| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Applied Dental Medicine, Southern Illinois University, School of Dental Medicine, Alton, Illinois 62002
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
cyclic adenosine monophosphate, cytokines, estradiol, granulosa cells, signal transducers
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The pituitary glycoprotein, FSH, promotes the differentiation of granulosa cells into E2-secreting cells, and the induction of cAMP signal transduction by FSH is a central intracellular cascade that controls E2 production [2]. Key events that stimulate the production of cAMP include the induction of stimulatory guanine-nucleotide binding proteins (Gs proteins), and the Gs protein-directed stimulation of adenylyl cyclase. Adenylyl cyclase then catalyzes the generation of cAMP, leading to the subsequent activation of cAMP-dependent protein kinases (PKA). In addition, FSH-stimulated cAMP-dependent phosphodiesterases (i.e., PDE4D) appear necessary in order to down-regulate cAMP levels, and hence maintain cAMP concentrations that are conducive to promoting steroidogenesis in granulosa cells [3, 4].
Importantly, FSH-dependent cAMP signaling, and the resultant production of E2, can be up- and down-modulated by intraovarian cytokines and growth factors that appear to work via a plethora of signal transduction mechanisms, to include the induction of cGMP [5–10]. With this in mind, it can be proposed that cohorts of positive- and negative-modulatory cytokines and growth factors aid in orchestrating follicle development by either promoting (positive-modulatory) or impairing (negative-modulatory) FSH bioactivity within granulosa cells. In so doing, intraovarian cytokines and growth factors may ensure that the onset and magnitude of FSH-stimulated E2 secretion by granulosa cells falls within the endocrine parameters that are essential for reproductive viability.
When considering the negative modulation of FSH-dependent E2 production in granulosa cells, the intraovarian cytokine, hepatocyte growth factor (HGF) [11, 12], attenuates FSH-stimulated E2 production in granulosa cells [9, 13]. The mechanism of action for HGF in granulosa cells is not known, however, HGF is the ligand for c-Met, a transmembrane receptor tyrosine kinase that has been localized in granulosa cells [9].
In contrast to the down-modulatory effects of HGF, insulin-like growth factor-I (IGF-I) is a positive-modulator that synergistically augments FSH-directed E2 production in granulosa cells [14]. Because stimulatory and inhibitory growth factors and cytokines coordinate the FSH responsiveness of granulosa cells, it can be suggested that cross talk between FSH-, HGF-, and IGF-I-stimulated signaling cascades coordinates any interactions between these hormones.
Although previous reports have elaborated the stimulatory actions of IGF-I during the course of granulosa cell differentiation in vitro [15], little is known about how HGF impairs E2 synthesis in granulosa cells. A previous report has shown that in cultured rat granulosa cells, HGF can selectively suppress FSH-dependent E2 synthesis, but not that stimulated by the cAMP analogue, N6, 2'-O-butyryl-cAMP (Bu2-cAMP) [9]. This suggested that HGF compromises the endogenous concentration of cAMP, which supports E2 production in granulosa cells. One potential mechanism of action for this could include an HGF-dependent disruption in the balance of FSH-stimulated cAMP that is maintained by adenylyl cyclase, by members of the cyclic nucleotide-PDE superfamily, or both.
In order to better understand how HGF suppresses E2 synthesis, this study investigated the effects of HGF on FSH-dependent and IGF-I-dependent steroidogenesis in rat granulosa cells in vitro. HGF-directed alterations in cyclic nucleotide-dependent signal transduction in granulosa cell cultures were also evaluated.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Recombinant human HGF (lyophilized with BSA as a carrier, >97% purity) was purchased from R&D Systems (Minneapolis, MN). Recombinant human FSH (AFP8468A) and human IGF-I were generously donated by the National Hormone and Pituitary Program of the National Institute of Diabetes and Digestive and Kidney Diseases, and by Dr. Parlow (Harbor-UCLA Medical Center, Torrance, CA). McCoy 5A medium (M5A, serum-free), Medium 199, fetal bovine serum, and cell culture antibiotics were purchased from Invitrogen (Grand Island, NY). The radioimmunoassay kits were obtained from the following vendors: E2, from Diagnostic Products Corporation (Los Angeles, CA); progesterone (P4) and estrone (E1), from Diagnostic Systems Laboratories, Inc. (Webster, TX); and cAMP and cGMP, from Biomedical Technology (Stoughton, MA). Tritiated cyclic nucleotides were purchased from Amersham-Pharmacia (Piscataway, NJ).
Granulosa Cell Culture
Animal procedures were approved by the Southern Illinois University-Edwardsville Institutional Animal Care and Use Committee. Granulosa cells were harvested from the ovaries of nonhormonally primed, immature (25–27 days old), intact Sprague-Dawley rats as previously described [9]. Cells were either incubated in serum-coated [16] 24-well plates in order to measure steroid (1 x 105 viable granulosa cells/well) and cyclic nucleotide production (1 x 106 viable granulosa cells/well), or in serum-coated 6-well plates (1 x 106 viable granulosa cells/well) in order to measure cyclic nucleotide PDE activity.
In all experiments, granulosa cells were allowed an overnight in vitro acclimation period before the addition of treatments. All incubations were conducted at 37°C in a humidified atmosphere containing 5% CO2 in air. After the acclimation period, cell-conditioned media were removed, and granulosa cells were replenished with fresh (37°C) M5A containing 0.1 µM androstenedione. Granulosa cells identified as controls in this report received androstenedione without additional treatments. In order to promote steroidogenic differentiation, granulosa cells were given FSH (3 ng/ml), and the effects of HGF (3–60 ng/ml), IGF-I (30 ng/ml), or both were tested as specified below. Separate cultures were preincubated for 30 min with 3-isobutyl-1-methylxanthine (IBMX, 10 µM) and then coincubated in the presence and absence of FSH, with and without HGF.
Incubations for the measurement of steroid secretion were terminated at 48 h following the addition of hormones. Radioimmunoassays were conducted according to the protocols provided by the kit manufacturers.
Determination of Cyclic Nucleotide Content
To measure cAMP levels, granulosa cells were incubated in the presence and absence of IBMX (10 µM) for 30 min; cultures in which cGMP was measured were not challenged with IBMX. Following the pretreatment, granulosa cells were challenged with FSH or FSH and HGF for 6, 24, and 36 h. Cell-conditioned media were collected, heated at 90°C for 10 min, and then frozen (-20°C).
Intracellular cyclic nucleotides were measured in the plated granulosa cells following an incubation with 5% trichloroacetic acid and ether extractions as described by the kit manufacturer. Cyclic nucleotide radioimmunoassays were conducted using acetylation according to the kit manufacturer's protocol. Data are presented as the sum of extracellular and intracellular cyclic nucleotide content for each treatment group.
Phosphodiesterase Activity Assay
The activities of cAMP-PDE and cGMP-PDE were measured in separate experiments essentially as described previously [17]. Briefly, granulosa cells were incubated in the presence and absence of FSH (3 ng/ml), HGF (30 ng/ml), IGF-I (30 ng/ml), or concomitant treatment with FSH, HGF, and IGF-I for 24 and 36 h. At these times, the culture plates were placed on ice, and conditioned media were aspirated and discarded. The adherent granulosa cells were washed once with cold (4°C) PBS pH 7.2. Next, 100 µl of PDE lysis buffer (25 mM Tris-HCl pH 8.0, 1 mM MgCl2, 2 mM PMSF, 0.1 mM EGTA, 10 mM ß-mercaptoethanol, and 1 mM imidazole) was added to each well, and the granulosa cells were scraped from the well bottoms using cell lifters. The granulosa cells (in lysis buffer) were gently shaken (4°C) for 10 min, and then centrifuged at 14 000 x g (4°C) for 15 min. Supernatants were collected, and protein concentrations were measured using Bio-Rad (Hercules, CA) protein dye reagent. Cyclic nucleotide-PDE activity was determined by adding equal volumes of PDE reaction cocktail (100 mM Tris-HCl pH 8.0, 10 mM MgCl2, 2 µM of either cAMP or cGMP, and 1 x 105 cpm of either [3H]cAMP or [3H]cGMP) to the cell lysates. Blank (background) control reactions were conducted in tubes containing lysis buffer and PDE reaction cocktail, without cell lysates. Additional controls were conducted using granulosa cell lysate that had been incubated in a boiling H2O bath for 5 min.
Reactions were incubated at 30°C for 10 min, and then heat-inactivated in a boiling H2O bath for 1 min. After cooling, 25 µg of Crotalus atrox venom was added to each tube, the reactions were mixed, incubated at 30°C for 10 min, and inactivated by incubation in a boiling H2O bath for 1 min. After cooling on ice, a 1:3 slurry (w:v) of AG1-X2 ion-exchange resin was added in order to separate [3H]adenosine (cAMP-PDE) or [3H]guanosine (cGMP-PDE) from the phosphate and nonhydrolyzed cyclic nucleotide in the reactions. Following a 1-min centrifugation at 12 000 x g (4°C), the equal volumes of the supernatants were collected from all reactions, and counts per minute of 3H were measured by spectrophotometry. Cyclic nucleotide PDE activities within different treatment groups were normalized based on their protein content.
Statistical Analyses
Duplicate treatments were administered for the measurement of steroidogenesis. Cyclic nucleotide PDE activity assays were conducted using single treatment groups per independent experiment. All experiments were repeated a minimum of 3 times. Mean values from independent experiments were statistically analyzed by unpaired t-test and multiple comparisons were performed using one-way ANOVA followed by the Tukey test. Values were determined to be significant when P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As shown in Figure 1a, FSH-dependent E2 synthesis was synergistically increased by IGF-I. Follicle-stimulating hormone-supported E2 synthesis was suppressed by HGF (30 ng/ml), and at all doses of HGF that were tested, the synergistic effect of IGF-I on FSH-stimulated E2 production was blocked.
|
A previous report has shown that HGF blocked the FSH-directed conversion of E1 into E2 [9]. Present data revealed that HGF suppressed the IGF-I-induced increase in FSH-supported E1 production in granulosa cells (Fig. 1b). Although the effect of IGF-I on FSH-stimulated E1 production was reduced by HGF, E1 concentrations remained significantly higher in these cultures compared with cells that received FSH and HGF in the absence of IGF-I.
Although preovulatory granulosa cells do not normally produce copious amounts of P4, a change from E2 to P4 production marks the onset of granulosa cell luteinization, and is therefore an important marker of terminal differentiation of these cells. As seen in Figure 1c, IGF-I-dependent augmentation of FSH-supported P4 production was not significantly altered by HGF.
Cyclic Nucleotide PDEs and HGF
The nonspecific cyclic nucleotide PDE inhibitor, IBMX, was used in an attempt to block the HGF-directed suppression of E2 in FSH-stimulated granulosa cells. Hepatocyte growth factor down-regulated FSH-dependent E2 accumulation (Fig. 2), and the suppressive effect of HGF was reduced by coincubation with IBMX. The inhibitor did not significantly alter the suppressive actions of HGF on FSH + IGF-I-supported E2 production at 48 h.
|
To further correlate changes in PDE function with the HGF challenge, we measured the activities of cAMP-PDE and cGMP-PDE. Compared with control granulosa cells, FSH caused an increase in cAMP-PDE activity at 24 and 36 h (Fig. 3a). At 24 h, HGF did not alter basal cAMP-PDE activity (data not shown), but it suppressed FSH-supported cAMP-PDE activity. However, at 36 h, FSH-supported cAMP-PDE activity was not reduced in the presence of HGF.
|
In the presence of IGF-I, FSH-stimulated cAMP-PDE activity was down-modulated at 24 and 36 h. It was interesting that HGF significantly alleviated the IGF-I-dependent reduction in FSH-stimulated cAMP-PDE activity at 24 and 36 h (Fig. 3a).
At 24 h in vitro, cGMP-PDE activity was not significantly altered from basal levels by FSH, or by the combinations of FSH, HGF, and IGF-I that were tested. FSH increased cGMP-PDE activity above basal levels at 36 h, whereas HGF impaired FSH-induced and FSH + IGF-I-induced cGMP-PDE activity at 36 h (Fig. 3b).
Cyclic Nucleotide Accumulation
In order to account for HGF-directed changes in cyclic nucleotide production, cAMP and cGMP concentrations in granulosa cell cultures were measured. Compared with control cultures, FSH stimulated cAMP accumulation at 6, 24, and 36 h (Fig. 4a). HGF induced a significant (28%) reduction in FSH-stimulated cAMP content at 24 h but not 36 h. In the presence of IBMX, HGF did not impede FSH-supported cAMP accumulation (Fig. 4b).
|
FSH did not significantly affect basal cGMP concentrations at 6, 24, or 36 h. Cyclic GMP levels were not altered by HGF at 6 h, but were increased by HGF (in the presence and absence of FSH) at 24 h (Fig. 4c). Compared with controls and FSH-stimulated cells, a moderate but nonstatistically significant elevation in cGMP content was detected in the presence of HGF at 36 h.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The present study showed that HGF attenuated the FSH-dependent and IGF-I-dependent production of E1 and E2. In contrast, de novo steroidogenesis (i.e., P4 production) was not impaired by HGF. With regard to the mechanism of action that is used by HGF to suppress E2 production in granulosa cells, it was previously reported that HGF did not disrupt E2 synthesis in the presence of Bu2-cAMP [9]. When this observation is combined with the findings presented here, a preliminary model can be developed for how HGF attenuates FSH-directed cAMP signal transduction (and ultimately E2 production) in granulosa cells. Consider that HGF suppressed FSH-stimulated cAMP production in a time-specific manner, and that this effect was blocked by IBMX. Furthermore, IBMX alleviated the suppressive effect of HGF on FSH-stimulated E2 production. These findings suggest that HGF-directed alterations in cyclic nucleotide PDE activity mediate, at least in part, the observed down-modulatory effect of HGF. Although data in this report show that HGF causes changes in cyclic nucleotide PDE activities, time-dependent effects of HGF at several loci within the cAMP-dependent signaling cascade are indicated. For example, the effects of HGF on cAMP-PDE activity do not explain the suppressive effect of HGF on FSH-stimulated cAMP content at 24 h. These data suggest that sites leading to the generation of cAMP (i.e., FSH receptor, Gs proteins, adenylyl cyclase, or a combination of these) are impaired. Whereas at 36 h neither FSH-stimulated cAMP content nor cAMP-PDE activity were reduced by HGF, and this indicates that if HGF indeed disrupts cAMP-directed signaling, the PKA phosphotransferase, its substrates, or both could be compromised.
The data herein show that HGF can comodulate cAMP-PDE and cGMP-PDE in granulosa cells; moreover, HGF stimulated cGMP accumulation in granulosa cell cultures. With regard to cGMP and granulosa cell steroidogenesis, it has been demonstrated that FSH-dependent E2 production in rat granulosa cells is blocked by exogenous cGMP, as well as by the activation of cGMP-mediated signal transduction [20]. Therefore, hormones that increase cGMP-PDE activity may dampen any rise in cGMP, and thus indirectly support E2 production. It is interesting that HGF stimulated cGMP accumulation at 24 h, but did not induce profound changes in cGMP-PDE activity. In contrast, at 36 h, HGF did not stimulate a statistically significant elevation in cGMP, whereas cGMP-PDE was impaired. Collectively, these results point toward HGF-orchestrated temporal changes in cGMP production and cGMP-PDE activity that support cGMP-dependent signal transduction in granulosa cells. This is admittedly a first step in identifying what appears to be a complex and intricate regulatory system involving time-dependent and hormone-dependent changes in the generation of cyclic nucleotides and the regulation of cyclic nucleotide PDE activities in granulosa cells.
The effects of HGF and IGF-I on PDEs are puzzling considering that HGF impaired FSH-stimulated and IGF-I-stimulated E2 production; however, IBMX did not block the HGF-dependent down-regulation. It has been shown that IGF-I stimulates FSH-dependent cAMP accumulation in rat granulosa cells [6, 21]; and in a simple model, this mechanism would explain how IGF-I augments FSH-dependent E2 production. However, other reports showed no stimulatory effect of IGF-I on FSH-stimulated cAMP concentrations in human and rat granulosa cells in vitro [22, 23]. In this regard, the present data showed that IGF-I modulates FSH-stimulated cAMP-PDE and cGMP-PDE activities in rat granulosa cells. On the basis of these results, it would appear that selective, time-dependent mechanisms mediate cyclic nucleotide PDE activity, depending on the hormonal milieu to which granulosa cells are exposed. If IGF-I-induced reductions in cAMP-PDE activity support an increase in intracellular cAMP concentrations (as has been documented in IGF-I-stimulated rat granulosa cells [21]), this would presumably promote a key signaling mechanism that promotes E2 production. However, HGF and IGF-I each reduced FSH-stimulated cAMP-PDE activity, yet HGF and IGF-I have opposite effects on FSH-stimulated E2 production.
Many of the aforementioned uncertainties may be clarified as more is understood about the precise complement (and regulation) of the cyclic nucleotide PDEs that are expressed in granulosa cells. With regard to the specific cAMP-PDE that are present in granulosa cells, a previous report has shown that granulosa cells from PDE4-deficient mice lose hormone responsiveness in vitro [4]. Furthermore, it has been suggested that inhibition of cyclic nucleotide PDE activity instigates a positive feedback loop in which lowered cAMP-PDE activity stimulates cAMP accumulation, which in turn, enhances cAMP-PDE activity [3]. In other words, this highly orchestrated cycle regulates the waning and waxing concentrations of cAMP that modulate cAMP-PDE activity. The resultant activation-inactivation pattern of cAMP signaling is, importantly, necessary in order for granulosa cells to maintain FSH responsiveness. By impairing the FSH-directed induction of cAMP-PDE (as was shown in this report), it is possible that HGF would then impede the previously suggested model for cAMP/cAMP-PDE feedback [3], and consequently, FSH-stimulated E2 synthesis would be impaired.
In order to better conceptualize the seemingly paradoxical relationships between FSH, HGF, and IGF-I and their regulation of cyclic nucleotide content and PDE activities, several additional factors must be considered. For example, virtually nothing is known about comodulation of and cross talk between a growing list of signaling molecules that are regulated by FSH, HGF, and IGF-I in granulosa cells. Experimental evidence has shown that HGF, IGF-I, and FSH can each affect phosphatidylinositol-3-kinase (PI-3K)/Akt-dependent signal transduction in granulosa cells [9, 16, 24]; thus providing one example of a shared signaling cascade. Whether or not PI-3K/Akt signaling can regulate cyclic nucleotide-dependent pathways in granulosa cells remains to be proven, however, we suggest that as more is understood about the signal transduction pathways that are affected by FSH, HGF, and IGF-I in granulosa cells, additional examples of cross talk, comodulation (or both), that ultimately couples to FSH-dependent E2 synthesis will be revealed.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by grant HD38277-01 from the National Institutes of Health and grants A1A 01-1 and A1A 02-1 from the Southern Illinois University School of Dental Medicine to R.J.Z. 
2 Correspondence: Rob J. Zachow, Southern Illinois University, School of Dental Medicine, 2800 College Ave., Alton, IL 62002. FAX: 618 474 7124; rzachow{at}siue.edu
Accepted: March 1, 2002.
Received: November 5, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
inhibits gonadotropin hormonal action in nontransformed ovarian granulosa cells. J Biol Chem 1989; 264:11591-11597 messenger ribonucleic content in rat ovarian granulosa cells by hepatocyte growth factor and follicle-stimulating hormone. Biol Reprod 2000; 62:1851-1857This article has been cited by other articles:
|
|
 |
|
  R. Zachow and M. Uzumcu The hepatocyte growth factor system as a regulator of female and male gonadal function J. Endocrinol., December 1, 2007; 195(3): 359 - 371. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Uzumcu, Z. Pan, Y. Chu, P. E Kuhn, and R. Zachow Immunolocalization of the hepatocyte growth factor (HGF) system in the rat ovary and the anti-apoptotic effect of HGF in rat ovarian granulosa cells in vitro. Reproduction, August 1, 2006; 132(2): 291 - 299. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O Gubbay, W Guo, M T Rae, D Niven, A F Howie, A S McNeilly, L Xu, and S G Hillier Anti-inflammatory and proliferative responses in human and ovine ovarian surface epithelial cells Reproduction, November 1, 2004; 128(5): 607 - 614. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
0.05) in estrogen content as a result of treatment are indicated by different letters
