help button home button Endocrine Society Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology, doi:10.1210/en.2005-0213
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
146/8/3247    most recent
Author Manuscript (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Couse, J. F.
Right arrow Articles by Korach, K. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Couse, J. F.
Right arrow Articles by Korach, K. S.
Right arrowPubmed/NCBI databases
*Gene*GEO Profiles
*HomoloGene*UniGene
*Compound via MeSH
*Substance via MeSH
Hazardous Substances DB
*ESTRADIOL
*MENOTROPINS
*PROGESTERONE
Endocrinology Vol. 146, No. 8 3247-3262
Copyright © 2005 by The Endocrine Society

Estrogen Receptor-ß Is Critical to Granulosa Cell Differentiation and the Ovulatory Response to Gonadotropins

John F. Couse, Mariana M. Yates, Bonnie J. Deroo and Kenneth S. Korach

Receptor Biology Section, Laboratory of Reproductive and Developmental Toxicology, 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{at}niehs.nih.gov.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
The process of granulosa cell differentiation that occurs in preovulatory follicles is dependent on FSH but requires augmentation by estradiol. To determine which estrogen receptor (ER) form mediates the effects of estradiol during gonadotropin-induced follicle growth, differentiation, and rupture, we characterized the response of ER{alpha}- and ERß-null mice to gonadotropin-induced ovulation. Immature mice were treated with an ovulatory regimen of exogenous gonadotropins and tissues were collected at distinct time points for morphological, biochemical, gene expression, and immunohistochemical analyses. Granulosa cells of ERß knockout (ERKO) preovulatory follicles exhibited an attenuated response to FSH-induced differentiation, as evident by reduced aromatase activity and estradiol synthesis, and insufficient expression of LH receptor. As a result, ßERKO ovaries were unable to fully respond to an ovulatory bolus of gonadotropin, leading to a reduced rate of follicle rupture; insufficient induction of prostaglandin-synthase 2 and progesterone receptor; an aberrant increase in aromatase activity and plasma estradiol; and incomplete expansion of the cumulus-oocyte complex. Parallel characterization of {alpha}ERKO females indicated a minimal role for ER{alpha} in granulosa cell differentiation, ovulation, and the concomitant changes in gene expression, although some abnormalities were revealed. These studies demonstrate that ERß-mediated estradiol actions are vital to FSH-induced granulosa cell differentiation; and in the absence of ERß, preovulatory follicles are deficient in the necessary cellular organization (i.e. antrum and cumulus oocyte complex), enzymatic activity (i.e. capacity to convert androgen precursor to estradiol), and receptor signaling pathways (i.e. LH receptor) to respond to a gonadotropin surge and expel a healthy oocyte.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
IN THE OVARY, folliculogenesis begins with the recruitment of a cohort of primordial follicles from the existing pool. Once enlisted, the selected follicles embark on an irreversible course of oocyte growth, granulosa cell proliferation, and thecal cell differentiation. The resulting preantral follicles become highly responsive to FSH from the anterior pituitary and are induced to differentiate to the preovulatory stage. Fully differentiated preovulatory follicles are characterized by an increased capacity to synthesize large amounts of estradiol via aromatization of thecal-derived androgens, formation of a large fluid-filled antrum, and the acquisition of LH-receptor and responsiveness. The synchronized assembly of mature follicles in the ovary provides for increasing levels of circulating estradiol, which ultimately peaks and acts on the hypothalamic-pituitary axis to elicit secretion of a massive bolus of LH. The acquisition of LH receptor by healthy preovulatory follicles allows them to selectively respond to the LH surge, leading to ovulation (rupture and expulsion of the oocyte) and terminal differentiation of the somatic cells to form a corpus luteum (1). Therefore, follicle growth and ovulation are remarkably synchronous processes that depend on the functions of multiple ovarian cell types and a myriad of endocrine signaling pathways. This complexity of follicular growth and ovulation in the mammalian ovary is illustrated by the growing number of gene-targeted mouse models that exhibit disruptions in one or both ovarian processes and hence have reduced fecundity or are infertile. For example, FSH-mediated follicular growth and differentiation are greatly impacted in mice null for IGF-I (Igf1) (2), growth differentiation factor (Gdf)-9 (3), cyclin D2 (Ccnd2) (4), {alpha}-inhibin (Inha) (5), FSH-ß (Fshb) (6), FSH receptor (Fshr) (7), or phosphodiesterase-4D (Pde4d) (8). In turn, mice null for connexin-37 (Gja4) (9), prostaglandin-synthase-2 (Ptgs2) (10), CCAAT/enhancer binding protein-ß (Cebpb) (11), or progesterone receptor (Pgr) (12) possess ovaries that exhibit fully developed preovulatory follicles that fail to rupture after LH stimulation.

We have previously reported that mice null for estrogen receptor (ER)-{alpha} (Esr1) or estrogen receptor-ß (Esr2) also exhibit distinct ovarian defects that lead to compromised fertility (13). Mice null for ER{alpha} ({alpha}ERKO) are anovulatory and invariably exhibit ovaries that possess multiple hemorrhagic/cystic follicles and elevated steroid synthesis, both of which are principally due to hyperstimulation of the ovary by elevated basal levels of circulating LH (14, 15). However, immature {alpha}ERKO females do successfully ovulate when treated with exogenous gonadotropins if administered before onset of the overt hypergonadotropic phenotypes in the ovary (14). In contrast, ßERKO mice exhibit ovaries that are grossly normal but are oligoovulatory and fail to exhibit efficient ovulation even when treated with exogenous gonadotropins (13). Therefore, the ovarian phenotypes in ERKO mice suggest that ERß plays a more critical intraovarian role than ER{alpha}. This hypothesis is supported by a plethora of evidence that ERß is the predominant ER form expressed in the granulosa cells of growing and mature follicles in the rodent ovary (16, 17, 18). It has been long recognized that estradiol augments the actions of FSH on granulosa cells and both hormones are necessary to establish a fully differentiated and healthy preovulatory follicle, but the mechanisms involved are less clear (19). For example, estradiol alone has little effect on rodent granulosa cells but is required for maximum FSH stimulation of the following: 1) aromatase (Cyp19) expression (20) and estradiol synthesis (21, 22), 2) LH receptor (Lhcgr) expression (23) and LH responsiveness (24, 25, 26), 3) antrum formation (27), 4) gap-junction formation (28, 29), and 5) prevention of atresia (30). Indeed, immature rats treated with an aromatase inhibitor during gonadotropin-induced ovulation exhibit a severely reduced number of healthy preovulatory follicles and ovulation rates (31).

Therefore, the following studies were undertaken to further characterize the ovulatory defects in ERKO mice with the aim of determining the contribution of each ER form to follicle growth and rupture. Immature control, {alpha}ERKO, and ßERKO females were induced to ovulate via treatment with exogenous pregnant mare serum gonadotropin (PMSG) followed 48 h later by human choriogonadotropin (hCG). Ovaries were collected at four specific points during the course of treatment for biochemical, immunohistochemical, and histomorphological analyses. Immature ßERKO females yielded a reduced number of ovulated oocytes, compared with similarly treated control and {alpha}ERKO females. Gene expression analyses demonstrated that ßERKO and {alpha}ERKO ovaries show relatively normal expression patterns for most genes during follicle growth and ovulation. However, PMSG-induced granulosa cell differentiation was clearly deficient in ßERKO ovaries as indicated by reduced estradiol synthesis and minimal Cyp19 and Lhcgr expression; hence, the response to subsequent hCG stimulation was severely compromised in ßERKO ovaries. Overall, these data indicate that ERß but not ER{alpha} plays an integral role in the processes of gonadotropin-induced granulosa cell differentiation that are necessary for a full response to LH-induced ovulation.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Animals and treatments
All procedures were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were preapproved by the National Institute of Environmental Health Sciences Institutional Animal Care and Use Committee. The generation of mice null for the Esr1 (ER{alpha}) or Esr2 (ERß) genes has been described previously (13). All animals were maintained in plastic cages under a 12-h light, 12-h dark schedule in a temperature-controlled room (21–22 C), fed National Institutes of Health 31 mouse chow and fresh water ad libitum. Immature {alpha}ERKO (Esr1–/–) female mice were generated via heterozygous (Esr1+/–) breeding pairs. Immature ßERKO (Esr2–/–) female mice were generated via breeding homozygous (Esr2–/–) males with heterozygous (Esr2+/–) females. All females were weaned at 21 d of age, genotyped as previously described (15), and subjected to gonadotropin-induced ovulation while between 23 and 30 d of age. The average animal weights at the time of necropsy were as follows: wild-type, 13.8 g; heterozygous, 14.3 g; {alpha}ERKO, 13.2 g; and ßERKO, 14.8 g. Gonadotropin-induced ovulation consisted of treating animals with 3.25 IU PMSG (Sigma Chemical Co., St. Louis, MO) between 1300 and 1500 h followed 48 h later with hCG (Sigma) of doses varying from 1.1 to 17.6 IU. Both drugs were dissolved in 0.85% saline solution and injected sc in a total volume of 0.1 ml. At necropsy, animals were killed by CO2-asphyxiation; whole blood was immediately collected via the descending vena cava; and the ovaries were dissected, trimmed of oviduct, and surrounding tissue and snap frozen, fixed in cold 10% buffered formalin, or processed for granulosa cell isolation.

Oocyte counts after induced ovulation
Animals were killed 16–20 h after hCG treatment and the ovaries and oviduct removed to M-2 medium (Specialty Media, Lavallette, NJ). The cumulus-oocyte complexes were surgically extracted from each oviduct, pooled into a single droplet of M-2 medium supplemented with 0.3% hyaluronidase (Sigma), and allowed to undergo enzymatic disassociation at 37 C for approximately 15 min. The number of oocytes per animal was then determined by manually counting using a MZ6 stereoscope (Leica, Québec, Canada). The data presented are the combination of multiple induced ovulation experiments carried out over a period of time.

Granulosa cell isolation
Ovaries were removed and immediately transferred to a 100-mm cell culture dish containing 15 ml ice-cold M199 medium supplemented with 1 mg/ml BSA, 2.5 µg/ml Fungizone, and 50 µg/ml gentamicin (all reagents from Invitrogen, Carlsbad, CA). Ovaries were pooled according to genotype and treatment. Once all ovaries were collected, the granulosa cells from each were expressed by manual puncture with 25-gauge needles followed by slight pressure applied with a sterile spatula. Follicular debris was then removed manually and the granulosa cell suspension filtered through a 150-µm Nitex nylon membrane (Sefar America Inc., Depew, NY) mounted in Swinnex filters (Millipore, Billerica, MA). The granulosa cells were then pelleted by centrifugation at 250 x g for 5 min at 4 C, followed by two washes in DMEM/F-12 medium. The final cell pellet was frozen at –80 C until RNA extraction.

RNA isolation and semiquantitative RT-PCR (SQ-RT-PCR) gene expression assays
Total RNA was isolated from frozen ovaries using TRIZOL reagent (Invitrogen) according to the manufacturer’s protocol. All RNA samples were rid of contaminating DNA using the DNA-free reagents (Ambion, Austin, TX) according to the manufacturer’s protocol, followed by normalization to a concentration of 0.2 µg/ml in RNase-free water. A SQ-RT-PCR approach was used to assess gene expression levels as previously described (15). For each individual ovarian RNA sample, 1 µg of RNA was used in a 25-µl cDNA reaction using random hexamers and the Superscript cDNA synthesis system (Invitrogen) according to the manufacturer’s protocol. PCRs were prepared using the equivalent of 1 µl cDNA per 15-µl reaction for each respective primer set (Table 1Go) using PCR reagents and Platinum Taq polymerase (Invitrogen) as previously described (15). PCR was carried out in a Thermo Hybaid multiblock system (Franklin, MA) at 95 C for 30 sec, 58 C for 45 sec, and 72 C for 30 sec for 26–32 cycles, depending on the level of gene expression. Primers for ribosomal protein L7 (Rpl7) were included in all reactions as an internal positive control and for normalization. All samples were electrophoresed on an agarose gel (2% NuSieve/0.7% SeaKem, BMA Bioproducts, Rockland, ME) in 1x Tris-borate EDTA followed by immobilization to BrightStar nylon membrane (Ambion) using the Royal Genie blotter (Idea Scientific, Minneapolis, MN). All blots were probed with a nested oligo specific to sequences internal to the PCR primers (Table 1Go) that was 5'-radiolabeled with [33P-{gamma}ATP] (Amersham Pharmacia Biotech, Buckinghamshire, UK) using T4 polynucleotide kinase (New England Biolabs, Beverly, MA). Hybridization of blots was carried out in Rapid-Hyb buffer (Amersham Pharmacia Biotech) with more than 3 x 106 cpm of probe per milliliter overnight in a rotisserie hybridization oven (Thermo Hybaid) at 42 C, followed by washing according to the manufacturer’s protocol. Final SQ-RT-PCR blots were exposed to a PhosphorImager screen and the data analyzed with a Storm 860 and accompanying ImageQuant Software (Molecular Dynamics, Sunnyvale, CA).


View this table:
[in this window]
[in a new window]
  		TABLE 1. Genes evaluated over the course of gonadotropin-induced ovulation in immature ERKO mice

 
RNA isolation and Northern blot analysis of gene expression
TRIZOL reagent (Invitrogen) was used according to the manufacturer’s protocol to extract total RNA from individual pairs of frozen ovaries or isolated granulosa cells from pooled ovaries. Northern blots were generated using 20 µg total RNA from whole ovaries or 5 µg total RNA from isolated granulosa cells using NorthernMax (Ambion) reagents according to the manufacturer’s protocol. RNAs were fractionated on a denaturing 1.2% agarose gel and immobilized to Brightstar nylon membrane (Ambion) by capillary action according to the manufacturer’s protocol. All membranes were prehybridized and hybridized in NorthernMax Pre/Hyb buffer (Ambion) in a hybridization oven (Thermo Hybaid) at 65 C. All blots were probed with strippable, radiolabeled riboprobes generated using Strip-EZ riboprobe transcription (Ambion) reagents and [32P]-{alpha}UTP (Amersham). Blots were probed with 1–5 x 107 cpm of purified probe per milliliter of hybridization buffer at 65 C overnight and washed according to the membrane manufacturer’s protocol. Blots were then exposed to a PhosphorImager screen and analyzed with a Storm 860 and accompanying ImageQuant Software (Molecular Dynamics). Once exposed, the blots were immediately stripped using the Strip-EZ (Ambion) reagents and prepared for reprobing. All probes used herein have been previously described: ß-actin (Actb) (15), Cyp19 (15), Lhcgr (32), Ptgs2, and Pgr (14).

Histological and immunohistochemical evaluation of ovaries
Formalin-fixed ovaries were embedded in paraffin, sectioned, and stained with hematoxylin and eosin according to standard laboratory protocols or left unstained and stored at 4 C. For immunohistochemical detection of Ptgs2, sections were subjected to deparaffinization/antigen retrieval using 1x Reveal (Biocare Medical, Walnut Creek, CA) in a decloaking chamber (Biocare Medical) according to the manufacturer’s protocol, followed by multiple washes in water and then 1x automation buffer (Biomeda, Foster City, CA). Endogenous peroxidase activity was then quenched by incubating the sections in 3% hydrogen peroxide for 10 min. Sections were then blocked for endogenous avidin and biotin using an avidin/biotin blocking reagents kit (Vector Laboratories, Burlingame, CA) and then placed under a general blocking reagent of 10% BSA in 1x automation buffer for 30 min. Sections were then incubated with goat-raised antirat Ptgs2 (M-19) (Santa Cruz Biotechnology, Santa Cruz, CA) at a dilution of 1:5000, followed by a 1:500 dilution of biotinylated antigoat IgG (Vector Laboratories), followed by a 1:50 dilution of extraavidin peroxidase (Sigma). All dilutions and subsequent washings were done in 1x automation buffer. Specific immunoreactivity was detected by applying NovaRED (Vector Laboratories) to sections for 5 min. Immunohistochemical detection of progesterone receptor (PR) followed a similar protocol except sections were first deparaffinized via xylene washes followed by antigen retrieval using 1x nuclear decloaker (Biocare Medical) in a decloaking chamber (Biocare Medical) according to the manufacturer’s protocol. Sections were then incubated with rabbit-raised antimouse PR (a gift from Dr. Dean P. Edwards, University of Colorado Health Sciences Center, Denver, CO) at a dilution of 1:100; followed by a 1:500 dilution of biotinylated antirabbit IgG (Amersham Pharmacia Biotech). All photomicrographs were collected using a BX-50 microscope and OLY-750 video camera (Olympus, Tokyo, Japan).

Western blot analysis
Cytosolic protein fractions were isolated from individual pairs of frozen ovaries using the NE-PER (Pierce Biotechnology, Rockford, IL) reagents according to the manufacturer’s protocol. Final protein preparations were quantified using the BCA-200 Protein Assay Reagents (Pierce Biotechnology) and BSA as a standard. All blots were generated using 30 µg of cytosolic protein per sample prepared in NuPage LDS sample buffer with reducing reagent (Invitrogen), fractionated on a 4–12% Tris-Bis acrylamide gel in 3[N-morpholino]propanesulfonic acid buffer (Invitrogen), and immobilized to nitrocellulose membrane (Invitrogen) according to the manufacturer’s protocol. For detection of Ptgs2, blots were probed with rabbit-raised antimouse cyclooxygenase-2 (Cayman Chemical Co., Ann Arbor, MI) at a dilution of 1:1,000, followed by a 1:20,000 dilution of horseradish peroxidase-conjugated antirabbit IgG (Cell Signaling Technology, Beverly, MA). For detection of PR, blots were probed with rabbit-raised antimouse PR (a gift from Dr. Dean P. Edwards) at 10 µg/ml, followed by a 1:20,000 dilution of horseradish peroxidase-conjugated antirabbit IgG (Cell Signaling Technology). Blocking was carried out in a solution of 5% nonfat dry milk in 1x Tris-buffered saline with Tween 20, and all postantibody washes were carried out in 1x Tris-buffered saline with Tween 20 alone. Final detection of immunoreactivity was carried out using the SuperSignal WestDura (Pierce Biotechnology) reagents according to the manufacturer’s protocol and captured via exposure to x-ray film or using an EpiChemi3 Darkroom (UVP Bioimaging Systems, Upland, CA).

Hormone RIAs
Upon necropsy, whole blood was collected from the inferior vena cava and mixed with 10 µl heparin (60 mg/ml) on ice. Plasma was then collected and stored at –70 C until assay. All hormone assays were carried out on plasma collected from individual animals. Plasma progesterone and estradiol were assayed in singlicate per animal on 25- and 200-µl aliquots, respectively, using the active progesterone RIA and ultrasensitive estradiol double-antibody RIA kits (Diagnostic Systems Laboratories, Webster, TX) according to the manufacturer’s protocol. All final assays were quantified using a Packard Multi-Prias 2 {gamma}-counter (PerkinElmer, Boston, MA). The following parameters applied for each RIA: estradiol (one assay; least detectable concentration 2.2 pg/ml; average intraassay coefficient of variation 18%; average interassay coefficient of variation not applicable); and progesterone (two assays; least detectable concentration 0.3 ng/ml; and average intraassay coefficient of variation 14%, and average interassay coefficient of variation 13%).

Statistical analysis
All data were analyzed for statistical significance (P < 0.05) using JMP software (SAS Institute, Cary, NC). Statistical analysis of oocyte counts indicated some difference between experimental trials but within the expected range given the variability of ovulation as an experimental end point. Therefore, our statistical analysis did not warrant the omission of any one experimental trial and allowed the data of all trials to be combined into a single set. Because the oocyte yield data were skewed, the Kruskal-Wallis ANOVA was used followed by the Mann-Whitney U test to determine where significant differences existed. Plasma steroid levels and gene expression data sets were first tested for homoscedasticity of variance using the Levene’s test and, if failed, were log transformed before further statistical analysis. All data sets were then evaluated using a one-way ANOVA followed by the Tukey-Kramer honestly significant difference post hoc test when applicable.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
ERß-null females exhibit reduced oocyte yield after induced ovulation
Statistical comparison of oocyte yields after treatment with a previously determined optimum regimen of PMSG (3.25 IU) and hCG (2.2 IU) indicated no difference between immature wild-type (Esr1+/+ or Esr2+/+) and heterozygous (Esr2+/–) females, allowing the results from all three genotypes to be combined into a single group that was defined thereafter as controls. As shown in Fig. 1AGo, the average oocyte yield following standard induced ovulation in immature ßERKO females (9.6 ± 2.1 oocyte/female) was significantly reduced (P < 0.05), compared with immature control (30.6 ± 1.7) and {alpha}ERKO (29.3 ± 6.5) females. The rate of ovulation in ßERKO females observed herein is consistent with our original description of the ßERKO in which assays were conducted on a much smaller number of animals (13). All three genotypes exhibited a broad range of oocyte yields, characteristic of gonadotropin-induced ovulation in mice. However, seven of 31 (22.5%) treated ßERKO females failed to ovulate even a single recoverable oocyte, whereas the lowest individual yield among control and {alpha}ERKO females was six and seven oocytes, respectively.



View larger version (16K):
[in this window]
[in a new window]
  		FIG. 1. Oocyte yield after induced ovulation in immature control and ERKO mice. Immature mice were treated with PMSG (3.25 IU) followed 48 h later with hCG at the dose indicated. Oviducts were then collected 16–20 h after hCG treatment and the oocytes extracted and counted. A, Results after treatment with gonadotropin doses previously determined to be optimum for induced ovulation. Each dot represents the oocyte yield of an individual animal within their respective genotype; vertical bars represent the average oocyte yield within each genotype. B, hCG dose-response for control and ßERKO females showing the average oocyte yield for each genotype after being treated with PMSG (3.25 IU) followed by increasing doses of hCG 48 h later. a, P < 0.05 vs. control with 2.2 IU hCG; b, P < 0.05 vs. oocyte yield with 2.2 IU hCG within the same genotype.

 
To determine whether an increased dose of hCG might improve the ovulatory response in immature ßERKO females, animals were treated with PMSG (3.25 IU) followed 48 h later with a dose of hCG ranging from 1.1 to 17.6 IU (Fig. 1BGo). As expected, 2.2 IU hCG proved to be the optimum dose in controls as the average yield after 1.1 IU hCG was significantly reduced (P < 0.05), and no significant improvement was observed among the three doses greater than 2.2 IU. Simply doubling the hCG dose to 4.4 IU did improve the ovulatory response in immature ßERKO females, but no further improvement was observed with 8.8 IU or 17.6 IU hCG. All three higher doses of hCG led to an improved frequency of recoverable oocytes from the oviducts of ßERKO females, with the lowest yield being three oocytes. Nonetheless, even at 17.6 IU hCG, 8-fold the optimum, the average oocyte yield among immature ßERKO females remained below that of controls at all doses. These data confirm our previous findings that ßERKO females exhibit an attenuated response to gonadotropin-induced ovulation in terms of oocyte yield and further demonstrates this deficit cannot be overcome with increased doses of hCG.

ER-null females exhibit aberrant plasma steroid levels during induced ovulation
The plasma levels of estradiol and progesterone fluctuate in dramatic but predictable patterns during folliculogenesis and ovulation and can therefore be used as parameters for evaluating the ovarian response to gonadotropin-induced ovulation. As shown in Fig. 2AGo, plasma estradiol levels in control females increased 6-fold (from 6.8 to 40.5 pg/ml) 48 h after a single PMSG treatment but dropped precipitously after hCG treatment to a nadir just before ovulation (~12–14 h after hCG). Plasma progesterone levels in control females were relatively unaffected by PMSG but rose rapidly within 4 h of hCG treatment and peaked at 10-fold the basal level and then gradually decreased thereafter (Fig. 2AGo). These data are congruent with the temporal changes in plasma steroid levels documented to occur during the rodent ovarian cycle (33).



View larger version (19K):
[in this window]
[in a new window]
  		FIG. 2. Plasma levels of ovarian steroids during gonadotropin-induced ovulation. Immature mice were treated with PMSG (3.25 IU) followed 48 h later with hCG (2.2 IU). Animals were killed at the indicated times over the course of gonadotropin-induced ovulation. Plasma was prepared from whole blood and assayed for estradiol and progesterone content by RIA. Shown are the average (±SEM) plasma levels of estradiol and progesterone at each time point for control (A), {alpha}ERKO (B), and ßERKO females (C). a, P < 0.05 vs. average control estradiol level at the same time point; b, P < 0.05 vs. average control progesterone at the same time point.

 
Immature {alpha}ERKO females exhibited a control-like pattern of plasma estradiol over the course of induced ovulation (Fig. 2BGo). However, plasma estradiol levels 48 h after PMSG and 4 h after hCG were much higher in {alpha}ERKO vs. controls and proved to be statistically significant (P < 0.05) at the latter time point. By 10–12 h after hCG treatment, plasma estradiol levels in {alpha}ERKO females were similar to that of control. The pattern and level of plasma progesterone among {alpha}ERKO females mirrored that of control, differing only at the final time point (20 h after hCG), at which {alpha}ERKO levels were approximately half that of control (9.2 vs. 20 ng/ml) (Fig. 2BGo). In contrast, immature ßERKO females exhibited a strikingly abnormal pattern of estradiol secretion such that plasma levels were abnormally low 48 h after PMSG treatment (P < 0.05 vs. control) but rose rapidly to a peak of more than 40 pg/ml within 4 h after hCG treatment (Fig. 2CGo), a significant divergence from the precipitous drop in plasma estradiol that followed hCG in control females (compare with Fig. 2AGo). Estradiol levels did eventually decline in ßERKO females, but the expected nadir at 10 h after hCG remained more than 3-fold higher (P < 0.05) than controls. The pattern and level of progesterone secretion in immature ßERKO females was similar to control, rapidly peaking 4 h after hCG and decreasing thereafter. Therefore, {alpha}ERKO females exhibit a normal pattern but somewhat aberrant levels of sex steroid secretion over the course of induced ovulation, whereas ßERKO females exhibit an abnormal pattern, marked by an aberrant increase in estradiol after hCG treatment that results in a simultaneous peak in estradiol and progesterone synthesis just hours before ovulation.

Ovarian histology during induced ovulation
Ovaries from immature untreated control, {alpha}ERKO, and ßERKO females were relatively indistinguishable because all exhibited a sparse number of growing follicles, most of which possessed multiple indications of atresia (data not shown). Some small antral follicles were observed in untreated {alpha}ERKO ovaries (data not shown), congruent with previous descriptions of premature stimulation by rising endogenous gonadotropins in this model (14, 32). PMSG treatment induced the appearance of multiple growing antral follicles in the ovaries of all three genotypes within 48 h. Compared with follicles in control and {alpha}ERKO ovaries, maturing follicles in ßERKO ovaries tended to exhibit slightly undersized antra (data not shown), but this was variable and difficult to quantify. Almost all {alpha}ERKO ovaries exhibited at least one large hemorrhagic and cystic follicle 48 h after PMSG treatment, similar to the cystic follicles described by others to occur spontaneously in older {alpha}ERKO females (32). None of the PMSG-stimulated control or ßERKO ovaries exhibited {alpha}ERKO-like hemorrhagic follicles.

By 10 h after hCG treatment, control and {alpha}ERKO ovaries exhibited multiple preovulatory follicles undergoing expansion of the cumulus-oocyte complex (COC) (Fig. 3Go, A–E) or already beginning corpus luteum formation (data not shown). In contrast, visible COC expansion among ovulatory follicles in ßERKO ovaries 12 h after hCG treatment (Fig. 3Go, C and F) was noticeably variable because observance of an ovulatory follicle undergoing COC expansion juxtaposed to another that was not in ßERKO ovaries was common.



View larger version (101K):
[in this window]
[in a new window]
  		FIG. 3. ßERKO follicles fail to exhibit cumulus-oocyte expansion 10 h after hCG-induced ovulation. Immature mice were treated with PMSG (3.25 IU) followed 48 h later with hCG (2.2 IU). Animals were killed 10 h after hCG treatment, just before the time of follicle rupture. Shown are representative cross-sections of control (A and D), {alpha}ERKO (B and E) and ßERKO (C and F) follicles. There is obvious expansion of the COC in the control and {alpha}ERKO examples, whereas in both ßERKO examples, the cumulus oophorus cells remain relatively packed around the oocyte and still retain some attachment to the mural granulosa cells.

 
SQ-RT-PCR for gene expression during induced ovulation
The mouse ovary exhibits a complex but synchronous pattern of gene expression during folliculogenesis and ovulation. Because the ERs are ligand-inducible transcription factors, the insufficient response of ßERKO follicles to gonadotropin-induced maturation and ovulation may be due to the loss of ERß-mediated expression of genes that are critical to these processes. To address this issue, we evaluated gene expression in the ovaries of immature control, {alpha}ERKO, and ßERKO females over the course of standard gonadotropin-induced ovulation with 3.25 IU PMSG followed 48 h later with 2.2 IU hCG. Ovaries were collected from animals of each genotype at five distinct time points during induced ovulation: 1) no treatment, 2) 48 h after PMSG, 3) 4 h after hCG, 4) 10 h after hCG, and 5) 16–20 h after hCG. The finite amount of RNA available from the ovaries of individual animals limited the number of genes that could be evaluated by traditional means. We therefore developed a SQ-RT-PCR method that allowed for evaluation of the expression of more than 35 different genes in the ovaries of individual animals. The SQ-RT-PCR method also permitted the sample numbers in this initial screen of gene expression to be more than eight per genotype per time point, making the assay more robust, allowing for proper statistical analysis of the data. Table 1Go lists all of the genes evaluated in this study, including those that exhibited little change in expression over the course of induced ovulation or in which changes occurred but no genotypic differences were observed. Genes that exhibited more dramatic changes in expression over the course of induced ovulation are discussed in greater detail below. Certain genes that are known to play an important role in folliculogenesis and ovulation were found to be aberrantly expressed in ßERKO ovaries by SQ-RT-PCR and were therefore evaluated further by: 1) Northern blot on RNA prepared from whole ovaries and isolated granulosa cells, 2) Western blot analysis of whole ovarian protein extracts, and 3) immunohistochemistry in whole ovaries at the most relevant time point. In general, Northern blot analyses confirmed the results initially generated by SQ-RT-PCR, thereby validating this method as an acceptable semi-high-throughput screen for changes and abnormalities in gene expression.

ERß-null granulosa cells exhibit insufficient preovulatory differentiation
The process of granulosa cell differentiation that occurs during progression from a preantral to preovulatory follicle is dependent on sufficient FSH stimulation (6, 7, 34) and is marked by the acquisition of increased aromatase activity and LH receptor. Because estradiol is known to augment the actions of FSH (19), we were specifically interested in evaluating the expression pattern of the genes encoding Fshr, Lhcgr, and Cyp19 in control and ERKO ovaries 48 h after PMSG (an FSH analog) treatment. As expected, Fshr expression was increased 3-fold in PMSG-treated vs. untreated control ovaries (Fig. 4AGo). Similar PMSG induction of Fshr expression was observed in {alpha}ERKO and ßERKO ovaries, indicating this effect of PMSG is independent of ER actions (Fig. 4AGo). Within 4 h after hCG treatment, Fshr expression returned to basal levels in all three genotypes, although post-hCG levels in ßERKO ovaries were elevated vs. control (Fig. 4AGo).



View larger version (70K):
[in this window]
[in a new window]
  		FIG. 4. Ovarian expression of Fshr, Lhcgr, and Cyp19 48 h after PMSG treatment in control and ERKO ovaries. Immature mice were treated with PMSG (3.25 IU) followed 48 h later with hCG (2.2 IU). Animals were killed at the indicated times over the course of gonadotropin-induced ovulation, whole ovaries collected, and total RNA extracted for gene expression analysis. A, Quantitative data from SQ-RT-PCR for Fshr, Lhcgr, and Cyp19. All data are expressed as the average fold control-none (untreated) ± SEM. B, Northern blot analysis for Cyp19 and Lhcgr expression in whole ovaries or granulosa cells isolated from control and ßERKO females during the indicated treatments and time points. a, P < 0.05 vs. control at the same time point.

 
Acquisition of LH receptor by preovulatory granulosa cells is a distinct marker of differentiation and is critical to their ability to respond to the LH surge and initiate ovulation. As expected, Lhcgr expression was increased 5-fold 48 h after PMSG treatment in control ovaries and remained relatively unchanged after hCG treatment when evaluated by SQ-RT-PCR in whole ovaries (Fig. 4AGo). Immature {alpha}ERKO ovaries exhibited a similar but exaggerated increase (~5-fold control levels) in Lhcgr expression 48 h after PMSG treatment, congruent with our previous descriptions of elevated Lhcgr expression in the ovaries and granulosa cells of adult {alpha}ERKO ovaries (14, 32). ßERKO ovaries also exhibited a relatively normal PMSG induction of Lhcgr expression when evaluated by SQ-RT-PCR in whole ovaries (Fig. 4AGo). However, whereas hCG had little effect on Lhcgr expression in control and {alpha}ERKO ovaries, it led to a further doubling in ßERKO ovaries after 4 h. Similar expression data for Lhcgr were generated when whole ovary RNA preparations were evaluated by Northern blot analysis, although some reduction in the levels of Lhcgr transcripts after PMSG treatment was apparent in ßERKO ovaries (Fig. 4BGo).

Because Lhcgr expression is constitutive in thecal cells, it is difficult to accurately assess granulosa cell-specific expression when evaluated in whole ovarian RNA preparations. Therefore, total RNA was prepared from enriched granulosa cell samples collected from control and ßERKO ovaries 48 h after PMSG and 4 h after hCG (which followed 48 h PMSG treatment). As shown in Fig. 4BGo, granulosa cell-specific levels of Lhcgr transcripts in ßERKO ovaries were dramatically reduced, compared with control ovaries after PMSG and PMSG/hCG treatment. Interestingly, the Lhcgr probe used for Northern blot analysis indicated two prominent transcripts in whole ovary RNA but only one in enriched granulosa cells (Fig. 4BGo). The specificity of the Lhcgr probe used herein has been previously documented (14, 32). The reduced level Lhcgr expression in PMSG-stimulated ßERKO granulosa cells was also confirmed by Taq-man quantitative RT-PCR (data not shown). These data strongly indicate that granulosa cells of preovulatory follicles in ßERKO ovaries fail to develop sufficient LH receptor levels.

Cyp19 is necessary for the final stages of estradiol synthesis and is specifically induced by FSH in the granulosa cells of preovulatory follicles (19). As expected, PMSG treatment led to a doubling of Cyp19 expression that was transiently reduced after hCG in immature control ovaries (Fig. 4AGo), a pattern that closely mimics the plasma estradiol levels shown in Fig. 2AGo. Northern blot analysis of RNA from whole ovary or isolated granulosa cells confirmed the induction of Cyp19 expression by PMSG in control ovaries (Fig. 4BGo). {alpha}ERKO ovaries exhibited a similar, control-like expression pattern of Cyp19 except that levels were much higher after PMSG treatment, strongly correlating with the above normal plasma estradiol levels observed at this time point in {alpha}ERKO females (Fig. 2BGo). In contrast to control and {alpha}ERKO females, immature ßERKO females exhibited an aberrant Cyp19 expression pattern such that PMSG treatment had little effect but hCG induced an almost 10-fold increase within 4 h (Fig. 4AGo). The minimal induction of Cyp19 expression by PMSG but aberrant increase after hCG in ßERKO ovaries was confirmed by Northern blot analysis (Fig. 4BGo) and Taq-man quantitative RT-PCR (data not shown) of whole ovary and isolated granulosa cell RNAs. Furthermore, plasma estradiol levels in immature ßERKO females (Fig. 2CGo) mirrored the observed Cyp19 expression pattern, including a striking increase 4 h after hCG treatment. These data indicate that preovulatory granulosa cells do not properly differentiate and develop sufficient aromatase activity in response to PMSG but then respond inappropriately to subsequent hCG treatment by dramatically increasing Cyp19 expression, as opposed to the hCG-elicited decrease observed in control and {alpha}ERKO ovaries.

ERß-null granulosa cells exhibit an attenuated response to hCG
Perhaps the most dynamic changes in ovarian gene expression occur shortly after the LH surge and can be mimicked by hCG administration in PMSG-primed mice. Several genes evaluated in this study exhibited robust induction 4 h after hCG exposure in control ovaries, only to return to near baseline levels within 10–20 h; these included amphiregulin (Areg), betacellulin (Btc), Cebpb, epiregulin (Ereg), hyaluronan synthase-2 (Has2), Pgr, Ptgs2, tissue plasminogen activator (Plat), and TNF-{alpha} induced protein 6 (Tnfaip6) (Table 1Go and Figs. 5–7GoGoGo). Such rapid and transient up-regulation in the ovary after hCG treatment or just before ovulation has been previously reported for each of these genes (1, 35). Several other genes exhibited equally impressive induction in control ovaries that was delayed until 10 h after hCG treatment and more proximal to the time of follicular rupture, including A disintegrin and metalloproteinase with thrombospondin-like repeats-1 (Adamts1), cyclin-dependent kinase inhibitor 1A (Cdkn1a), matrix metalloproteinase (Mmp) 19, and tissue-inhibitor of metalloproteinase (Timp) (Table 1Go and Fig. 5Go). Finally, several genes exhibited peak expression in the postovulatory ovary (>20 h after hCG treatment), including 17ß-hydroxysteroid dehydrogenase type VII (Hsd17b7), Mmp2, and Timp3 (Table 1Go and Fig. 5Go). Once again, the hCG induction and expression patterns observed herein for each of the above genes has been previously reported (36, 37, 38).



View larger version (46K):
[in this window]
[in a new window]
  		FIG. 5. Ovarian gene expression over the course of gonadotropin-induced ovulation in control and ERKO ovaries. Immature mice were treated with PMSG (3.25 IU) followed 48 h later with hCG (2.2 IU). Animals were killed at the indicated times over the course of gonadotropin-induced ovulation, whole ovaries collected, and total RNA extracted for gene expression analysis. Shown is the quantitative data from SQ-RT-PCR for Areg, Btc, Cebpb, Ereg, Hsd17b7, Mmp19, Timp, and Tnfaip6. All data are expressed as the average fold control-none (untreated) ± SEM. a, P < 0.05 vs. control at the same time point.

 


View larger version (97K):
[in this window]
[in a new window]
  		FIG. 6. ßERKO ovaries exhibit an attenuated induction of the Ptgs2 gene after hCG treatment during gonadotropin-induced ovulation. Immature mice were treated with PMSG (3.25 IU) followed 48 h later with hCG (2.2 IU). Animals were killed at the indicated time points over the course of gonadotropin-induced ovulation, whole ovaries collected for RNA or protein extraction, and immunohistochemistry. A, Quantitative data from SQ-RT-PCR for Ptgs2 gene expression showing the rapid rise in expression 4 h after hCG treatment that is attenuated in ßERKO ovaries. Data are expressed as the average fold control-none (untreated) ± SEM. B, Northern blot analysis for Ptgs2 gene expression in whole ovaries or granulosa cells isolated from control and ßERKO females during the indicated treatments and time points. Ptgs2 expression peaks at 4 h after hCG and decreases precipitously thereafter; Ptgs2 transcripts are essentially undetectable in ßERKO ovaries 4 h after hCG treatment. C, Western blot analysis for Ptgs2 protein in control, {alpha}ERKO, and ßERKO ovaries 48 h after PMSG treatment and 4 h after subsequent hCG treatment. D, Immunohistochemistry for Ptgs2 in control, {alpha}ERKO, and ßERKO ovaries collected 4 and 10 h after hCG treatment. Shown are representative photomicrographs of low- and high-power magnification for each genotype. Ptgs2 immunoreactivity is clearly localized to the granulosa cells of preovulatory follicles in control and {alpha}ERKO ovaries and peaks at 4 h after hCG but remains 10 h after hCG treatment. The representative ßERKO ovaries illustrate the severely attenuated induction of Ptgs2 in the granulosa cells of preovulatory follicles. Some Ptgs2 immunoreactivity is detectable in ßERKO preovulatory follicles, but this is generally restricted to the cumulus-oophorus. a, P < 0.05 vs. control at the same time point.

 


View larger version (71K):
[in this window]
[in a new window]
  		FIG. 7. ßERKO ovaries exhibit an attenuated induction of the Pgr gene after hCG treatment during gonadotropin-induced ovulation. Immature mice were treated with PMSG (3.25 IU) followed 48 h later with hCG (2.2 IU). Animals were killed at the indicated time points over the course of gonadotropin-induced ovulation, whole ovaries collected for RNA or protein extraction, and immunohistochemistry. A, Quantitative data from SQ-RT-PCR for Pgr gene expression showing the rapid rise in expression 4 h after hCG treatment that is attenuated in ßERKO ovaries. Data are expressed as the average fold control-none (untreated) ± SEM. B, Northern blot analysis for Pgr gene expression in whole ovaries or granulosa cells isolated from control and ßERKO females during the indicated treatments and time points. Pgr expression peaks at 4 h after hCG and is undetectable at all other time points. Pgr transcripts are essentially undetectable in ßERKO ovaries 4 h after hCG treatment. C, Immunohistochemistry for PR in control, {alpha}ERKO, and ßERKO ovaries collected 4 h after hCG treatment. Shown are representative photomicrographs of low- and high-power magnification for each genotype. PR immunoreactivity is clearly localized to the granulosa cells of preovulatory follicles in control and {alpha}ERKO ovaries. The representative ßERKO ovaries illustrate the severely attenuated induction of PR in the granulosa cells of preovulatory follicles. D, Quantitative data from SQ-RT-PCR for Adamts1 gene expression, a PR-regulated gene in the ovulatory follicle, showing the lack of increased expression 12 h after hCG treatment in ßERKO ovaries. Data are expressed as the average fold control-none (untreated) ± SEM.

 
In general, immature {alpha}ERKO ovaries exhibited a control-like expression pattern for all of the genes evaluated (Table 1Go and Fig. 5Go). Some minor differences in the level of peak expression were observed in {alpha}ERKO vs. control ovaries, such as an above-normal increase in Areg and Tnfaip6 expression at 4 h post hCG and Timp expression at 10 h after hCG and a slightly lower increase in Mmp19 expression at 20 h after hCG (Fig. 5Go). One striking abnormality observed in {alpha}ERKO ovaries was premature induction of Hsd17b7 expression with PMSG alone such that levels were increased more than 15-fold vs. approximately 3-fold in control ovaries. The Hsd17b7 gene encodes the enzyme 17ß-HSD VII, which was recently described as specific to luteal cells and necessary for the final step in estradiol synthesis in the corpus luteum (38). Immature control ovaries exhibited a slight increase in Hsd17b7 expression after PMSG treatment but a much greater response after hCG, which elicited a steady induction that peaked at more than 12-fold 20 h after hCG (Fig. 5Go). Immature {alpha}ERKO ovaries also expressed considerable levels of Hsd17b3 (Table 1Go); which encodes a Leydig cell-specific enzyme involved in testosterone synthesis and is previously described to be uniquely expressed in {alpha}ERKO ovaries (15).

Immature ßERKO ovaries exhibited appropriate hCG induction of several genes, including Areg, Cebpb, Ereg, and Has2. However, hCG failed to fully induce expression of several other genes in ßERKO ovaries relative to control, including Adamts1, Btc, Mmp19, Pgr, Ptgs2, and Timp (Figs. 5–7GoGoGo). Interestingly, Timp and Mmp19 expression in ßERKO ovaries did eventually increase to control-like levels but not until 20 h after hCG, a considerable delay, compared with the peak at 4 h after hCG observed in control ovaries (Fig. 5Go).

Ptgs2 and PR are both obligatory to follicular rupture (10, 12), and each exhibited a severely blunted response to hCG in ßERKO ovaries. Therefore, the expression pattern of Ptgs2 and Pgr in control and ßERKO ovaries after hCG treatment was evaluated in greater detail. SQ-RT-PCR analysis indicated that Ptgs2 mRNA levels were induced in ßERKO ovaries 4 h after hCG treatment but that peak levels were more than 60% lower (P < 0.05) than that observed in control and {alpha}ERKO ovaries (Fig. 6AGo). These data were confirmed by Northern blot analysis, which demonstrated that Ptgs2 transcripts were barely detectable in ßERKO ovaries and isolated granulosa cells collected 4 h after hCG treatment of PMSG-primed mice vs. the extraordinary induction observed in similarly treated control ovaries (Fig. 6BGo). Western blot analysis of cytosolic protein preparations from control, {alpha}ERKO, and ßERKO ovaries further illustrated that the levels of immunoreactive Ptgs2 (cyclooxygenase-2) after hCG induction were barely detectable in ßERKO ovaries but easily observed in control and {alpha}ERKO ovaries (Fig. 6CGo). Immunohistochemical detection of Ptgs2 in control and {alpha}ERKO ovaries collected at 4 and 10 h after hCG treatment indicated that expression was localized to the granulosa cells of ovulatory and ovulating follicles in both genotypes (Fig. 6DGo). Ptgs2 immunoreactivity was especially strong among granulosa cells lining the antrum and forming the cumulus oophorus (Fig. 6DGo), in agreement with previous descriptions of ovarian Ptgs2 expression during induced ovulation in mice (39). In contrast, a minimal number of granulosa cells of ßERKO preovulatory follicles 4 and 10 h after hCG treatment exhibited Ptgs2 immunoreactivity, and these cells were sparse and exhibited no apparent pattern (Fig. 6DGo). Although a discernible range of Ptgs2 expression was apparent when comparing ßERKO preovulatory follicles, even within the same ovary, none exhibited control-like levels of immunoreactivity. Furthermore, the preovulatory rise in Tnfaip6 expression (Fig. 5Go), which is reported to be partially dependent on prerequisite increases in Ptgs2 function within the ovulating follicle (40), was blunted in ßERKO ovaries vs. control ovaries at 10 h post-hCG treatment, suggesting insufficient Ptgs2 action in the former (Fig. 5Go).

Assays for Pgr expression over the course of induced ovulation in mice indicated an extraordinary increase in Pgr mRNA levels at 4 h post-hCG treatment in PMSG-primed mice, followed by a return to barely detectable levels within 6 h (Fig. 7Go). This dramatic and transient induction of Pgr expression after hCG has been well described (1, 19) and is critical to follicular rupture (12). In turn, although Pgr mRNA levels were increased in ßERKO ovaries 4 h after hCG treatment, the level of peak induction was more than 70% less (P < 0.05) than that observed in control ovaries (Fig. 6AGo) when assessed by SQ-RT-PCR. Northern blot analysis of RNA from whole ovaries or isolated granulosa cells of control mice confirmed the induction of Pgr expression 4 h after hCG treatment in control females but only barely detectable levels in similarly treated ßERKO ovaries and granulosa cells (Fig. 7BGo). Immunohistochemistry for PR immunoreactivity during induced ovulation indicated a pattern that was strikingly similar to that of Ptgs2 (Fig. 7CGo). Control and {alpha}ERKO ovaries collected 4 h after hCG treatment possessed multiple growing follicles; however, only preovulatory follicles exhibited strong PR immunoreactivity (Fig. 7CGo). PR expression was evenly distributed throughout the mural granulosa cells but notably absent from the cumulus oophorus in both genotypes, in agreement with previous descriptions of hCG-induced PR expression in preovulatory follicles (41). In contrast, the multiple preovulatory follicles present in immature ßERKO ovaries collected 4 h after hCG exposure exhibited minimal PR immunoreactivity (Fig. 7CGo). Because expression of the Adamts1 gene is reported to depend on increased PR functions in the preovulatory follicle (36), we were interested in determining whether expression of this gene in ßERKO preovulatory follicles was altered as well. Interestingly, ßERKO ovaries exhibited a normal initial rise in Adamts1 mRNA levels after hCG exposure but failed to exhibit any further increase thereafter vs. a more than 4-fold rise observed in control ovaries 10 h after hCG (Fig. 7DGo). Recent studies have indicated a bimodal mechanism of Adamts1 induction after hCG exposure or the LH surge such that the acute increase in expression (within 4 h) is LH dependent, but the more dramatic increase that provides for peak expression at 10 h after hCG is PR dependent (36). These data support our findings that PR functions in ßERKO ovaries after hCG treatment are insufficient.

PR expression is reduced in PMSG-treated ER{alpha}-null ovaries
To date, most investigations of ovarian PR expression during folliculogenesis have focused on the dramatic rise that occurs in granulosa cells shortly after the LH surge or hCG treatment. However, we observed a small but measurable (2-fold) increase in Pgr mRNA levels in control ovaries 48 h after PMSG treatment (Fig. 8AGo). Furthermore, Western blot analyses indicated that both PR-A and PR-B were detectable in whole ovarian homogenates after PMSG treatment (Fig. 8BGo). However, {alpha}ERKO ovaries exhibited reduced basal Pgr mRNA levels relative to control ovaries and no change after PMSG treatment when assessed by SQ-RT-PCR (Fig. 8AGo). Western blot analysis confirmed that PR-A and PR-B were barely detectable before and after PMSG treatment in {alpha}ERKO ovaries, whereas both forms were easily observed in preparations of PMSG-treated control ovaries (Fig. 8BGo). ßERKO ovaries exhibited slightly higher basal levels of Pgr mRNA and minimal induction after PMSG treatment, but levels of PR protein in ßERKO ovaries were comparable with control before and after PMSG treatment (Fig. 8BGo). Due to high background, it was difficult to localize the basal and PMSG-induced PR expression by immunohistochemistry in ovaries from any one of the three genotypes (data not shown). Still, these data are the first to demonstrate that constitutive PR expression in the murine ovary is ER{alpha} dependent.



View larger version (44K):
[in this window]
[in a new window]
  		FIG. 8. {alpha}ERKO ovaries exhibit reduced Pgr expression 48 h after PMSG treatment. Immature mice were treated with PMSG (3.25 IU) followed 48 h later with hCG (2.2 IU). Animals were killed at the indicated time points over the course of gonadotropin-induced ovulation and whole ovaries collected for RNA and protein extraction. A, Quantitative data from SQ-RT-PCR for Pgr gene expression showing a 2-fold induction after PMSG treatment in control ovaries but lower basal levels and no increase in similarly treated {alpha}ERKO ovaries. Data are expressed as the average fold control-none (untreated) ± SEM. B, Western blot analysis for PR protein in control, {alpha}ERKO, and ßERKO ovaries 48 h after PMSG treatment illustrating that PR-A and PR-B are undetectable in {alpha}ERKO ovaries but clearly present in control and ßERKO ovaries after PMSG treatment. A nonspecific (n.s.) immunoreactive band migrates just above the PR-B band and appears at equal amounts in all samples, thereby acting as an internal loading control. Ut, Wild-type mouse uterus to serve as a positive control for PR. a, P < 0.05 vs. control at the same time point.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The primary cause of reduced fecundity in ERß-null female mice is infrequent and inefficient ovulation (13, 42). In the current study, we provide evidence that the ovulatory phenotype in ßERKO females is due to the loss of intraovarian ERß functions that facilitate gonadotropin-induced differentiation of preovulatory follicles. Two classic indices of preovulatory granulosa cell differentiation, increased Cyp19 expression and aromatase activity and acquired Lhcgr expression and LH responsiveness, were markedly deficient in ßERKO ovaries after PMSG treatment. As a result, the response of PMSG-primed ßERKO ovaries to an ovulatory dose of hCG was severely attenuated, as evident by a reduced rate of ovulation, insufficient induction of Ptgs2 and PR functions, an aberrant increase in Cyp19 expression and plasma estradiol, and incomplete expansion of the COC. Our parallel characterization of ER{alpha}-null females indicates that loss of intraovarian ER{alpha} function had minimal impact on gonadotropin-induced ovulation and the concomitant changes in gene expression, although some abnormalities were revealed. Therefore, we conclude that ERß is more important in mediating the intraovarian actions of estradiol that are critical to maximizing the follicular response to gonadotropins.

ERß-null follicles fail to differentiate after PMSG treatment
FSH is undoubtedly the primary stimulus for differentiation of follicles from the preantral to preovulatory stage. Female mice null for FSH signaling due to the loss of hormone (6, 43) or receptor (7, 44) exhibit follicles that fail to differentiate beyond the preantral stage and produce minimal estradiol and hence are infertile. However, FSH alone does not provide for complete follicle and granulosa cell differentiation in rodents. Instead, the synergistic actions of estradiol are required to maximize FSH induction of antrum formation (27), Cyp19 expression and aromatase activity (20, 21, 22), and LH responsiveness (25, 26, 45). Because ERß is the more prominent ER form present in rodent granulosa cells (16, 17, 18) and exhibits an expression pattern similar to FSH receptor, it is the most plausible candidate to mediate the cooperative actions of estradiol with FSH. Our thorough analysis of induced ovulation in immature ßERKO ovaries supports this role of ERß in amplifying the granulosa cell response to FSH.

Substantial Cyp19 expression and estradiol synthesis is a hallmark of healthy preovulatory follicles in the mammalian ovary (46). FSH stimulation of the Cyp19 gene is augmented by estradiol (20, 21, 22) and several findings suggest this to be an ERß-mediated effect, including the highly correlative expression pattern of ERß and Cyp19 in granulosa cells (47, 48, 49); the preservation of synergism in isolated granulosa cells, ruling out a paracrine type mechanism; and the inhibition of FSH-induced aromatase activity in granulosa cells by ER antagonists (50). Our data provide the most definitive evidence to date that ERß mediates this effect of estradiol and is required for maximum FSH stimulation of aromatase activity by demonstrating that PMSG-stimulated ßERKO ovaries and granulosa cells exhibit reduced Cyp19 expression and that ßERKO females fail to present the expected preovulatory rise in plasma estradiol. This conclusion is further supported by our recent findings that individual ßERKO follicles maintained in vitro secrete significantly less estradiol relative to wild-type follicles when stimulated with FSH (51). Furthermore, {alpha}ERKO ovaries (Fig. 4Go) (15) and individual {alpha}ERKO follicles (51) exhibit elevated Cyp19 expression and estradiol synthesis, excluding a direct role for ER{alpha}.

Direct estradiol/ERß regulation of the Cyp19 gene is unlikely because there are no known EREs in the human or rat Cyp19 promoter II and estradiol alone has no stimulatory effect (20). In turn, FSH stimulation of aromatase activity in granulosa cells involves the binding of cAMP/protein kinase A (PKA)-activated cAMP response element binding protein (CREB) and steroidogenic factor (SF)-1 to their cognate response elements within the Cyp19 promoter II (19, 52, 53, 54). LRH-1, which may substitute for SF-1 in granulosa cells (55), and GATA-4 (56) have also been postulated to be involved. We found that SF-1 (Nr5a1), LRH-1 (Nr5a2) and Gata4 expression were not overtly altered in immature ßERKO ovaries during gonadotropin-induced ovulation (Table 1Go), nor were the levels of CREB protein (data not shown).

It should be noted that PMSG-stimulated ßERKO females did exhibit a slight but detectable rise in plasma estradiol levels despite reduced Cyp19 expression. Furthermore, we previously reported that adult ßERKO females exhibit an estrogenized reproductive tract (13) and relatively normal plasma estradiol levels (15). Although FSH action alone may account for some Cyp19 expression in adult ßERKO ovaries, it is likely that thecal cell-derived androgens may activate granulosa cell androgen receptor (AR) and partially compensate for the loss of ERß’s cooperative action with FSH. AR-mediated testosterone and dihydrotestosterone actions are as effective as estradiol in enhancing FSH induction of Cyp19 expression in rat granulosa cells (20, 57). We previously reported that androgen synthesis is normal in ßERKO ovaries (15). Furthermore, AR immunoreactivity is reportedly elevated in ßERKO granulosa cells (58), although we were unable to substantiate these findings and found relatively normal Ar mRNA levels in ßERKO ovaries in vivo (Table 1Go) and ßERKO follicles in vitro (51). Interestingly, androgen/FSH actions may also promote the elevated Cyp19 expression and estradiol synthesis that is characteristic of {alpha}ERKO ovaries (Figs. 2Go and 4Go) (15), which possess the capability to synthesize extremely high levels of testosterone (15).

Another imperative characteristic of preovulatory granulosa cells is the acquisition of LH-receptor and responsiveness, which provides the mechanism by which follicles that are suitable for ovulation are selected to respond to the LH surge and ovulate. Like the Cyp19 gene, induction of Lhcgr expression in preovulatory granulosa cells is primarily stimulated by FSH but requires prior exposure to estradiol (23, 24, 25, 59, 60). However, unlike Cyp19 regulation, FSH-induced Lhcgr expression is augmented by estradiol only (24, 25, 60), whereas nonaromatizable androgens have little effect (25, 60). Further evidence that FSH-induced Lhcgr expression is dependent on estradiol comes from demonstrations that cotreatment of estrogen-primed granulosa cells with FSH and aromatase inhibitors (26), ER-antagonists (45), or estradiol-specific antisera (24) blocks FSH induction of LH-receptor expression and that Lhcgr mRNA levels are decreased 5-fold in the ovaries of Cyp19-null mice despite elevated FSH and testosterone (61). Herein we provide strong evidence that ERß mediates the synergistic effect of estradiol on FSH regulation of LH receptor levels in preovulatory granulosa cells by demonstrating that PMSG-primed ßERKO females exhibit a severely attenuated ovulatory and genomic response to hCG and that preovulatory granulosa cells isolated from ßERKO ovaries 48 h after PMSG treatment possess reduced Lhcgr mRNA levels relative to wild type. We recently reported similar data in individual ßERKO follicles grown and propagated in vitro, which also exhibit reduced Lhcgr expression and respond poorly to hCG-induced follicle rupture (51). In contrast, granulosa cells in {alpha}ERKO ovaries exhibited abnormally high levels of Lhcgr mRNA that are maintained even after the follicles become enlarged and cystic (32).

The complexity of the Lhcgr promoter is illustrated by the need to maintain constitutive expression in thecal cells, while allowing for FSH/estradiol mediated increases in granulosa cells. FSH induction of the Lhcgr gene in granulosa cells occurs via adenylyl cyclase/cAMP signaling (62, 63, 64). However, unlike the Cyp19 promoter II, the proximal promoters of the human and rat Lhcgr genes lack a consensus cAMP response element but possess multiple novel cAMP response elements within a GC-rich region that are able to confer cAMP induction (65). This same region of the rat Lhcgr promoter contains an estrogen response element (ERE) half-site but does not bind ER or respond to estradiol and is in fact inhibitory to transcription in vitro (66). Interestingly, FSH induction of Lhcgr but not Cyp19 expression is lost in estrogen-primed granulosa cells when dispersed in culture but is restored on estradiol-induced reaggregation, suggesting that estradiol may augment FSH induction of Lhcgr expression by stimulating and/or maintaining cell-cell interactions between granulosa cells (60). Indeed, estradiol is known to increase the number of gap-junctions among granulosa cell populations (28, 29), and this may be ERß mediated.

In essence, several aspects of the ßERKO ovarian phenotype after gonadotropin-induced ovulation are similar to but less severe than those described in mice null for FSH signaling (6, 7). Our findings are supported by a recent report that an ERß-specific agonist (8ß-VE2) but not an ER{alpha}-specific agonist (16{alpha}-LE2) effectively mimics estradiol treatment during gonadotropin-induced ovulation in hypophysectomized rats (67). Previous work has shown that estradiol has little effect on FSH receptor levels in preovulatory granulosa cells (19, 59), and our demonstration of normal Fshr mRNA levels in untreated and PMSG-stimulated ßERKO ovaries agree (Fig. 3Go). Instead, the long-recognized synergism of estradiol on FSH-induced granulosa cell differentiation is likely due to its capacity to enhance the levels (19, 25, 59, 68) and action of intracellular cAMP after FSH stimulation (25, 59). Indeed, in vitro and in vivo studies have shown cooperative synergism between the ER and cAMP/PKA signaling pathway (69), including functional interaction between ERß and CREB (70) or CREB-binding protein (71, 72), both of which are implicated in Cyp19 regulation in granulosa cells (20, 53, 54). Interestingly, GnRH stimulation of the cAMP/PKA pathway leading to phosphorylation of CREB was recently shown to require functional ERß in hypothalamic neurons (73), suggesting that similar cooperative action may exist in other cell types. Therefore, given the extensive evidence that estradiol is required to facilitate the ovarian response to FSH, it is not surprising that the ßERKO ovarian phenotype is similar to those of mice null for FSH signaling, indicating the requirement of ERß in this synergism.

ERß-null follicles exhibit a compromised response to hCG
The evidence of disrupted follicle rupture in ERß-null female mice is indisputable because this phenotype is twice reported previously (13, 42) and confirmed herein. Congruent with the earlier studies, we found the average oocyte yield among immature ßERKO females to be less than 30% that of controls. Increasing the dose of hCG did improve the average oocyte yield among ßERKO females but yields never equaled that of similarly treated control females. As in the previous studies (13, 42), we also observed a broad range in oocyte yield among individual ßERKO females such that some exhibited a control-like response, whereas almost one quarter did not ovulate at all. This variation in oocyte yield is mirrored by the results of fertility studies in ßERKO females, in which some females exhibit a normal frequency of pregnancy but reduced litter size (i.e. subfertile), whereas others fail to exhibit even a single pregnancy over the course of several months (i.e. infertile) (13, 42). Such variation is difficult to explain but remains consistent between the two reported lines of ERß-null mice (13, 42).

The periovulatory period in the rodent ovary is marked by dramatic temporal and quantitative changes in gene expression. A recent study using serial analysis of gene expression in the mouse ovary revealed that more than 700 genes exhibit a significant change in expression within 12 h after hCG administration (74). Herein we chose to limit our evaluation to the expression of several genes that are well documented to dramatically increase in the periovulatory rodent ovary 4 or 12 h after stimulation by hCG or LH. In general, PMSG-primed {alpha}ERKO and ßERKO females exhibited the exp