| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Division of Reproductive Sciences, Oregon Regional Primate Research Center, Beaverton, Oregon 97006
b Department of Physiology and Pharmacology, Oregon Health Sciences University, Portland, Oregon 97201
ABSTRACT
Progesterone produced in response to the midcycle gonadotropin surge is essential for ovulation and luteinization of the primate follicle. Because cell-cycle arrest is associated with the initiation of luteinization, this study was designed to determine the dynamics and regulation of granulosa cell proliferation by gonadotropin and progesterone during the periovulatory interval in the primate follicle. Granulosa cells or ovaries were obtained from macaques undergoing controlled ovarian stimulation either before (0 h) or as long as 36 h following the administration of an ovulatory hCG bolus with or without a 3ß-hydroxysteroid dehydrogenase inhibitor with or without a nonmetabolizable progestin. The percentage of cells staining positive for Ki-67, a nuclear marker for cell proliferation, decreased (P < 0.05) within 12 h of hCG administration in a steroid-independent manner. Levels of cyclin D2 and E mRNA did not decline during the periovulatory interval; however, cyclin B1 mRNA was reduced significantly by 12 h. Steroid depletion increased (P < 0.05) cyclin B1 mRNA at both 12 and 36 h post-hCG and was reversible by progestin replacement at 36 h. The cyclin-dependent kinase inhibitor p21Cip1 was transiently increased 12 h post-hCG, whereas p27Kip1 mRNA levels increased at 36 h in a steroid-independent fashion. These data suggest that a gonadotropin bolus inhibits mitosis in granulosa cells early (12 h) in the periovulatory interval, whereas progesterone may play a later, antiproliferative role in luteinized cells of primates.
corpus luteum, granulosa cells, human chorionic gonadotropin, ovulation, progesterone
INTRODUCTION
Cell proliferation or progression through the cell cycle [1] is controlled by stimulatory and inhibitory factors. Cyclins promote cell-cycle progression by interacting with specific cyclin-dependent kinases (cdk) to induce phosphorylation of target proteins. Cyclin D/cdk4,6 is essential for the early transition through the first gap phase of the cell cycle (G1), whereas cyclin E/cdk2 is important in movement across the G1/DNA synthesis (S) boundary. Cyclin B/cdk1 activity is associated with the second gap phase (G2) exit and mitosis [1]. The activity of these complexes is inhibited through interaction with specific proteins termed cyclin-dependent kinase inhibitors (CKI). Two groups of CKI have been characterized to date, the Ink4 and the Cip/Kip families, both of which have broad substrate specificity [1]. Whereas many current concepts regarding control of the cell cycle have emerged from studies of cancer cells [2], investigators are examining their relevance to cycles of cell proliferation and differentiation in normal tissues, including follicular growth and luteinization in the ovarian cycle [3].
Follicle rupture and formation of the corpus luteum is the end result of a cascade of events initiated by the ovulatory gonadotropin surge. In nonprimate species, luteinization of granulosa cells is associated with reduced proliferative capacity [4–7], and this may be an obligatory step in formation of the corpus luteum. Recent gene knockout models have provided valuable information regarding the function of key cell-cycle regulatory machinery during follicle growth and differentiation. Notably, granulosa cells from cyclin D2-/- mice are unable to proliferate in response to FSH, although they do luteinize following an LH bolus [8]. In contrast, ovarian follicles grow normally following FSH stimulation in p27-null animals, but luteal cells continue to proliferate following an ovulatory stimulus [9]. Follicles from cdk4-deficient mice develop and luteinize normally, but they display impaired ovulation [10]. Thus, as Robker and Richards [6] have suggested, follicular growth, differentiation, and ovulation require the coordinated actions of cyclins, cdk, and CKI.
An ovulatory stimulus in rodent species attenuates granulosa cell proliferation [3], but this same action of gonadotropins has not been well established in species with a long periovulatory interval and a functional luteal phase, such as primates. Nor is it known if this action is a direct effect of the gonadotropin surge or is mediated by locally produced intrafollicular factors. In all mammalian species examined to date, LH-stimulated progesterone is essential for ovulation (e.g., mouse [11], rat [12], ewe [13], monkey [14]), but to our knowledge, relatively few studies have addressed the possibility of a function for this steroid in early luteal formation and function. A role for progesterone was suggested by Chaffkin et al. [15, 16], who showed that progesterone inhibited in vitro proliferation of human granulosa cells obtained following an ovulatory stimulus. Progesterone also exerts significant cell-cycle control in breast cancer cells via the regulation of specific cell-cycle components [2, 17].
The goals of the present study were to determine the time course of changes in the proliferation of granulosa cells from preovulatory follicles following an ovulatory stimulus in primates undergoing controlled ovarian stimulation and its regulation by gonadotropins versus progesterone. Granulosa cell proliferation was assessed with Ki-67 immunohistochemistry, and mRNA levels of cyclins B1, D2, E as well as cdk inhibitors p21Cip1 and p27Kip1 were used as markers of gonadotropin and progestin regulation of cell-cycle machinery.
MATERIALS AND METHODS
Animals
The general care and housing of monkeys at the Oregon Regional Primate Research Center (ORPRC) was described previously [18]. Animal protocols and experiments were approved by the ORPRC Animal Care and Use Committee, and studies were conducted in accordance with the Guide for the Care and Use of Laboratory Animals [19]. Adult female rhesus monkeys were treated with recombinant human (r-h) gonadotropins (r-hFSH ± r-hLH for 9 days; Laboratoires Serono SA, Aubonne, Switzerland) to promote the development of multiple preovulatory follicles. Antide (Serono) was also administered daily to prevent an endogenous LH surge, as described in detail elsewhere [20]. Animals were assigned randomly to receive no ovulatory stimulus or 1000 IU of r-hCG (single injection i.m.; Serono) to initiate periovulatory events. Ovaries were removed or follicles greater than 4 mm aspirated during laparotomy of anesthetized animals either the morning after the last LH/FSH treatment (0 h) or 12 and 36 h following administration of 1000 IU of r-hCG (n = 4, 3, and 3 monkeys/time point, respectively). An additional group of monkeys (n = 3 per time point) was stimulated in an identical fashion but also received the 3ß-hydroxysteroid dehydrogenase (3ß-HSD) inhibitor trilostane (TRL; Sanofi Research Division, Malvern, PA) orally (1 g in 8 ml of orange Kool-Aid [Kraft General Foods, Inc., White Plains, NY] containing 1% [w/v] gum tragacanth [Sigma, St. Louis, MO]) beginning 4 h before hCG administration and for every 12 h thereafter until the time of ovariectomy. A third group of animals (n = 3 per time point) received TRL plus the nonmetabolizable progestin R5020 (2.5 mg in sesame oil s.c. once daily starting at the time of hCG; Promegestrone; DuPont/NEN, Boston, MA). We have shown previously that this treatment suppresses completely the periovulatory rise in intrafollicular progesterone concentration and blocks ovulation, and that coadministration of R5020 with TRL is sufficient to restore ovulation [14, 21].
Immunohistochemistry for Ki-67
Following ovariectomy, half of one ovary was microwaved in 0.5 ml of Hanks balanced salt solution for 7 sec, then chilled on ice in 10% sucrose dissolved in 0.1 M PBS, embedded in Tissue Tek II Optimal Cutting Temperature mixture (OCT; Miles, Inc., Elkhart, IN), frozen in liquid propane, and cryosectioned at 15 µm. The remaining ovary was processed for other studies [22]. The proportion of granulosa cells entering the cell cycle was determined as previously reported using Ki-67 antigen, a nuclear matrix protein expressed in proliferating but not in quiescent cells [23–25]. Cryosections were microwaved for an additional 3 sec before fixation (15 min at 4°C) in 2% (w/v) paraformaldehyde, 1.22% (w/v) picric acid, 1.5% (w/v) polyvinyl-pyrrolidone (PVP), and 0.1 M phosphate buffer (pH 7.3). Postfixed tissues were then treated with 85% (w/v) ethanol/1.5% (w/v) PVP, followed by PBS/1.5% PVP. Sections were rinsed with 0.37% glycine in PBS/PVP to eliminate aldehyde groups, and after additional PBS/1.5% PVP rinses, tissues were incubated with glucose oxidase (Sigma) to quench endogenous peroxidase activity at a concentration of 1 U/ml for 45 min at 4°C. The tissues were washed in PBS/1.5% PVP/0.1% (w/v) gelatin and incubated in 2% normal horse serum for 20 min. Pretreatment of tissues with avidin/biotin blocking kit (Vector Laboratories, Burlingame, CA) was used to block endogenous biotin or biotin-binding proteins (15 min each of biotin and avidin). Incubation with a specific monoclonal antibody to Ki-67 (10 pg/µl; Dako Corp., Carpinteria, CA) was performed overnight at 4°C. Controls included the use of an irrelevant primary antibody against timothy grass pollen [25]. Tissues were rinsed with PBS/0.1% gelatin/0.075% (w/v) BRIJ 35 (lauryl ether; Sigma) and blocked with additional horse serum for 20 min before incubation with the biotinylated horse anti-mouse IgG (1:100 dilution) for 30 min at room temperature. Detection of the biotinylated secondary antibody was performed using avidin-biotin peroxidase (Vector) for 60 min at room temperature. After subsequent washes with PBS/0.1% gelatin and 0.038 M Tris buffer, sections were treated with 0.025% diaminobenzidine tetrahydrochloride (Dojindos DAB; Wako Chemicals, Richmond, VA) in 0.038 M Tris buffer and 0.001% (w/v) H2O2 (pH 7.6). Sections were rinsed several times with water and then exposed to 0.026% (w/v) osmium tetroxide for 1 min to increase stain intensity, followed by washes with water before postfixation in 2% paraformaldehyde/1.22% picric acid. Slides were then rinsed with several changes of water and counterstained with hematoxylin for 15 min. Sections were dehydrated in a graded series of ethanol, cleared with xylene, and cover-slipped.
The percentage of Ki-67-positive granulosa cells was determined by averaging the proportion of positive cells from four regions (100 cells/region) within healthy, large antral (diameter, 
4 mm) follicles (excluding cumulus cells). Data from individual follicles were pooled among animals within a treatment and expressed as the percentage (mean ± SEM) of Ki-67-positive granulosa cells per treatment group (n = 3–9 follicles/treatment group).
Total RNA Isolation and Reverse Transcription-Polymerase Chain Reaction
Granulosa cell isolation and reverse transcription-polymerase chain reaction (RT-PCR) are described in detail elsewhere [26]. The concentration of MgCl2, primers, cDNA, and the number of cycles were determined empirically as part of the validation process. Temperature profile of the PCR was denaturation at 94°C for 30 sec, annealing at 60°C for 1 min, and extension at 72°C for 1 min. Primers sequences were as follows: cyclin D2, up: 5'-TCATGACTTCATTGAGCA, dn: 5'-CACTTCCTCATCCTGCTG [27]; cyclin B1, up: 5'-GAAGTGACTGGAAACATG, dn: 5'-AGTATGTTGCTCGACATC [27]; cyclin E, up: 5'-AATAGAGAGGAAGTCTGG, dn: 5'-AGATATGCAACCTGCATG [27]; p27, up: 5'-AGGATGTCAGCGGGAGCCGG, dn: 5'-CTTCTTGGGCGTCTGCTCCA [28]; and p21; up: 5'-ACTGTGATGCGCTAATGGC, dn: 5'-ATGGTCTTCCTCTGCTGTCC [29]. Aliquots of each PCR reaction were electrophoresed through a 2% agarose Gel stained with 0.1 µG/ml of ethidium bromide. Gels were visualized on an ultraviolet transilluminator, photographed using 667 Polaroid (Cambridge, MA) film, and the photographs analyzed by densitometry. All values were normalized to the internal standard glyceraldehyde-3-phosphate dehydrogenase (GAPDH) [26]. Validation of the RT-PCR assay was performed using RNA from granulosa cells aspirated 27 h following hCG during routine in vitro fertilization (IVF) protocols [18] (data not shown). The amount of coamplified product for experimental and internal standard primer sets was linear and parallel with increasing amount of cDNA, and both sets of primers were in the exponentially increasing phase relative to the number of cycles. To control for interassay variability, total RNA from granulosa cells of three monkeys was combined and reverse transcribed as described to form a pool that was amplified in triplicate during each PCR with the appropriate set of primers. The coefficient of variation calculated using the triplicate pool samples was typically less than 15%. Sequence analysis by the Molecular and Cellular Biology Core Laboratory at the ORPRC confirmed the identity of the PCR products. Because data for each set of primers was collected in two to three rounds of PCR reactions, the pool triplicates were also used to normalize data between reactions.
Statistical Analysis
The percentage of Ki-67-positive granulosa cells was analyzed by one-way ANOVA followed by the Newman-Keuls test. To test for heterogeneity of variance, RT-PCR data were subjected to a Bartlett 
2 test and subsequently transformed (to log+2) before analyses. Control values at various time points were analyzed by one-way ANOVA followed by Newman-Keuls test for comparison between means. Because TRL and TRL + R5020 data were collected at only 12 and 36 h post-hCG, separate comparisons were made between treatments within a time point for both Ki-67 and RT-PCR data by one-way ANOVA followed by Newman-Keuls test. Differences were considered to be significant at P < 0.05. Values are presented as the mean ± SEM.
RESULTS
Ki-67 Immunocytochemistry
The Ki-67-positive cells were detected in the granulosa and theca layers of healthy follicles before and throughout the 36-h interval after administration of an ovulatory hCG stimulus (Fig. 1, A and B). The use of anti-timothy grass pollen in place of anti-Ki-67 resulted in the absence of DAB precipitate (data not shown). No differences in Ki-67 staining were noted with respect to mural versus antral granulosa cells at any time point. Whereas many cells in the granulosa layer stained positive for Ki-67, fewer cells in theca tissue, including follicular vasculature, appear to stain for Ki-67 (Fig. 1A).
|
Approximately half the total cells in the granulosum before an ovulatory stimulus were Ki-67-positive (Fig. 1C). The percentage of Ki-67-positive granulosa cells decreased significantly by 12 h (P < 0.05) and remained at this level 36 h post-hCG. Treatment with trilostane did not change the percentage of Ki-67-positive cells 12 h post-hCG. However, by 36 h, Ki-67-positive granulosa cells were not observed following TRL treatment (P < 0.05), whereas R5020 returned the percentage of positive cells to control levels.
Cyclin mRNA
Cyclin D2 mRNA was detectable by RT-PCR in granulosa cells before hCG administration and increased modestly (1.8-fold, P < 0.05) by 12 h after an ovulatory stimulus (P < 0.05) (Fig. 2A). By 36 h post-hCG, cyclin D2 mRNA levels were not different from those at the 0- or 12-h time points. Administration of TRL with or without R5020 did not alter cyclin D2 mRNA levels at 12 or 36 h.
|
Cyclin E mRNA was also present in all samples before hCG administration and levels did not change following the ovulatory stimulus (Fig. 2B). Steroid depletion increased cyclin E mRNA 12 h post-hCG (P < 0.05), whereas R5020 returned levels to those of time-matched controls. Steroid depletion tended (P = 0.09) to increase levels at 36 h, but TRL with R5020 resulted in mRNA levels not different from those of time-matched controls or TRL treatment groups.
In contrast, cyclin B1 mRNA levels were highest in granulosa cells before hCG administration and declined threefold by 12 h post-hCG (P < 0.05) (Fig. 3, A and B). Steroid depletion following TRL treatment resulted in significantly elevated levels of cyclin B1 mRNA 12 h after hCG administration (P < 0.05), although R5020 did not reverse this effect. Cyclin B1 mRNA tended (P = 0.14) to increase 36 h after hCG with TRL administration, whereas TRL with R5020 treatment resulted in a significant decrease versus TRL alone (P < 0.05).
|
p21 and p27 mRNA Levels
Granulosa cell levels of p21 mRNA increased ninefold by 12 h (P < 0.05) (Fig. 4A) and were intermediate between 0- or 12-h values at 36 h. Steroid depletion did not alter p21 mRNA levels at 12 or 36 h, although at 12 h, TRL with R5020 tended (P = 0.08) to reduce mRNA levels.
|
In granulosa cells aspirated before an ovulatory stimulus, p27 mRNA was detectable by RT-PCR in only one of four samples (Fig. 4B). Whereas p27 mRNA levels did not change between 0 and 12 h post-hCG, these levels did increase by 22-fold near the expected time of ovulation (36 h) relative to 12-h samples (P < 0.05). Depletion of steroids with TRL increased p27 mRNA levels 12 h after an ovulatory stimulus, an effect that R5020 replacement completely reversed. However, whereas TRL resulted in highly variable levels of p27 mRNA levels at 36 h, the addition of R5020 resulted in a significant reduction in mRNA levels compared to the 36-h control group (P < 0.05).
DISCUSSION
The developing primate corpus luteum increases in size following an ovulatory stimulus due primarily to cell hypertrophy rather than hyperplasia [16, 24, 25, 30]. Although exit of granulosa cells from the cell cycle is a hallmark of luteinization in nonprimate species [3, 31], little is known regarding the control of primate granulosa cell proliferation following an ovulatory stimulus. In the present study, we established that nearly half the granulosa cell population before an ovulatory stimulus is proliferative, whereas less than 10% remain so within 36 h of an hCG bolus. The reduction in proliferation is not steroid-dependent but, rather, is initiated by the ovulatory gonadotropin bolus. The reduction in granulosa cell proliferation following hCG administration is associated with a complex and diverse set of changes in mRNAs encoding various components of cell-cycle machinery, with early (12 h) changes in cyclin B1 and p21 and a later (36 h) increase in p27 mRNA. Notably, specific mRNAs for cell-cycle components are gonadotropin- and steroid-dependent, including cyclin B1 and p27, whereas others (e.g., p21) are gonadotropin-dependent but steroid-independent. Due to extremely limited amounts of tissue, only mRNA levels of the various cyclins and CKI were determined. Future studies will address issues of protein levels and phosphorylation states.
In primates, granulosa cells exit the cell cycle within 12 h of an ovulatory gonadotropin stimulus. Exactly how soon granulosa cells cease dividing following receipt of the ovulatory signal is not known, but it is probable that this represents one of the first steps in the initiation of luteinization. In species with a short periovulatory interval relative to that of primates (e.g., rat), granulosa cells exit the cell cycle within 4 h of the ovulatory stimulus [3]. Regardless, approximately 10% of granulosa cells in primates remain Ki-67-positive 36 h after the hCG bolus. Consistent with these data, 5% to 15% of nonendothelial cells in early corpora lutea of macaques, marmosets, and humans are proliferative but do not express 3ß-HSD [32–34]. These cells may represent, at least in part, a subset of granulosa cells that enter the cell cycle and arrest in G1, perhaps due to increased p21, fail to luteinize in response to hCG and, thus, do not express 3ß-HSD [20]; or are proliferating granulosa-lutein cells that do not synthesize progesterone during the cell cycle [34].
Notably, levels of cyclin D2 mRNA do not decline in granulosa cells following an ovulatory stimulus in rhesus monkeys undergoing controlled ovarian stimulation. Cyclin D/cdk4,6 complexes phosphorylate retinoblastoma protein (pRb), resulting in activation of the E2F transcription factor and passage through G1 and G1-S transition [1, 35]. The expression of cyclin D2 mRNA in primate granulosa cells before the ovulatory stimulus is consistent with the proliferative action of this gene product [8], although the persistent expression after hCG administration is surprising. It is noteworthy that preliminary studies using immunohistochemistry for cyclin D2 parallel the mRNA studies presented herein (data not shown), suggesting that cyclin D2 protein levels do not change following hCG. This, however, does not rule out posttranslational changes to cyclin D2 that could result in reduced activity. It is interesting to note that in rats, cyclin D2 is rapidly down-regulated in granulosa cells following an ovulatory stimulus [6]. This may represent either a species difference, a posttranslational level of control in the primate, or relate to changes in cellular localization. The issue of species difference is especially relevant, because granulosa cells from rats do not express cyclins D1 and D3 whereas mice and primates do [6, 7] (unpublished data). Regardless, data from the current study do not indicate reduced expression of cyclins (D2 and E) controlling G1-S transition following the ovulatory hCG bolus in primates. Rather, cyclin D2 mRNA levels are moderately increased at 12 h post-hCG. Evidence from human fibroblasts suggests that contact inhibition of proliferation is associated with increased expression of cyclin D [36]. The granulosum in primates does not expand until 24 h after hCG [22]; thus, cyclin D2 may function to progress the cell cycle before hCG and to inhibit it through a contact-mediated mechanism during the periovulatory interval. Alternatively, levels of cyclin D2 may be related to the 10% of Ki-67-positive granulosa cells observed following the ovulatory stimulus, thus continuing to function in a proliferative capacity.
Similarly, levels of cyclin E mRNA do not change in granulosa cells during the periovulatory interval in primates. Like cyclin D2, cyclin E facilitates G1-S transition, although in the rat model, cyclin E is not down-regulated until 48 h after hCG [3]. Cyclin E is hypothesized to act in a feedback manner to phosphorylate pRb, thus activating the E2F transcription factor, which in turn increases transcription of cyclin E [35]. Interestingly, mRNA levels of another cyclin E substrate, the phosphatase CDC25A, are also unchanged in macaque granulosa cells following hCG (unpublished data). The lack of change in levels of cyclin D2, cyclin E, and CDC25A collectively suggest that pRb expression/activity is not altered following an ovulatory bolus in primates. This is consistent with a report that pRb is expressed in large antral follicles as well as early corpus luteum in humans [37]. In contrast, phosphorylated pRb is absent from granulosa cells 72 h post-hCG in mice [7], suggesting important species differences. It is possible that loss of granulosa cell proliferation following the ovulatory stimulus in primates is mediated by other G1-arresting mechanisms, such as through a reduction in c-Myc [38]. Alternatively, the observation that mRNA levels do not change could reflect translational regulation, although substantial evidence exists demonstrating transcriptional regulation of cyclins in breast cancer cells [39].
In contrast, cyclin B1 mRNA levels in granulosa cells of macaque periovulatory follicles were significantly reduced following hCG administration. This cyclin facilitates the G2-M transition and, thus, represents an important checkpoint for completion of the cell cycle [1]. Cyclin B plays a role in progestin-induced G1 arrest in T47D-YB breast cancer cells [40] as well in proliferation of uterine endometrial cells [41, 42], supporting the hormonal regulation of this gene. It is noteworthy that the percentage of human luteinizing granulosa cells in S, G2, and M phases is reduced compared to cells obtained before an ovulatory gonadotropin bolus [5]. Thus, we hypothesize that luteinizing granulosa cells are arrested in G1, partly by reduced expression of cyclin B1.
Depletion of intrafollicular steroids does not play a significant role in the early (12 h) decline in granulosa cell proliferation in the macaque periovulatory follicle. Nevertheless, evidence is convincing that steroids (notably progesterone) may influence granulosa cell proliferation through regulation of specific components of the cell cycle. Interestingly, a role for steroids in the control of proliferation has been shown in many endocrine systems, including the ovary. For example, progesterone has antiproliferative actions in cultured human luteinizing granulosa cells [15, 16], although culture of granulosa cells obtained during IVF protocols (i.e., after an ovulatory stimulus) likely results in a phenotype more closely resembling luteal cells rather than periovulatory granulosa cells. In contrast, steroid depletion results in the complete absence of Ki-67-positive granulosa cells in macaque follicles by 36 h post-hCG. We recently observed that steroid depletion increases the percentage of atretic follicles 36 h post-hCG, an effect fully reversible by R5020 [22]. The attenuation of Ki-67 expression may thus be associated with the high percentage (70%) of atretic follicles rather than with a direct regulation of proliferation by steroids. Alternatively, progesterone may permit or promote proliferation of a subpopulation of granulosa cells during luteal development [2].
The mechanism of progesterone control of proliferation appears to be via regulation of components of the cell cycle. Ample evidence exists that progesterone alters the expression of cyclins D and E as well as cyclin E/cdk2 activity in breast cancer cells [17], mouse uterine epithelial cells [43], and human endothelial cells [44]. Consistent with these data, progesterone limits the expression of cyclin E mRNA following hCG in primate granulosa cells. This may represent a balance between gonadotropin-stimulated and progestin-suppressed cyclin E expression and may facilitate G1 arrest. Similarly, cyclin B1 mRNA levels are strongly steroid regulated in granulosa cells following an ovulatory stimulus. The initial (12 h) suppression of cyclin B1 appears to be by androgen or estrogen, because R5020 does not return values to control levels. We have reported previously that androstenedione and estradiol levels in follicular fluid rise between 0 and 12 h after hCG, making these steroids possible early regulators of cyclin B1 [20]. Consistent with this concept, estrogen induces degradation of cyclin B1 mRNA in MCF-7 breast cancer cells [45]. By 36 h post-hCG, steroid control shifts in favor of progesterone suppression of cyclin B1 mRNA expression, although the effects of progesterone on other markers are divergent (i.e., increased Ki-67, decreased p27 [see below]). The effects of progesterone are likely dictated by cell-specific parameters (e.g., degree of luteinization) and may be associated with the 10% Ki-67-positive granulosa cells observed at this time.
The mRNAs for the CKI p21 and p27 increase following the ovulatory stimulus, but their temporal expression profile is different: p21 increases within 12 h of hCG administration and declines thereafter, whereas p27 does not increase until 36 h. A similar expression profile for p21 and p27 exists in rats [6], supporting a general role for these genes in the control of granulosa cell proliferation during the periovulatory interval. In p27-/- mice, luteal cells continue to proliferate, suggesting that the rise in p27 is essential to prevent the cell cycle in luteal cells [9]. However, the expression of p27 increases after the apparent drop in proliferation in rats [3, 6] and primates (current study), whereas the increase in p21 correlates well in both species with the decline in proliferation. Thus, the early rise in p21 appears to be essential for the initial drop in mitosis in granulosa cells. In contrast, p27 inhibits both proliferation and apoptosis following inflammatory injury in renal glomerular cells [46], suggesting that p27 induces cell-cycle arrest during stress conditions, of which follicle rupture could qualify. The late (36 h) increase in p27 may play an important role in the development and function of the primate corpus luteum.
Steroids do not have a clear role in the regulation of CKI expression in granulosa cells from primates undergoing controlled ovarian stimulation. However, p27 mRNA is suppressed by progestins 12 h after hCG administration, and p21 mRNA tends to be reduced following steroid depletion at this time as well. Evidence exists that progesterone regulates the timing of indices relating to luteinization [22], and progestin suppression of p27 at 12 h post-hCG supports this hypothesis. Similarly, unopposed progestin (TRL + R5020) at 36 h post-hCG results in p27 mRNA levels that are not different than those at 0 and 12 h, suggesting that progesterone plus other steroids interact to regulate p27 in granulosa-lutein cells. Progesterone can have biphasic growth-regulatory properties in breast cancer cells, with the initial progestin stimulus accelerating cells through one round of division and subsequent exposure inducing arrest at the G1/S boundary [2]. Similarly, in primate granulosa cells, progestins may regulate the temporal sequence of events leading to cell-cycle arrest, such as by holding p27 and cyclin E mRNA levels in check.
In summary, granulosa cell proliferation is markedly reduced within 12 h of the administration of an ovulatory hCG bolus during controlled ovarian stimulation cycles in rhesus monkeys. Figure 5 illustrates gonadotropin versus steroid-mediated changes in mRNA levels for various components of the cell cycle. Notably, gonadotropin sequentially increases first p21, followed by p27, whereas steroids suppress the expression of cyclin B1. Whether protein levels for these cyclins change during luteinization is not known. However, mRNA levels of the cell-cycle inhibitor p21 increase in a time frame coincident with the loss of proliferation, whereas a second inhibitor, p27, increases just before the time of ovulation. We thus hypothesize that p21 provides important initial control of the cell cycle following hCG, and that p27 plays a role in maintaining G0/G1 arrest of granulosa-lutein cells around the time of rupture. Steroids do not have a clear role in the decline of granulosa cell proliferation, but they do regulate mRNA levels of key components of the cell cycle and, in particular, regulate the timing of p27 expression. Thus, both the gonadotropin surge and steroids, notably progesterone, likely interact to regulate granulosa cell proliferation during luteinization in primates.
|
ACKNOWLEDGMENTS
The authors appreciate the expert service provided by the Division of Animal Resources and the surgical team of Dr. John Fanton and the Endocrine Services Core Laboratory. Recombinant human LH, FSH, CG, and Antide were generously provided by Ares Advanced Technology, Inc., a member of the Ares-Serono Group. Trilostane was graciously supplied by Sanofi Pharmaceutical Inc., Great Valley, Malvern, PA.
FOOTNOTES
First decision: 25 January 2001.
1 Supported by NIH/NICHD HD20869, U54 HD18185 as part of the Specialized Cooperative Centers Program in Reproductive Research, HD8302, and RR00163. 
2 Correspondence: Richard L. Stouffer, Division of Reproductive Sciences,
Oregon Regional Primate Research Center, 505 NW 185th St., Beaverton,
OR 97006. FAX: 503 690 5563; stouffri{at}ohsu.edu
3 Current address: Department of Physiology, Medical College of Georgia,
Augusta, GA 30912.
Accepted: April 12, 2001.
Received: December 12, 2000.
REFERENCES
luteal transition. Cell Growth Differ 1998; 9:787-792[Abstract]This article has been cited by other articles:
|
|
 |
|
  M. B. Stanek, S. M. Borman, T. A. Molskness, J. M. Larson, R. L. Stouffer, and P. E. Patton Insulin and Insulin-Like Growth Factor Stimulation of Vascular Endothelial Growth Factor Production by Luteinized Granulosa Cells: Comparison between Polycystic Ovarian Syndrome (PCOS) and Non-PCOS Women J. Clin. Endocrinol. Metab., July 1, 2007; 92(7): 2726 - 2733. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.N. Fru, M. Cherian-Shaw, M. Puttabyatappa, C.A. VandeVoort, and C.L. Chaffin Regulation of granulosa cell proliferation and EGF-like ligands during the periovulatory interval in monkeys Hum. Reprod., May 1, 2007; 22(5): 1247 - 1252. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M Quirk, R. G Cowan, and R. M Harman The susceptibility of granulosa cells to apoptosis is influenced by oestradiol and the cell cycle. J. Endocrinol., June 1, 2006; 189(3): 441 - 453. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Cannon, M. Cherian-Shaw, and C. L. Chaffin Proliferation of Rat Granulosa Cells during the Periovulatory Interval Endocrinology, January 1, 2005; 146(1): 414 - 422. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Cherian-Shaw, R. Das, C. A. VandeVoort, and C. L. Chaffin Regulation of Steroidogenesis by p53 in Macaque Granulosa Cells and H295R Human Adrenocortical Cells Endocrinology, December 1, 2004; 145(12): 5734 - 5744. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Quirk, R. G. Cowan, and R. M. Harman Progesterone Receptor and the Cell Cycle Modulate Apoptosis in Granulosa Cells Endocrinology, November 1, 2004; 145(11): 5033 - 5043. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Borman, C. L. Chaffin, K. M. Schwinof, R. L. Stouffer, and M. B. Zelinski-Wooten Progesterone Promotes Oocyte Maturation, but Not Ovulation, in Nonhuman Primate Follicles Without a Gonadotropin Surge Biol Reprod, July 1, 2004; 71(1): 366 - 373. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Quirk, R. G. Cowan, R. M. Harman, C.-L. Hu, and D. A. Porter Ovarian follicular growth and atresia: The relationship between cell proliferation and survival J Anim Sci, January 1, 2004; 82(13_suppl): E40 - 52. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. L. Chaffin, R. S. Brogan, R. L. Stouffer, and C. A. VandeVoort Dynamics of Myc/Max/Mad Expression during Luteinization of Primate Granulosa Cells in Vitro: Association with Periovulatory Proliferation Endocrinology, April 1, 2003; 144(4): 1249 - 1256. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Gillio-Meina, Y. Y. Hui, and H. A. LaVoie GATA-4 and GATA-6 Transcription Factors: Expression, Immunohistochemical Localization, and Possible Function in the Porcine Ovary Biol Reprod, February 1, 2003; 68(2): 412 - 422. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
