCG plus PGE2 (500 ng/ml PGE2 added 4 h after hCG). *, P < 0.001 vs. respective wild-type value. The number of follicles used for each group is shown in parentheses. B–G, Histological analysis of follicles cultured for 5 d. B, Wild-type follicle; C, ?ERKO follicle; D, nonovulated wild-type follicle 16 h after stimulation with hCG (2 IU/ml) (note the expanded COC and thinning of the follicle wall); E, nonovulated ?ERKO follicle 16 h after hCG (2 IU/ml) (less than robust COC expansion and thinning of the wall compared with wild-type follicle); F, nonovulated wild-type follicle treated with hCG plus PGE2, 16 h after hCG pulse; G, nonovulated ?ERKO follicle treated with hCG plus PGE2, 16 h after hCG pulse showing similar COC expansion and thinning of the follicle wall. Scale bar, 100 μm. These data are representative of results obtained from analyzing at least five follicles per experimental group per genotype.
Discussion
In the present study, we have used the method of in vitro follicle culture to gain further insight into the intraovarian role of ER and ER? in follicle growth, ovulation, and luteinization. ?ERKO females exhibited an abnormal growth pattern, decreased estradiol production, and an attenuated response to an ovulatory dose of hCG, whereas follicles from ERKO ovaries exhibited wild-type-like growth and ovulatory efficiency but an elevated capacity to synthesize estradiol. These data confirm as well as advance our previous descriptions of ovarian phenotypes in ERKO mice that were derived from in vivo studies. The capacity of ERKO follicles to behave normally when placed in culture, removed from the chronically high levels of LH, estradiol, and androgens that are characteristic of this animal, supports our previous hypothesis that the more dramatic phenotypes in the ERKO ovary are secondary to the loss of ER-mediated negative feedback in the hypothalamic-pituitary axis (23, 29). The reduced efficiency of ?ERKO follicles to ovulate in vitro after hCG exposure corroborates our earlier in vivo superovulation data (4) and confirms that reduced fertility in these animals is primarily a result of the loss of ER? actions in the ovary.
A sizeable percentage of ?ERKO follicles, both in vivo and in vitro, fail to reach maximal size after gonadotropin exposure. The reduced size of mature ?ERKO follicles is not a result of a dramatic decrease in the granulosa cell population, although no quantitative assays for cell number were carried out. Instead, the in vitro growth curve demonstrates that ?ERKO follicles exhibit wild-type-like increases in diameter up to the small antral stage, when granulosa cell proliferation is most rapid and the primary contributor to increased follicle size (19). However, during the later stages of follicular growth when granulosa cell proliferation is slowed but follicle size continues to increase because of antrum formation (19), ?ERKO follicles begin to exhibit retarded growth. The reduced number of large antral follicles in immature ?ERKO ovaries after gonadotropin stimulation observed in vivo corroborates the in vitro findings. Dupont et al. (32) also described follicles with an underdeveloped antrum after gonadotropin treatment in the ovaries of a separate strain of ER?-null mice. Granulosa cell proliferation and antrum development are predominantly under the positive regulation of FSH and estradiol (19, 39). ER? is localized to the granulosa cells of growing follicles and expression peaks during the progression from small to large antral size (24, 25, 28, 40). Therefore, our combined in vitro and in vivo data indicate that ER? may contribute little to granulosa cell proliferation but is important to antrum formation during follicle maturation. This is consistent with the findings of Burns et al. (41) showing that ER? is not a determining factor in the growth of granulosa cell tumors in mice. Because some but not all ?ERKO follicles do progress through the large antral stage, ER? is not absolutely required for antrum formation.
In addition to gonadotropin-stimulated growth of follicles in vitro, we also compared the ability of mature follicles from all three genotypes to respond to a surge of hCG and successfully expel the oocyte in vitro as well as demonstrate the expected changes in gene expression and steroid production. Ovulation is a tightly coordinated process involving multiple factors and signaling pathways (37). In vivo, ovulation is triggered by a surge of LH upon the ovary, which simultaneously stimulates follicular somatic cells to cease proliferation and terminally differentiate toward a luteinized cell type as well as the processes necessary to rupture the follicle and release the oocyte into the extraovarian environment. ?ERKO follicles that reach a large antral size are less likely to ovulate in response to hCG than wild-type follicles, whereas ERKO follicles showed no deficit. Increasing the dose of hCG could not improve the poor ovulation rate of ?ERKO follicles. Although most ?ERKO follicles failed to rupture in response to hCG, the somatic cells of the follicles did luteinize as shown by the rapid increase in progesterone secretion regardless of ovulatory success and increased expression of Cdkn1b, a gene that is known to be increased in luteinized granulosa cells (42). And although initially decreased in ?ERKO and ERKO follicles, TIMP-1, another gene that is likely to play a role in corpus luteum formation (43), is properly up-regulated 24 h after hCG in ERKO follicles. These findings are consistent with earlier in vivo studies demonstrating that hCG-stimulated ?ERKO ovaries possessed corpora lutea and luteinized follicles that still contained trapped oocytes (4, 32). Therefore, we conclude that ER? facilitates follicle rupture but not luteinization.
Ovulation is preceded by rapid and transient increases in gene expression in the follicle, including several genes whose products are obligatory to follicle rupture via the study of gene-targeted mice, including Pgr (44) and Ptgs2 (38). HAS2 (Has2), the enzyme necessary for synthesis of glycosaminoglycan hyaluronan, the main component of the expanded COC, is also rapidly induced by LH just before ovulation (45). To determine whether the abnormal failure rate among ?ERKO follicles to ovulate was because of a lack of necessary gene expression, we evaluated the transient expression of Pgr, Ptgs2, and Has2 mRNA in wild-type and ERKO follicles in vitro. Whereas Pgr and Has2 induction by hCG was similar in wild-type, ERKO, and ?ERKO follicles, hCG-induced Ptgs2 levels among ?ERKO follicles were significantly reduced. Ptgs2 encodes the enzyme prostaglandin-endoperoxide synthase 2 necessary for synthesis of prostaglandins, including PGE2, which are critical to COC expansion and follicle rupture (38). The less than robust COC expansion observed herein among ?ERKO follicles in vitro correlates with a lack of sufficient Ptgs2 induction. In addition, we and others have previously reported impaired mucification of the COC among ?ERKO oocytes in vivo after superovulation of ER?-null females (4, 32). Therefore, the reduced induction of Ptgs2 after hCG exposure in ?ERKO follicles may be a causal factor in reduced ovulation rates observed both in vitro and in vivo. The ovulatory defect in Ptgs2-null mice can be restored by treatment with PGE2 during superovulation (38). Herein, similar treatment of ?ERKO follicles in vitro improved expansion of the COC and thinning of the follicle wall, but no improvement of follicle rupture was observed. This suggests that ?ERKO follicles may have reduced levels of PGE2 but intact prostaglandin-receptor signaling pathways. However, no quantitative assays for PGE2 content or PGE2 target genes were carried out to elucidate this.
Although Ptgs2 was the only hCG-regulated gene evaluated herein to show a reduced response, we did observe that preovulatory ?ERKO follicles in vitro exhibit deficits in granulosa cell differentiation, as indicated by a significant reduction in preovulatory level of Cyp19, as was estradiol output. Cyp19 expression in granulosa cells of large antral follicles is primarily dependent on FSH, but this requires estradiol (11). A similar role for estradiol has been shown to occur in the FSH regulation of Lhcgr mRNA expression (14, 15, 16), which was slightly reduced in preovulatory ?ERKO follicles in vitro but within normal range. These data suggest that the long-known synergistic effect of estradiol on FSH actions may be mediated by ER?. Additional support for this hypothesis comes from Boa et al. (40) and their report of increased ER? expression just before initiation of Cyp19 and Lhcgr mRNA expression in granulosa cells of preovulatory follicles. Another indication of a failure of ER?-null granulosa cells to fully differentiate is reports of elevated androgen receptor expression in antral ?ERKO follicles (46), which is normally reduced in differentiated granulosa cells (10). However, we were unable to verify this abnormality in individual ?ERKO follicles in vitro.
Interestingly, in vitro cultured ER-null follicles behaved similarly to wild-type follicles among all of the parameters evaluated with the exception of estradiol synthesis. Thus, ER has a minimal intrafollicular role in terms of regulating growth and ovulatory capacity of the individual follicle. However, the increased estradiol output by individual ERKO follicles when placed in a controlled hormonal environment was surprising and in contrast to our earlier hypotheses (23). We have previously shown that by reducing the chronically high plasma LH levels in the ERKO female via a GnRH antagonist, several aspects of the ERKO ovarian phenotype are prevented, including the increased Cyp19 expression and estradiol synthesis (23, 29). Although the present in vitro data further support the in vivo data in terms of rescuing ERKO follicle growth when placed in a more physiological hormonal milieu, the continued elevated level of estradiol secretion was puzzling. No increase in level of Cyp19 mRNA was observed between wild-type and ERKO follicles. However, the availability of androgen precursors from the theca is considered limiting to estradiol synthesis by granulosa cells in preovulatory follicles (9). We have previously shown that ERKO ovaries have increased Cyp17 mRNA expression, and this may persist in vitro (23). Alternatively, the explanation for this observation may be that the follicles were prepared from 21- to 25-d-old ovaries, at which time ERKO females already show elevated levels of serum LH (29). Exposure of ERKO follicles to higher endogenous LH before removal and culture may enhance the steroidogenic capacity of these follicles when compared with wild type.
In summary, our combined use of in vitro follicle culture methods with ERKO mice has proved to be a powerful tool to expand our knowledge of the role of ER and ER? in follicular development and ovulation. Our in vitro data suggest that folliculogenesis and ovulation can normally occur in the absence of ER but not in the absence of ER?. Several indicators of follicle maturity, e.g. antrum formation as well as acquisition of Cyp19 expression, fail to fully develop in ER?-null follicles. These findings suggest that ER? may not be directly mediating the ovulatory response to the LH surge but instead is most important to propagating the necessary differentiation of the follicle that leads up to the LH surge and successful ovulation.
Acknowledgments
We thank Vickie Walker and Linwood Koonce for management of the ERKO colonies and Wendy Jefferson and Dr. David Schomberg for advice and comments on the manuscript.
References
Couse JF, Korach KS 1999 Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev 20:358–417
Britt KL, Drummond AE, Dyson M, Wreford NG, Jones ME, Simpson ER, Findlay JK 2001 The ovarian phenotype of the aromatase knockout (ArKO) mouse. J Steroid Biochem Mol Biol 79:181–185
Lubahn DB, Moyer JS, Golding TS, Couse JF, Korach KS, Smithies O 1993 Alteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor gene. Proc Natl Acad Sci USA 90:11162–11166
Krege JH, Hodgin JB, Couse JF, Enmark E, Warner M, Mahler JF, Sar M, Korach KS, Gustafsson JA, Smithies O 1998 Generation and reproductive phenotypes of mice lacking estrogen receptor ?. Proc Natl Acad Sci USA 95:15677–15682
Schomberg DW, Couse JF, Mukherjee A, Lubahn DB, Sar M, Mayo KE, Korach KS 1999 Targeted disruption of the estrogen receptor- gene in female mice: characterization of ovarian responses and phenotype in the adult. Endocrinology 140:2733–2744
Green S, Walter P, Kumar V, Krust A, Bornert JM, Argos P, Chambon P 1986 Human oestrogen receptor cDNA: sequence, expression and homology to v-erb-A. Nature 320:134–139
Kuiper GG, Enmark E, Pelto-Huikko M, Nilsson S, Gustafsson JA 1996 Cloning of a novel receptor expressed in rat prostate and ovary. Proc Natl Acad Sci USA 93:5925–5930
Mosselman S, Polman J, Dijkema R 1996 ER?: identification and characterization of a novel human estrogen receptor. FEBS Lett 392:49–53
Gore-Langton RE, Amstrong DT 1994 Follicular steroidogenesis and its control. In: Knobil E, Neil JD, eds. The physiology of reproduction. 2nd ed. New York: Raven Press; 571–617
Tetsuka M, Hillier SG 1996 Androgen receptor gene expression in rat granulosa cells: the role of follicle-stimulating hormone and steroid hormones. Endocrinology 137:4392–4397
Fitzpatrick SL, Richards JS 1991 Regulation of cytochrome P450 aromatase messenger ribonucleic acid and activity by steroids and gonadotropins in rat granulosa cells. Endocrinology 129:1452–1462
Zhuang LZ, Adashi EY, Hsuch AJ 1982 Direct enhancement of gonadotropin-stimulated ovarian estrogen biosynthesis by estrogen and clomiphene citrate. Endocrinology 110:2219–2221
Adashi EY, Hsueh AJ 1982 Estrogens augment the stimulation of ovarian aromatase activity by follicle-stimulating hormone in cultured rat granulosa cells. J Biol Chem 257:6077–6083
Segaloff DL, Wang HY, Richards JS 1990 Hormonal regulation of luteinizing hormone/chorionic gonadotropin receptor mRNA in rat ovarian cells during follicular development and luteinization. Mol Endocrinol 4:1856–1865
Kessel B, Liu YX, Jia XC, Hsueh AJ 1985 Autocrine role of estrogens in the augmentation of luteinizing hormone receptor formation in cultured rat granulosa cells. Biol Reprod 32:1038–1050
Knecht M, Brodie AM, Catt KJ 1985 Aromatase inhibitors prevent granulosa cell differentiation: an obligatory role for estrogens in luteinizing hormone receptor expression. Endocrinology 117:1156–1161
Goldenberg RL, Vaitukaitis JL, Ross GT 1972 Estrogen and follicle stimulation hormone interactions on follicle growth in rats. Endocrinology 90:1492–1498
Wang XN, Greenwald GS 1993 Synergistic effects of steroids with FSH on folliculogenesis, steroidogenesis and FSH- and hCG-receptors in hypophysectomized mice. J Reprod Fertil 99:403–413
Hirshfield AN 1991 Development of follicles in the mammalian ovary. Int Rev Cytol 124:43–101
Burghardt RC, Anderson E 1981 Hormonal modulation of gap junctions in rat ovarian follicles. Cell Tissue Res 214:181–193
Billig H, Furuta I, Hsueh AJ 1993 Estrogens inhibit and androgens enhance ovarian granulosa cell apoptosis. Endocrinology 133:2204–2212
Banks PK, Meyer K, Brodie AM 1991 Regulation of ovarian steroid biosynthesis by estrogen during proestrus in the rat. Endocrinology 129:1295–12304
Couse JF, Yates MM, Walker VR, Korach KS 2003 Characterization of the hypothalamic-pituitary-gonadal axis in estrogen receptor (ER) null mice reveals hypergonadism and endocrine sex reversal in females lacking ER but not ER?. Mol Endocrinol 17:1039–1053
Byers M, Kuiper GG, Gustafsson JA, Park-Sarge OK 1997 Estrogen receptor-? mRNA expression in rat ovary: down-regulation by gonadotropins. Mol Endocrinol 11:172–182
Sharma SC, Clemens JW, Pisarska MD, Richards JS 1999 Expression and function of estrogen receptor subtypes in granulosa cells: regulation by estradiol and forskolin. Endocrinology 140:4320–4334
Fitzpatrick SL, Funkhouser JM, Sindoni DM, Stevis PE, Deecher DC, Bapat AR, Merchenthaler I, Frail DE 1999 Expression of estrogen receptor-? protein in rodent ovary. Endocrinology 140:2581–2591
Tetsuka M, Milne M, Hillier SG 1998 Expression of oestrogen receptor isoforms in relation to enzymes of oestrogen synthesis in rat ovary. Mol Cell Endocrinol 141:29–35
Sar M, Welsch F 1999 Differential expression of estrogen receptor-? and estrogen receptor- in the rat ovary. Endocrinology 140:963–971
Couse JF, Bunch DO, Lindzey J, Schomberg DW, Korach KS 1999 Prevention of the polycystic ovarian phenotype and characterization of ovulatory capacity in the estrogen receptor- knockout mouse. Endocrinology 140:5855–5865
Huynh K, Jones G, Thouas G, Britt KL, Simpson ER, Jones ME 2004 Estrogen is not directly required for oocyte developmental competence. Biol Reprod 70:1263–1269
Rosenfeld CS, Murray AA, Simmer G, Hufford MG, Smith MF, Spears N, Lubahn DB 2000 Gonadotropin induction of ovulation and corpus luteum formation in young estrogen receptor- knockout mice. Biol Reprod 62:599–605
Dupont S, Krust A, Gansmuller A, Dierich A, Chambon P, Mark M 2000 Effect of single and compound knockouts of estrogen receptors (ER) and ? (ER?) on mouse reproductive phenotypes. Development 127:4277–4291
Nayudu PL, Osborn SM 1992 Factors influencing the rate of preantral and antral growth of mouse ovarian follicles in vitro. J Reprod Fertil 95:349–362
Rose UM, Hanssen RG, Kloosterboer HJ 1999 Development and characterization of an in vitro ovulation model using mouse ovarian follicles. Biol Reprod 61:503–511
Wu J, Nayudu PL, Kiesel PS, Michelmann HW 2000 Luteinizing hormone has a stage-limited effect on preantral follicle development in vitro. Biol Reprod 63:320–327
Ochsner SA, Day AJ, Rugg MS, Breyer RM, Gomer RH, Richards JS 2003 Disrupted function of tumor necrosis factor--stimulated gene 6 blocks cumulus cell-oocyte complex expansion. Endocrinology 144:4376–4384
Richards JS, Russell DL, Ochsner S, Hsieh M, Doyle KH, Falender AE, Lo YK, Sharma SC 2002 Novel signaling pathways that control ovarian follicular development, ovulation, and luteinization. Recent Prog Horm Res 57:195–220
Davis BJ, Lennard DE, Lee CA, Tiano HF, Morham SG, Wetsel WC, Langenbach R 1999 Anovulation in cyclooxygenase-2-deficient mice is restored by prostaglandin E2 and interleukin-1?. Endocrinology 140:2685–2695
Rao MC, Midgley Jr AR, Richards JS 1978 Hormonal regulation of ovarian cellular proliferation. Cell 14:71–78
Bao B, Kumar N, Karp RM, Garverick HA, Sundaram K 2000 Estrogen receptor-? expression in relation to the expression of luteinizing hormone receptor and cytochrome P450 enzymes in rat ovarian follicles. Biol Reprod 63:1747–1755
Burns KH, Agno JE, Chen L, Haupt B, Ogbonna SC, Korach KS, Matzuk MM 2003 Sexually dimorphic roles of steroid hormone receptor signaling in gonadal tumorigenesis. Mol Endocrinol 17:2039–2052
Robker RL, Richards JS 1998 Hormone-induced proliferation and differentiation of granulosa cells: a coordinated balance of the cell cycle regulators cyclin D2 and p27Kip1. Mol Endocrinol 12:924–940
Curry Jr TE, Osteen KG 2001 Cyclic changes in the matrix metalloproteinase system in the ovary and uterus. Biol Reprod 64:1285–1296
Lydon JP, DeMayo FJ, Funk CR, Mani SK, Hughes AR, Montgomery CA, Shyamala G, Conneely OM, O’Malley BW 1995 Mice lacking progesterone receptor exhibit pleiotropic reproductive abnormalities. Genes Dev 9:2266–2278
Ochsner SA, Russell DL, Day AJ, Breyer RM, Richards JS 2003 Decreased expression of tumor necrosis factor--stimulated gene 6 in cumulus cells of the cyclooxygenase-2 and EP2 null mice. Endocrinology 144:1008–1019
Cheng G, Weihua Z, Makinen S, Makela S, Saji S, Warner M, Gustafsson JA, Hovatta O 2002 A role for the androgen receptor in follicular atresia of estrogen receptor ? knockout mouse ovary. Biol Reprod 66:77–84(Judith M. A. Emmen, John )