| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Centre for Reproductive Health, Department of Obstetrics and Gynaecology, University of Adelaide, Woodville, South Australia 5011, Australia
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
androgen receptor, follicle, granulosa cells, growth factors, ovary
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Oocyte-secreted factors (OSFs) play a role in determining phenotypic differences between the two subpopulations of granulosa cells that emerge when a follicle enters the antral stage of folliculogenesis. These two subpopulations are morphologically distinct and are partly defined by their proximity to the oocyte or to the basement membrane: cumulus cells (CCs) surround and maintain direct contact via gap junctions with the oocyte, forming the cumulus-oocyte complex (COC), while mural granulosa cells (MGCs) form the follicle wall, separated from surrounding theca cells by the basement membrane. Within this model there exist at least two chemical gradients to which granulosa cell subtypes are differentially exposed: one arising from the oocyte, and one arising from the theca cells and surrounding vasculature. Soluble factors from the oocyte have been shown to impart temporal differences in granulosa cell responses to environmental factors such as FSH and insulin-like growth factor 1 (IGF1) [5–7], which are both essential for normal folliculogenesis. These studies show that CCs proliferate primarily in response to the presence of these factors, whereas MGCs have a more steroidogenic phenotype. Moreover, CCs will become more steroidogenic in the absence of an oocyte [7, 8] and MGCs will become more proliferative when cocultured with denuded oocytes [9, 10]. It appears that GDF9 and its homologue GDF9b can account for some but not all of these particular OSF-induced effects [11, 12], and an interaction between the two may be necessary for normal follicular development in vivo through as-yet-unresolved mechanisms [13] that are likely to differ between animal species with different ovulation rate phenotypes [3].
Locally produced ovarian steroid hormones also regulate folliculogenesis through modulation of gonadotropin and growth factor activities. During antral follicle growth, theca cells predominantly secrete aromatizable androgens, which accumulate in micromolar quantities within follicular fluid. In the early growth phase, granulosa cells preferentially metabolize these androgens to 5
-dihydrotestosterone (DHT) [14–16], a nonaromatizable androgen that has the highest affinity for the androgen receptor (AR). Granulosa cells of all mammalian species examined to date express the AR and direct, AR-mediated activity has been implicated in the control of folliculogenesis through mechanisms that involve both FSH [17] and IGF1 [18, 19]. This steroid receptor has also been identified in the oocytes of some mammalian species [20–22] and may be involved in nonclassical steroid receptor signaling pathways that influence oocyte maturation [23]. In a previous study, we observed differential effects of DHT on mitogenic and steroidogenic responses of pig cumulus-oocyte complexes (COCs) and MGCs to FSH and IGF1 that were dependent on follicle size and variably associated with classic steroid receptor-mediated mechanisms [24]. The latter study led to our current hypothesis that there is an interaction between OSF- and androgen-stimulated signaling in granulosa cells. To address this hypothesis, we have employed an established bioassay for OSFs in which denuded oocytes (DO) are cocultured with MGCs to ascertain whether factors derived from the oocyte can induce responses in MGCs that are characteristic of the cumulus cell phenotype. This technique has been employed primarily using mouse models [9, 10] and, to a lesser extent, domestic animal models [6, 7], both under various culture conditions that may or may not involve FSH, IGF1, or both. We also examined whether recombinant GDF9, a known OSF, had the same effect on porcine granulosa cells as that elicited by denuded oocytes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Collection of Follicular Cells
Prepubertal pig ovaries were collected from a local abattoir and follicular cells comprised of separate COC and aggregated MGC cell fractions were aspirated from small antral follicles (1–3 mm diameter) and isolated as previously described [24]. For the current study, COCs in warm 25 mM Hepes-buffered tissue culture medium-199 (H-TCM; ICN, Costa Mesa, CA) supplemented with sodium pyruvate (2 mM), penicillin G (100 U/ ml), streptomycin sulfate (100 mg/ml), and polyvinyl alcohol (PVA; 0.3 mg/ml) were further processed to remove the surrounding cumulus cells to obtain DO. This was achieved by collecting pooled COCs in 1 ml of H-TCM with supplements and vortexing them for 4 min in a 15-ml centrifuge tube (Falcon, Franklin Lakes, NJ). The oocytes were allowed to pellet by unit gravity for 1 min before being transferred to a new Petri dish containing warm H-TCM. Oocytes that were free of cumulus cells and had an intact zona pellucida were further isolated and washed twice in warm bicarbonate-buffered TCM-199 (B-TCM supplemented as per H-TCM; ICN). At the end of their isolation procedure, a small aliquot of the MGC suspension was manually dispersed by repeated pipetting to disassociate aggregated granulosa cells to enable accurate determination of cell numbers by counting with a hemocytometer.
Production and Partial Purification of GDF9
Because GDF9 is not commercially available, bioactive recombinant mouse GDF9 was produced in house from a transfected human embryonic kidney-293H cell line, generously donated by Olli Ritvos (University of Helsinki) [11, 25]. In preliminary experiments, materials secreted by the parent cell line were found to be inhibitory to pig, but not to mouse [11] MGC proliferation. Therefore, conditioned media were subjected to purification by hydrophobic interaction chromatography (HIC).
Control conditioned media from untransfected 293H cells, and conditioned media from 293H cells expressing recombinant mouse GDF9 (maximum 200 ml for a 1-ml column), were concentrated approximately 20-fold by ultrafiltration using a YM10 (10 000 MWCO) membrane (Millipore Corporation, Bedford, MA). After addition of ammonium sulfate to a final concentration of 1 M, the medium was loaded onto a 1-ml Phenyl Sepharose (low sub) column (Amersham Biosciences, Sydney, Australia), pre-equilibrated with 50 mM sodium phosphate pH 7.0 and 1 M ammonium sulfate. The column was washed with 13 ml of the same buffer and then eluted with a linear gradient of 50 mM sodium phosphate pH 7.0. The procedure was carried out at room temperature at a flow rate of 1 ml/ min. One-milliliter fractions containing GDF9 were detected by immunoblotting using monoclonal antibody GDF9-53, and by using a mouse MGC proliferation bioassay, both as previously described [11]. Fractions that were both immunoreactive and bioactive were pooled, concentrated, and dialyzed with PBS using a Centriprep concentrator (10 000 MWCO; Millipore). After filter sterilization, the samples were stored at –80°C. Concentration of GDF9 was determined by immunoanalysis using a rat GDF9 preparation as the standard, as previously described [11].
The relative bioactivities of conditioned media from 293H cells expressing GDF9, and media from untransformed 293H cells, before and after HIC purification, were determined by addition at increasing concentrations to cultures of MGCs alone or in the presence of 50 ng/ml recombinant human IGF1 (rhIGF1; Gropep, Adelaide, Australia). Cells were cultured and [3H]-thymidine incorporation was determined as described below.
Follicle Cell Culture
MGCs were cultured as aggregates in 125 µl of media (serum-free B-TCM, supplemented as above) in 96-well flat-bottom plates (Falcon) at a density of 0.5 x 106 cells/ml. Variable combinations of the following treatments were added, depending upon the particular experiment: 50 mIU/ ml recombinant human FSH (rhFSH; Puregon, N.V. Organon, Oss, The Netherlands), 50 ng/ml recombinant human IGF1 (rhIGF1), 5–500 nM DHT (Sigma), 5–1000 nM testosterone (Sigma), DO (5–40/well), and recombinant mouse GDF9 (15–120 ng/ml). To block action of the AR, some experiments included the addition of a nonsteroidal antiandrogen, hydroxyflutamide (OHF, Sigma). In these latter experiments, cells were preincubated for 1 h with 100 or 1000 nM OHF, plus other nonandrogen treatments, before the addition of DHT or testosterone. Cells were cultured with treatments in an atmosphere of 38.5°C, 96% humidity, 5% CO2 in air for 18 h followed by a 6-h pulse of 0.8 µCi tritiated thymidine ([3H]-thymidine; ICN Biomedicals, NSW, Australia) in the same conditions. At 24 h, a fraction of the culture medium was removed and frozen (–20°C) for steroid analysis and plates were kept at 4°C until cell harvest. Experiments were performed three to six times, in which each treatment was represented in triplicate, except in a few instances in the OHF experiments, in which insufficient numbers of DO necessitated some treatments in duplicate.
Measurement of DNA Synthesis and Progesterone Secretion
Incorporation of [3H]-thymidine was measured in cells as an indication of the degree of cellular DNA synthesis and potential proliferation using a previously described method and validation procedure [24]. Conditioned medium samples were assayed for progesterone content using a radioimmunoassay kit (Diagnostic System Laboratories, Webster, TX) in accordance with the manufacturer's instructions. The kit has a sensitivity of 0.25 pmol/ml, an intraassay coefficient of 8.4%, and an interassay coefficient of 12%.
Data Analysis
Statistical analyses for data represented in Figure 2, were performed with SAS software (SAS Institute, Cary, NC) using four-way analyses of variance (ANOVAs) on mean data, derived from six replicate experiments in which each individual mean represented the average of triplicate data points for each treatment group, with blocking on experiment. Where significant treatment effects were detected, post hoc t-tests were used for comparison of adjusted means for those significant effects. All other experimental data were analyzed with SigmaStat software (Version 2, SPSS Inc., Chicago, IL) using 1- or 2-way ANOVAs, as appropriate, on log-transformed raw data, derived from three replicate experiments. Differences between groups were subsequently assessed by all pair-wise multiple comparison procedures (Tukey test). Statistical significance in all instances was set at P < 0.05.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
When GDF9-expressing 293H conditioned media were subjected to HIC, GDF9 bound to the Phenyl Sepharose (low sub) column. Both the precursor protein (57 kDa as monomer) and the more abundant mature form (17.7 kDa as monomer) were eluted with between 640 mM and 180 mM ammonium sulfate (Fig. 1A). All fractions were examined for immunoreactive and bioactive GDF9, and only those fractions that contained immunoreactive GDF9 (fractions 35–40; fractions 1–29 not shown) (Fig. 1B) also contained bioactive GDF9, as demonstrated by potent stimulation of [3H]-thymidine incorporation in cultured mouse MGCs (data not shown). After dialysis of the pooled immunoreactive fractions, GDF9 was shown to be enriched approximately 5-fold (from 2 to 6 µg GDF9/mg total protein to 10–28 µg GDF9/mg total protein). HIC successfully removed the inhibitory factors expressed by the parent 293H cell line present in the conditioned media. While control (untransfected) 293H-conditioned medium before HIC dose-dependently suppressed [3H]-thymidine incorporation in pig MGCs cultured with IGF1, control medium after HIC had no such inhibitory effect (Fig. 1C). Furthermore, GDF9 post-HIC medium was twice as active as pre-HIC medium in enhancing IGF1-stimulated DNA synthesis. To determine whether specific inhibitory substances remained in the post-HIC preparation that affected FSH- but not IGF1-regulated MGC functions, we performed experiments with post-HIC 293H conditioned media, at equivalent concentration to the highest dose of GDF9 used in the main experiments, and found no inhibition of FSH-stimulated DNA synthesis or progesterone secretion (data not shown). Hence, partial purification of GDF9 by HIC was effective and necessary to generate GDF9 that is bioactive on pig MGCs; this HIC purification was also necessary for GDF9 to promote cumulus expansion in the mouse [26].
|
Effects of DO or GDF9 with or Without DHT on MGC Responses
Experimental results are summarized in graphs (Fig. 2) in which all data are expressed as the average of raw means from six replicate experiments ± SEM. Due to the complexity of the data set, statistical significance is indicated only for responses related to the effects of DO or GDF9 alone and in combination with DHT, as these represent the more novel responses that are the main focus of this study. In most instances, the actions of FSH, IGF1, and DHT, alone and in combination, were as previously described [24]. In summary, FSH and IGF1 both stimulated DNA synthesis in MGCs and had an additive effect when present together; DHT had little additional effect, with the exception of a small, <2-fold increase in IGF1-stimulated DNA synthesis. The effects of FSH, IGF1, and DHT, alone and in combination, on progesterone secretion from MGCs were also consistent with previous observations [24], whereby FSH more potently stimulated progesterone secretion as compared to IGF1, and combined hormone had a synergistic effect. However, in the current study, DHT significantly inhibited (P < 0.05) FSH-stimulated steroid secretion, an effect that was not consistently observed in the previous study.
Before the current study, we observed that 500 nM DHT potently augmented IGF1-stimulated DNA synthesis and suppressed FSH-stimulated progesterone secretion in porcine COCs from 1- to 3-mm antral follicles, but the same treatment of MGCs from the same follicles had little, or no effect, respectively [24]. As a continuation of this line of investigation, we herein repeated those same experimental conditions, substituting a coculture of MGCs with DO for the COC to determine whether similar responses could be elicited. When HIC-purified GDF9 became available, we then repeated the same experiments with this known OSF factor to ascertain whether it behaved similarly to the DO in the previous experiment. The doses of DO (20/well) and GDF9 (60 ng/ml) used in these exploratory experiments were chosen on the basis of similar experiments performed in mice that did not include the DHT treatment [4].
Effects of Denuded Oocytes
[3H]-thymidine uptake In the absence of any other growth factors, DO induced a moderate (2-fold) stimulation of DNA synthesis in MGCs; however, they potently stimulated [3H]-thymidine incorporation in the presence of IGF1 (P < 0.0001) (Fig. 2A), independent of the presence of FSH; this occurred in a dose-dependent manner (Fig. 3). The overall stimulatory effect of DO was significantly enhanced by DHT (P = 0.01), irrespective of the presence of IGF1 or FSH (Fig. 2A). Interestingly, DO tended to further increase DNA synthesis in the presence of combined FSH and IGF1, as compared to either mitogen alone, in marked contrast to previous observations with intact COCs, in which suppression of DNA synthesis is a hallmark response to this combined treatment [5, 24].
|
Progesterone secretion Denuded oocytes also suppressed progesterone secretion from MGCs, an effect that was most significant in the presence of FSH (P < 0.001), independent of the presence of IGF1 (Fig. 2C). In addition, DO significantly suppressed IGF1-stimulated progesterone secretion (P = 0.04) in a manner independent of FSH. The inhibitory effects of DHT and DO on progesterone secretion appear to be independent because there was not a significant interaction between these two factors, although in combination, they induced the most potent suppressive cocktail in the presence of FSH, IGF1, or both as compared to either factor alone.
In summary, these DO-induced effects on hormone stimulated proliferation and progesterone secretion are indicative of a partial reversal of the differentiated MGC phenotype toward a less differentiated, more proliferative phenotype. Furthermore, this reversal was enhanced in all instances by the additional presence of DHT, particularly and most significantly in terms of proliferation.
Effects of GDF9
[3H]-thymidine uptake Like DO, GDF9 alone moderately stimulated DNA synthesis in MGCs, and significantly interacted with IGF1 (P < 0.0001) (Fig. 2B), enhancing [3H]-thymidine incorporation in a dose-dependent manner (Fig. 3). However, unlike DO, the recombinant agent did not affect FSH-stimulated mitogenesis, and did not further enhance the mitogenic effects of combined IGF1 and FSH (Fig 2B). Furthermore, GDF9 did not have statistically significant interactions with DHT when assessed across all treatment options, although the combination of GDF9 and DHT more potently enhanced IGF1-stimulated DNA synthesis compared to either agent alone (P = 0.02), and there was a tendency for this combination to inhibit IGF1 + FSH-stimulated DNA synthesis when either agent alone had no effect.
Progesterone secretion In a manner similar to DO, GDF9 significantly suppressed FSH-stimulated progesterone secretion (P < 0.0001) independent of IGF1 (Fig. 2D). Unlike DO, GDF9 increased basal progesterone secretion 2-fold, and did not have any significant independent interactions with IGF1 in the modulation of progesterone secretion. Furthermore, significant interactions were not observed between GDF9 and DHT; however, as with DO, there was an obvious pattern of enhanced suppression of FSH-stimulated progesterone secretion when both of these factors were present as compared to either factor alone, and this pattern was more pronounced in the FSH + IGF1 treatment group.
In summary, recombinant GDF9 mimicked many of the actions induced by DO in porcine MGCs, particularly in terms of enhancing IGF1-stimulated DNA synthesis and suppressing FSH-stimulated progesterone secretion. Interactions between the recombinant protein and DHT also showed marked similarities to those observed with DO, but confined to instances where GDF9 behaved in a DO-like manner.
Dose Effects of DHT and Testosterone on DO or GDF9-Stimulated MGCs
To ascertain whether both of the primary AR agonists, DHT and testosterone, show evidence of an interaction with mitogenic signaling initiated by DO, GDF9, or both, a range of doses for each androgen was employed under the following constant conditions that were replicated from the initial experiments: IGF1 (50 ng/ml) ± DO (20/well) or GDF9 (60 ng/ml), as depicted in Figure 4. In the absence of DO or GDF9, only the highest dose (500 nM) of DHT (Fig. 4A) and the two higher doses of testosterone (100 nM and 1000 nM) (Fig. 4B), increased IGF1-stimulated DNA synthesis, (less than 2-fold in all instances). However, in the presence of DO+IGF1 or GDF9+IGF1, low doses of both androgens that were ineffective with IGF1 alone, led to substantial increases in DNA synthesis. The dose-dependent nature of the three-way interactions between androgen x IGF1 x GDF9/DO were similar with DHT or testosterone.
|
Opposition of Androgen-Induced Effects with the Antiandrogen OHF
To determine whether the observed androgen effects on mitogenic signaling in granulosa cells were mediated through the AR, cells were pretreated for 1 h with the antiandrogen OHF (at double the androgen dose) before the addition of particular doses of DHT or testosterone. These experiments were performed in the presence of IGF1 and DO (20/well) or GDF9 (60 ng/ml), but only the results of the GDF9 experiments are presented, because OHF alone at either dose had highly variable inhibitory effects on the activity of DO. In the presence of IGF1+GDF9, the stimulatory action of 50 nM DHT was completely inhibited by 100 nM OHF, but 500 nM DHT was only partially inhibited by 1000 nM OHF, possibly due to a small agonistic effect of the higher dose of OHF alone (Fig. 5A). The same doses of testosterone had potent stimulatory effects that were substantially, but not fully, inhibited by the antiandrogen (Fig. 5B).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Oocyte control of folliculogenesis via OSFs now represents a firmly established concept in ovarian biology, but the mechanics of this control still require characterization, especially in nonmurine species. As previously documented in mice and cows (reviews in [2, 4]), as well as pigs [30, 31], coculture of MGCs with DO in the present study induced changes in GC responses that correlate with a less-luteinized phenotype, as characterized by increased proliferation and diminished progesterone secretion. The mitogenicity of pig DO was largely dependent on the simultaneous presence of IGF1, having little or no stimulatory effect in the absence of any other growth factors or in the presence of FSH alone. These observations concur with those reported by Li et al. under similar culture conditions in the cow [7] and Brankin et al. [31] under long-term culture in the pig, but differ from those observed in the mouse, in which oocytes alone exert a potent mitogenic effect that is similar to, but does not augment IGF1-stimulated MGC proliferation [10]. In contrast to the species-specific differences in mitogenicity of oocyte secretions, the inhibitory effect of oocytes on progesterone secretion from MGCs in vitro occurs mainly in the presence of FSH in all species studied (reviewed in [4]).
Overall, the nonaromatizable androgen DHT universally potentiated the interactions of an unknown cocktail of OSFs secreted by DO with IGF1, FSH, or both: in essence, mitogenesis was greatest and progesterone secretion was most attenuated in the combined presence of DHT and DO. While it is clear that DHT and oocytes can independently influence the activities of FSH and IGF1, our data suggest that there is an interaction between OSFs and androgen that is independent of their individual effects on signaling by these growth factors: one that nonetheless appears to bootstrap those independent effects. Therefore it is feasible to propose that OSFs either potentiate androgen signaling, or androgens potentiate oocyte signaling, or both. Because most identified OSFs are members of the TGFß superfamily, there is precedence for such interaction from studies in the prostate whereby androgen regulates expression of TGFß1 and its cognate receptors [32–34], and the AR interacts with intracellular SMAD signaling molecules [35, 36]. Additionally, Killian et al. [37] report that an androgen-regulated protease cleaves latent TGFß1 in osteoblast cells and proteolytically modulates cell surface receptors. In the ovary, DHT administered in vivo to hamsters induces TGFß receptor type II mRNA in follicular cells [38, 39]. Although TGFß1 is expressed by the oocyte, it is unlikely to be the factor responsible for the effects of oocytes and the interaction of its secreted factors with androgen in the current study, as predicated by previous studies in the cow [40] and mouse [11], and supported by our unpublished observations using the current pig model.
In the mouse model, many of the effects of OSFs on GCs can be mimicked by, but not necessarily attributed to, GDF9 [11, 26, 41]. GDF9 is a member of the TGFß superfamily, whose expression is oocyte-specific in many species, but possibly not in primates [42, 43] or pigs [44], in which expression has also been detected to a lesser degree in granulosa cells. Due to lack of commercial availability, the specific effects of GDF9 have not been widely explored outside of the mouse model. Using recombinant mouse GDF9 generated in-house, we demonstrate that GDF9 generally mimicked the nonluteinizing effects of oocytes in our culture system, especially in terms of dose-dependent increases in IGF1-stimulated proliferation. This specific mitogenic effect of GDF9 or DO was significantly augmented by DHT and testosterone, the two most potent AR agonists. Moreover, the effects of both androgens on GDF9-stimulated cells could be inhibited by the antiandrogen OHF, implicating an AR-mediated mechanism. The variable influence of OHF in the DO coculture model (data not shown) probably indicates a direct effect of OHF, androgens, or both on naked oocytes, as this variability was not observed when COCs were similarly treated [24]. Although AR expression has been documented in pig oocytes [21], the dynamics of this expression has not been fully explored as in the rat, in which expression is differentially localized at various developmental stages [22]. A transcription-independent, AR-mediated mechanism has been implicated in mouse oocyte maturation, under conditions in which oocytes were stripped of surrounding GCs [45], giving support to the possibility that androgens in our coculture system caused direct effects within the oocyte. It will be interesting to determine whether such effects influence the expression and/or secretion of GDF9 and/or other OSFs.
One of the striking differences between the actions of DO and GDF9 in the current study was their opposing effect on proliferation in the presence of FSH and IGF1. Inhibition of IGF1-stimulated DNA synthesis by FSH is a characteristic feature of COCs that distinguishes them from MGCs [5, 46], which can be replicated by coculturing oocytectomized cumulus complexes with denuded oocytes in the cow [7]. Curiously, in the current study, GDF9 tended to induce this cumulus cell-like effect in MGCs, but DO did not; in fact, DO further enhanced DNA synthesis under these conditions. In addition, DHT potentiated both of these opposing effects. There are a few possible explanations for these differences: 1) the additional complexity of other factors secreted by the oocyte, GCs, or both in response to those factors; 2) the exact amount of GDF9 released by the DO during culture could not be determined, and was potentially in constant renewal over the culture period, in contrast to a fixed dose of recombinant GDF9; and 3) mouse oocytes seem to predominantly secrete GDF9 in the unprocessed form, while the recombinant molecule represents mainly the processed form [11]. It is provocative to speculate that the specific role of GDF9, its interaction with DHT, or both, changes as antral follicles become more vascularized and FSH intervenes on activities largely driven by IGF1 alone in earlier stages of folliculogenesis. These complex interactions, and the additional effects of other OSFs such as GDF9b, are to be explored in future studies.
Consequences of interruption to the interaction between androgens and OSFs may be inferred from studies of mice with perturbations of androgen signaling. Ablation of 5--reductase [47] or inactivating mutations of the AR [48], induce a subfertility phenotype in mice that has not yet been explored at the ovarian level. However, a similar subfertility is also evident in AR-knockout mice, characterized by a reduction of GC numbers in large antral follicles that may result in luteal deficiency [49], suggesting an important albeit nonobligatory role for androgens in GC proliferation. Dominant follicles are characterized by exponential increases in GC numbers, and a critical mass of GCs may be necessary for ovulation to occur [50]. Perhaps androgens provide the extra boost necessary to achieve this optimal growth. Indeed, pigs treated with DHT have an increased rate of ovulation [51]. Nonhuman primates treated with excess androgens have increased follicular growth [52], and elevated androgen in women is associated with an abnormal accumulation of antral follicle numbers (polycystic ovaries) (reviewed in [53]), supporting a growth-promoting role for androgens in an in vivo context in which cells are exposed to oocyte-secreted factors. However, in polycystic ovary syndrome (PCOS), ovulation is impaired due to a follicular arrest by unknown mechanisms. Such follicles are characterized by a deficient GC layer [54], and perhaps reduced expression of GDF9 in oocytes [55], giving rise to speculation that elevated androgens disturb normal GDF9 signaling [56]. Herein we provide evidence of a novel interaction between androgen and GDF9 that substantiates the feasibility of this postulate, although the mechanistic profile of this newly discovered interaction, and its species-specific peculiarities, have yet to be determined.
In conclusion, we have demonstrated that androgens directly interact with GDF9 and possibly other unknown factors secreted by the oocyte to amplify GC proliferation at early antral stages of folliculogenesis in the pig. Oocyte signaling may therefore be necessary for androgens to stimulate a proliferative rather than a differentiative pathway in follicular granulosa cells.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Theresa E. Hickey, Research Centre for Reproductive Health, Department of Obstetrics and Gynaecology, The University of Adelaide, The Queen Elizabeth Hospital, 1st Floor Maternity Building, Woodville Road, Woodville, South Australia 5011, Australia. FAX: 61 8 8222 7521; theresa.hickey{at}adelaide.edu.au
Received: 20 December 2004.
First decision: 15 January 2005.
Accepted: 3 June 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. B. Gilchrist, M. Lane, and J. G. Thompson Oocyte-secreted factors: regulators of cumulus cell function and oocyte quality Hum. Reprod. Update, March 1, 2008; 14(2): 159 - 177. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.A. Walters, C.M. Allan, and D.J. Handelsman Androgen Actions and the Ovary Biol Reprod, March 1, 2008; 78(3): 380 - 389. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. I. Durnerin, K. Erb, R. Fleming, H. Hillier, S.G. Hillier, C.M. Howles, J.-N. Hugues, A. Lass, H. Lyall, P. Rasmussen, et al. Effects of recombinant LH treatment on folliculogenesis and responsiveness to FSH stimulation Hum. Reprod., February 1, 2008; 23(2): 421 - 426. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. X. Yeo, R. B. Gilchrist, J. G. Thompson, and M. Lane Exogenous growth differentiation factor 9 in oocyte maturation media enhances subsequent embryo development and fetal viability in mice Hum. Reprod., January 1, 2008; 23(1): 67 - 73. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Benoit, H. A. LaVoie, G. L. McCoy, and C. A. Blake Expression of Sperm Protein 22 (SP22) in the Rat Ovary During Different Reproductive States Experimental Biology and Medicine, July 1, 2007; 232(7): 910 - 920. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P M. Lokman, K. A N George, S. L Divers, M. Algie, and G. Young 11-Ketotestosterone and IGF-I increase the size of previtellogenic oocytes from shortfinned eel, Anguilla australis, in vitro Reproduction, May 1, 2007; 133(5): 955 - 967. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. A. Dragovic, L. J. Ritter, S. J. Schulz, F. Amato, J. G. Thompson, D. T. Armstrong, and R. B. Gilchrist Oocyte-Secreted Factor Activation of SMAD 2/3 Signaling Enables Initiation of Mouse Cumulus Cell Expansion Biol Reprod, May 1, 2007; 76(5): 848 - 857. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. J. Diaz, K. Wigglesworth, and J. J. Eppig Oocytes determine cumulus cell lineage in mouse ovarian follicles J. Cell Sci., April 15, 2007; 120(8): 1330 - 1340. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. B. Gilchrist, L. J. Ritter, S. Myllymaa, N. Kaivo-Oja, R. A. Dragovic, T. E. Hickey, O. Ritvos, and D. G. Mottershead Molecular basis of oocyte-paracrine signalling that promotes granulosa cell proliferation J. Cell Sci., September 15, 2006; 119(18): 3811 - 3821. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. S. Hussein, D. A. Froiland, F. Amato, J. G. Thompson, and R. B. Gilchrist Oocytes prevent cumulus cell apoptosis by maintaining a morphogenic paracrine gradient of bone morphogenetic proteins J. Cell Sci., November 15, 2005; 118(22): 5257 - 5268. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||
