Adenosine 5 (cid:1) -Monophosphate-Activated Protein Kinase Regulates Progesterone Secretion in Rat Granulosa Cells

The AMP-activated protein kinase (AMPK) is a major regula- tor of energy metabolism involved in fatty acid and cholesterol synthesis. In the ovary, cholesterol plays a key role in steroid production. We report the presence of AMPK in rat ovaries, and we have investigated its role in granulosa cells. We show using RT-PCR and Western blot that the mRNAs for the (cid:1) 1/2 and (cid:2) 1/2 subunits and the proteins are found in the ovaries. Immunohistochemistry localized the (cid:1) 1 AMPK subunit in granulosa cells, corpus luteum, and oocyte and less abundantly in theca cells. Treatment with 1 m M 5-amino-im-idazole-4-carboxyamide-1- (cid:2) - D -ribofuranoside (AICAR), an ac- tivator of AMPK, increased dose-dependent and time-depen-dent phosphorylation of AMPK (cid:1) 1 on Thr172 in primary granulosa cells. Simultaneously, phosphorylation of acetyl-coenzyme A carboxylase at Ser79 was also increased. AICAR treatmentfor48hhalvedprogesteronesecretion,3 (cid:2) -HSDpro-tein and mRNA levels, and phosphorylation of both basal MAPK ERK1/2 and p38 and in response to IGF-I and/or FSH in granulosa cells. AICAR treatment (1 m M ) had no detectable effect on basal and FSH- and/or IGF-I-induced estradiol production and on granulosa cell proliferation or viability. Adenovirus-mediatedexpressionofdominantnegativeAMPKtotallyabolishedtheeffectsofAICARonprogesteronesecre-tion,3 (cid:2) -HSD protein production, and MAPK ERK1/2 and p38 phosphorylation. Moreover, we showed using specific in- hibitors of ERK1/2 and p38 MAPK that the MAPK ERK1/2 and not p38 is involved in progesterone secretion and 3 (cid:2) -HSD expression, strongly suggesting that the activation of AMPK in response to AICAR reduces progesterone production throughtheMAPKERK1/2signalingpathwayinratgranulosacells.( Endocrinology and/or IGF-I receptor (FSH-R and IGF-I-R) activation lead to stimulation of the MAPK ERK1/2 and p38 signaling pathways. However, only the MAPK ERK1/2 is involved in the 3 (cid:2) -HSD expression and progesterone production. B, AMPK activation induced by AICAR treatment (1 m M ) for 24 or 48 h reduces markedly the MAPK ERK1/2 and p38 phosphorylation in response to FSH and/or IGF-I stimulation. The reduction in MAPK ERK1/2 phosphorylation leads to an inhibition in the 3 (cid:2) -HSD expression and progesterone secretion, whereas the consequences of the inhibition in the p38 MAPK phosphorylation are yet unknown.

AMPK is a trimeric enzyme consisting of a catalytic subunit, ␣, and two regulatory subunits, ␤ and ␥ (20); there are isoforms of each subunit (␣1, ␣2, ␤1, ␤2, ␥1, ␥2, and ␥3) with multiple possible combinations (21,22). AMP binds to the ␥-subunit of AMPK and facilitates phosphorylation of threonine 172 (Thr172) of the ␣-subunit by an upstream kinase, AMPK kinase (AMPKK), now known as LKB1 (23), resulting in increased enzyme activity (24). The adenosine analog, 5-aminoimidazole-4-carboxamide-1-␤-d-ribonucleoside (AICAR), is a potent activator of the AMPK system (25). The metabolic effects of AICAR have been studied extensively in skeletal muscle, adipose tissue, and liver. Recently, the presence of AMPK and the effect of AICAR have been investigated in oocytes (26). In this cell type, AMPK may play an important role in meiotic induction (26). However, until now, the presence and the roles of AMPK have never been studied in other ovarian structures, such as granulosa cells.
Growth and differentiation of granulosa cells are regulated by gonadotropins such as FSH (27) and local regulators such as IGF-I. FSH binds to its cognate G protein-coupled receptor and activates the membrane-associated adenyl cyclase, resulting in a rise in the intracellular cAMP level. cAMP subsequently activates cAMP-dependent protein kinase A, leading to the phosphorylation or expression of cellular proteins controlling granulosa cell proliferation and differentiation (28). For example, FSH stimulates the expression of the proliferating cell nuclear antigen (PCNA) or cyclin D2 (28) and also the expression of steroidogenic enzymes such as the cholesterol side-chain cleavage cytochrome p450 (P450scc), 3␤-hydroxysteroid dehydrogenase (3␤-HSD), and the steroidogenic acute regulatory protein (StAR), which is a protein that participates in the transport of cholesterol from the mitochondrial outer membrane to the inner membrane (29). AMPK is involved in the cholesterol synthesis (1) and cell proliferation in response to different factors in various cell types (7,8). Thus, it could regulate cell growth and also steroidogenesis of rat granulosa cells in response to FSH.
The actions of IGF-I on granulosa cells are mediated by the type I receptor, which activates two main signaling pathways, MAPK, and more importantly ERK1/2, and phosphatidyl inositol 3Ј kinase/Akt (30 -32). Studies with cultured granulosa cells have suggested that IGF-I, in addition to its growth-promoting activity, plays an important role in affecting progesterone secretion in synergy with FSH. IGF-I increases FSH-stimulated cAMP accumulation in rat granulosa cells (33), and interactions between AMPK activation and the effects of IGF-I have been reported in various cell systems (34 -36). Thus, AMPK activation could affect IGF-I actions in granulosa cells.
The aims of this study were, first, to characterize AMPK in the rat ovary and, second, to investigate its role in IGF-I-induced and FSH-induced proliferation and differentiation of granulosa cells. We found large amounts of AMPK␣ and -␤ in corpus luteum, oocyte, and granulosa cells, but less in theca cells. In granulosa cells, the activation of AMPK in response to AICAR reduced production of progesterone but not estradiol. This was associated with a reduction in the amounts of 3␤-HSD protein and mRNA and a decrease in phosphorylation of both MAPK p38 and ERK1/2. However, using specific inhibitors of MAPK kinase (MEK1/2) and p38, we showed that MAPK ERK1/2 and not p38 is involved in progesterone secretion and 3␤-HSD expression, strongly suggesting that the activation of AMPK in response to AICAR reduces progesterone production through the MAPK ERK1/2 signaling pathway in rat granulosa cells.

Hormones and reagents
Purified ovine FSH-20 (lot no. AFP-7028D, 4453 IU/mg; FSH activity ϭ 175 times activity of ovine FSH-S1) used for culture treatment was a gift from the National Hormone Pituitary Program (National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD

Animals
All procedures were approved by the Agricultural Agency and the Scientific Research Agency and conducted in accordance with the guidelines for Care and Use of Agricultural Animals in Agricultural Research and Teaching.
Immature female rats of the Wistar strain were purchased from Janvier Laboratories (Genest St. Isle, France). They were housed with controlled temperature and photoperiod (10 h dark, 14 h light; lights on from 0600 -2000 h). The animals had ad libitum access to food and water. Ovaries were also collected from immature (21-d-old) rats after one treatment with 25 IU pregnant mare serum gonadotropin (PMSG) for 48 h to induce follicle growth. Some rats received a single ip injection of 25 IU human chorionic gonadotropin (hCG) after the PMSG treatment to induce ovulation and luteinization, and some of these ovaries were collected for immunohistochemistry analyses. Ovaries were then fixed in 4% paraformaldehyde, dehydrated in alcohol baths, and embedded in paraffin. Other ovaries were used to collect oocytes. The oviducts were dissected out, and oocytes were recovered by oviductal flushing. The cumulus mass surrounding ovulated oocytes was dispersed using 0.1% hyaluronidase (Sigma) in M2 medium (37). The oocytes were then examined under a phase contrast microscope and collected in Trizol reagent for total RNA preparation.

Isolation and culture of rat granulosa cells
Immature female rats were injected sc with diethylstilbestrol (DES, 1 mg/d) every day for 3 d. On the third day of DES treatment, the animals were killed and the ovaries removed aseptically and transferred to culture medium. Granulosa cells were harvested by puncturing the follicles allowing expulsion of the cells. Cells were recovered by centrifugation, washed with fresh medium, and counted in a hemocytometer. The culture medium used was McCoy's 5A supplemented with 20 mmol/liter HEPES, 100 U/ml penicillin, 100 mg/liter streptomycin, 3 mmol/liter l-glutamine, 0.1% BSA, 50 g/liter insulin, 0.1 mol/liter androstenedione, 5 mg/liter transferrin, and 20 g/liter selenium and 5% fetal bovine serum (FBS). The cells were initially cultured for 48 h with no other treatment and then incubated in fresh culture medium with or without test reagents for the appropriate time. All cultures were performed under a water-saturated atmosphere of 95% air/5% CO 2 at 37 C.

Thymidine incorporation into granulosa cells
Granulosa cells (2 ϫ 10 5 viable cells/500 l) were cultured in 5-ml polypropylene Falcon tubes with McCoy's 5A medium in the presence or absence of AICAR (1 mm), FSH (10 Ϫ8 m) and IGF-I (10 Ϫ8 m). Cultures were maintained at 37 C under 5% CO 2 in air. After 24 h of culture with 1 Ci/tube of [ 3 H]thymidine (Amersham Life Science, Arlington Heights, IL), cells were washed once and resuspended in ice-cold PBS by centrifugation at 2000 ϫ g for 30 min. The radioactivity in the cells was determined after resuspension by scintillation counting in a ␤-photomultiplier.

Adenoviruses and infection of rat granulosa cells
Dominant negative AMPK adenovirus (␣1-DN) was constructed from AMPK␣1 carrying the Asp-157 to Ala (D157A) mutation as described previously (38). Recombinant adenovirus was propagated in HEK293 cells, purified by cesium chloride density centrifugation, and stored as described previously (38). Rat granulosa cells were infected with 2 or 20 pfu/cell adenovirus in serum-starved McCoy's 5A. After 2 h, one additional volume of serum-starved McCoy's 5A was added, and the cells were cultured for 24 h in the presence or absence of FSH and/or IGF-I and 1 mm AICAR. Preliminary studies revealed that within 24 h of infection (20 pfu/cell) with a GFP-expressing virus, the majority of granulosa cells (Ͼ90%) expressed GFP.

RNA isolation and RT-PCR
Total RNA was extracted from whole tissue (ovary and dissected corpus luteum) or from cultured granulosa cells using Trizol reagent according to the manufacturer's procedure (Invitrogen). RNA was quantified by measuring the absorbance at 260 nm. Samples were stored at Ϫ80 C until use.
Total RNA (1 g) was reverse transcribed in a 20-l reaction mixture containing 50 mm Tris-HCl (pH 8.3), 75 mm KCl, 3 mm MgCl 2 , 200 m of each deoxynucleotide triphosphate (Amersham, Piscataway, NJ), 50 pmol oligo(dT)15, 5 U ribonuclease inhibitor, and 15 U Moloney murine leukemia virus reverse transcriptase. RT reactions were carried out at 37 C for 1 h. Single-strand cDNAs were amplified with specific sets of primer pairs designed to amplify parts of the different AMPK isoforms as described in Table 1. PCRs were carried out using 2 l of the RT reaction mixture in a volume of 50 l containing 10 mm Tris-HCl (pH 9.0), 50 mm KCl, 1.5 mm MgCl 2 , 200 mm of each deoxynucleotide triphosphate, 10 pmol of each primer, and 1 U Taq polymerase. The samples were processed for 35 PCR cycles (95 C for 1 min, 58 C for 1 min and 72 C for 1 min), with a final extension step at 72 C for 10 min. PCR products were visualized in an agarose gel (1.5%) stained with ethidium bromide, and the DNAs were extracted from the agarose using the QIAEX II (QIAGEN, Hilden, Germany) gel extraction kit and sequenced in both direction using the Dye terminator kit on an ABI Prism automated sequencer, model 377 (Biomolecular Research Facility, University of Virginia, Charlottesville, VA). PCR amplifications with RNA were performed in parallel as negative controls (data not shown).

Northern blot
Total RNA from granulosa cells (10 g) was separated by denaturing formaldehyde electrophoresis and then transferred to a nylon membrane by capillarity overnight and immobilized by exposure to UV light as previously described (39,40). Blots were prehybridized for 2 h at 42 C in a buffer containing 50% formamide, 5ϫ Denhardt's, 1% SDS, 5ϫ standard saline citrate, and 100 g/ml denatured salmon sperm. Blots were then hybridized overnight at 42 C with 2 ϫ 10 6 cpm/ml [␣-32 P]dCTP-labeled DNA probe in a buffer containing 50% formamide, 2.5ϫ Denhardt's, 1% SDS, 5ϫ standard saline citrate, 1% dextran, and 100 g/ml denatured salmon sperm. 3␤-HSD and StAR probes obtained by RT-PCR using the primers described in Table 1 were labeled using the Rediprime labeling kit (Amersham, Piscataway, NJ). After high-stringency washings, membrane-incorporated radioactivity was also quantified using a STORM apparatus. The integrity and the quantification of different transcripts were assessed using the human RNA 18S probe as a control (Ambion, Austin, TX).

Immunohistochemistry
Ovaries embedded in paraffin were serially sectioned at a thickness of 7 m. Sections were deparaffinized, hydrated, and microwaved for 5 min in antigen unmasking solution (Vector Laboratories, Inc., AbCys, Paris, France) and then left to cool to room temperature. After washing in a PBS bath for 5 min, sections were immersed in peroxidase blocking reagent for 10 min at room temperature to quench endogenous peroxidase activity (Dako Cytomation; Dako, Ely, UK). After two washes in a PBS bath for 5 min, nonspecific background was eliminated by blocking with 5% lamb serum in PBS for 20 min, followed by incubation overnight at 4 C with PBS containing rabbit primary antibody raised against either AMPK␣1 or pACC (1:200; Upstate Biotechnology Inc., Lake, Placid, NY). Sections were washed twice for 5 min each time in a PBS bath and were incubated for 30 min at room temperature with a ready-to-use labeled polymer-HRP antirabbit (Dako Cytomation Envi-

Progesterone and estradiol RIA
The concentration of progesterone and estradiol in the culture medium of granulosa cells was measured after 48 h of culture by a RIA protocol as previously described (41) and adapted to measure steroids in cell culture media. The limit of detection of progesterone was 12 pg/tube (60 pg/well), and the intra-and interassay coefficients of variation were less than 10 and 11%, respectively. The limit of detection of estradiol was 1.5 pg/tube (7.5 pg/well), and the intra-and interassay coefficients of variation were less than 7 and 9%, respectively. Results were expressed as the amount of steroids secreted for 48 h per 100 g protein.

Statistical analysis
All experimental data are presented as means Ϯ sem. One-way ANOVA was used to test differences. If ANOVA revealed significant effects, the means were compared by Newman's test, with P Ͻ 0.05 considered significant.

AMPK␣ and -␤ subunit mRNA and protein in rat ovary
RT-PCR analysis with RNA from adult rat ovary, dissected corpus luteum, fresh granulosa cells and oocyte resulted in the amplification of four cDNAs corresponding to fragments of two different isoforms of the catalytic ␣-subunit, ␣1 (710 bp) and ␣2 (1652 bp), and two different isoforms of the regulatory ␤-subunit, ␤1 (802 bp) and ␤2 (811 bp), of AMPK (Fig. 1A). Immunoblotting of protein extracts revealed one band corresponding to the AMPK␣1 (62 kDa) and two bands corresponding to the AMPK␤1 (40 kDa) and -␤2 (34 kDa) (Fig. 1B), showing that both subunits ␣ and ␤ of AMPK are produced in rat ovary and more particularly in granulosa cells and corpus luteum. Immunohistochemistry with ovarian sections from immature rats treated with PMSG alone or PMSG and then hCG or with neither confirmed the immunoblot findings and demonstrated AMPK␣1 in oocytes and theca cells but less abundantly than in granulosa cells (Fig.  1C). Furthermore, we have also shown that ACC phosphorylated on Ser79 (Fig. 1C) or ACC total protein (data not shown), a target of AMPK␣ colocalized with AMPK␣1. Thus, AMPK is present in the different structures of the ovarian follicle.

Phosphorylation of the AMPK␣ in granulosa cells in response to AICAR treatment
We next determined the effects of stimulation with various concentrations and various times by the AMPK activator AICAR on the phosphorylation state of the ␣-subunit of AMPK in immature granulosa cells from DES-primed rats. We used a specific anti-phospho-Thr172 AMPK antibody. Treatment of granulosa cells with AICAR elicited a phosphorylation of AMPK in a dose-and time-dependent manner (Fig. 2, A and B). AMPK was phosphorylated maximally (about 3-fold) by 1 mm AICAR, and this concentration was therefore used for all subsequent experiments. The AICAR treatment (1 mm) rapidly increased AMPK phosphorylation (30 min), and the effect reached a maximum 3.5-fold stimulation after 48 h. We also assessed the AMPK activity by measuring the phosphorylation of its downstream target, ACC. Indeed, it is well known that AMPK inhibits ACC by increasing ACC phosphorylation at Ser79 (4). Western blot analysis showed that AICAR markedly induced phosphorylation of ACC at Ser79 in a time-dependent fashion, paralleling the stimulation of Thr172 phosphorylation of AMPK␣ at 30 min, 1 h (Fig. 2C), and 48 h (data not shown). Thus, AMPK is active in rat granulosa cells in culture.

Effects of the AICAR treatment on FSH-and IGF-Istimulated progesterone and estradiol production in rat granulosa cells
We next investigated the effect of AICAR treatment on progesterone and estradiol production in rat granulosa cells. Cells were cultured for 48 h in serum-free medium with 1 mm AICAR in the presence and in the absence of IGF-I and/or FSH. As shown in Fig. 3A, in the presence and in the absence of IGF-I and/or FSH, AICAR treatment (1 mm) decreased progesterone secretion by at least half (P Ͻ 0.001), whereas it did not significantly affect estradiol secretion (Fig. 3B). Identical results were obtained when 1 mm AICAR and FSH/or IGF-I were included in the culture medium for 6, 12, and 24 h (data not shown).
We next examined whether this inhibitory effect of AICAR on progesterone production was a result of less of the two key enzymes of steroidogenesis (3␤-HSD and P450scc) and/or of StAR, an important cholesterol carrier. AICAR treatment (1 mm) in basal or IGF-I and/or FSH treatment halved (P Ͻ 0.001) the production of 3␤-HSD protein (Fig.  4A), whereas it had no effect on P450scc (Fig. 4B) and StAR proteins (Fig. 4C). Furthermore, Northern blot analysis indicated a decrease by about half in the level of 3␤-HSD mRNA in presence of AICAR, whether alone or combined with IGF-I and/or FSH (Fig. 4D). AICAR, in the same condition, had no effect on StAR mRNA (Fig. 4E). Thus, the decrease in progesterone secretion in response to AICAR treatment appears to be caused by a reduction in the amounts of 3␤-HSD protein and mRNA.

Effects of the overexpression of dominant negative AMPK on the progesterone and estradiol production in rat granulosa cells
Although AICAR has been widely used as a pharmacological activator of AMPK, AICAR alone has many cellular effects (25,42). Therefore, we tested whether the AICARinduced decrease in the progesterone production and 3␤-HSD expression was indeed mediated by AMPK. We infected rat granulosa cells for 24 h with dominant negative (␣1-DN) AMPK using an adenovirus vector. After infection with control (Ad.GFP) or Ad.␣1-DN viruses, granulosa cells were analyzed by Western blotting for production of mutant and endogenous AMPK␣1 subunits. ␣1-DN and GFP proteins were detected in granulosa cells infected with Ad.␣1-DN and Ad.GFP (2 or 20 pfu/cell) for 24 h, respec-tively (Fig. 5A). Furthermore, ␣1-DN significantly attenuated basal AMPK Thr172 phosphorylation (Fig. 5A). Infection of granulosa cells with Ad.GFP had no effect on AMPK␣1 expression or AMPK Thr172 phosphorylation (Fig. 5A). Expression of ␣1-DN in rat granulosa cells strongly reduced the AICAR-induced decrease in the production of both progesterone and 3␤-HSD protein in response to FSH (24 h) (Fig. 5B) or IGF-I (24 h) (Fig. 5C). Infection of cells with a control GFP virus for 24 h had no effect (data not shown). Expression of ␣1-DN had no effect on estradiol production in response to FSH, AICAR, or IGF-I (Fig. 5, B and C).

Effects of the AICAR treatment on granulosa cell number
We also investigated whether the dose of AICAR used affected the number of granulosa cells in culture, either by induction of mitosis or by increasing the cell death by apoptosis. [ 3 H]Thymidine incorporation by granulosa cells treated with 1 mm AICAR was tested after 24 h in culture in the presence or in the absence of IGF-I and/or FSH. As expected, IGF-I and FSH treatment significantly increased [ 3 H]thymidine incorporation (Fig. 6A (43). However, AICAR treatment had no effect on either the basal and IGF-I-or  ). B, Detection of AMPK␣1 and -␤1/2 proteins by immunoblotting. Protein extracts (50 g) were resolved by SDS-PAGE, transferred to a nitrocellulose membrane, and incubated with specific antibodies against either AMPK␣1 or AMPK␤1/2 isoforms and tubulin-␣. C, Localization of AMPK␣1 isoform and ACC phosphorylated on Ser79 in rat ovary by immunohistochemistry: AMPK␣1 in small antral follicle in immature rat ovary (a. ϫ10), in large follicle in ovary from rat treated with PMSG for 48 h (b, ϫ10; and c, ϫ40), in corpus luteum in ovary from rat treated with PMSG for 48 h and then with hCG for 24 h (d, ϫ4; and e, ϫ20) and phospho-ACC Ser79 in corpus luteum in ovary from rat treated with PMSG for 48 h and then with hCG for 24 h (h, ϫ4), in large follicle in ovary from rat treated with PMSG for 48 h (i, ϫ10; and j, ϫ20). Note that AMPK␣1 and pACC are observed in granulosa cells (G), in corpus luteum (CL) and oocyte (O), and less abundantly in theca cells (T). Negative controls included a section incubated with nonimmune serum (f, ϫ4) and sections incubated with AMPK␣1 (g, ϫ10) and pACC (k, ϫ10) antibodies that were first preincubated with their respective blocking peptides (20 g/ml). AT, Antrum FSH-stimulated state (Fig. 6A). These results were confirmed by evaluating the PCNA level by Western blotting (Fig. 6B). As revealed by the Annexin-V-Fluos staining kit, there was no significant difference in the apoptosis rates between controls and cells treated with 1 mm AICAR for 48 h (data not shown). Thus, AICAR treatment (1 mm) for 48 h did not affect granulosa cell number or death.

Signaling pathways involved in the inhibitory effect of AICAR mediated by AMPK on progesterone production in rat granulosa cells
It has been suggested that AICAR can either stimulate or inhibit the MAPK ERK1/2 and p38 and Akt signaling pathways in different cell types (34, 44 -46). These signaling pathways are important for the proliferation and steroidogenesis of rat granulosa cells. Therefore, we first examined the ability of AICAR treatment to modulate the activation of these signaling pathways and, second, determined which signaling pathways may be involved in reducing progesterone production induced by AICAR treatment in rat granulosa cells (Figs. 7, 8, and 9).
Quantitative analysis of Western blots indicated that AICAR (1 mm) strongly inhibited phosphorylation of ERK1/2 MAPK, maximal after 60 min in the basal state when the cells were in presence of serum (5% FBS) (Fig. 7A). AICAR treatment (1 mm) also approximately halved ERK1/2 phosphorylation in the basal and FSH-and/or IGF-I-stimulated conditions used for assaying progesterone and estradiol production (48 h of stimulation in serum-free medium) (Fig. 7B). We observed similar results when the cells were stimulated or not with FSH or IGF-I for 24 h in serum-free medium (Fig. 7C). Furthermore, in these latter conditions,

FIG. 2. Effect of AICAR on AMPK and ACC phosphorylation in rat granulosa cells. A and B, Immature granulosa cells from DES-primed rats were incubated with the indicated concentrations of AICAR for 24 h (A) or incubated with 1 mM AICAR for various times (B).
Lysates were then prepared and subjected to Western blotting using anti-phospho-AMPK-Thr172 and anti-AMPK␣1 as a protein loading control. Note that AMPK phosphorylation in control medium (without AICAR) did not vary significantly over 48 h. The results are expressed as the means Ϯ SEM of the pAMPK/AMPK ratio in comparison with the basal state. *, P Ͻ 0.05 relative to the value in the absence of AICAR. These experiments are representative of three independent experiments. C, Cells were incubated with 1 mM AICAR for various times. Lysates were then prepared and subjected to Western blotting using anti-phospho-ACC-Ser79 and anti-ACC as a protein loading control. The results are expressed as means Ϯ SEM of the pACC/ACC ratio in comparison with the basal state. *, P Ͻ 0.05; ***, P Ͻ 0.001 relative to the value in the absence of AICAR. These experiments are representative of three independent experiments. AICAR-decreased MAPK ERK1/2 phosphorylation was significantly restored by expression of the ␣1-DN adenovirus (Fig. 7C). However, no such effect was seen on infecting the cells with a control GFP virus (data not shown). We showed, using the MEK1/2-specific inhibitor U0126 that the MAPK ERK1/2 signaling pathway is partly involved in progesterone secretion and 3␤-HSD production in basal (Fig. 7D) and FSH-and IGF-I-stimulated conditions (Fig. 7E). Indeed, in the basal state, U0126 dose-dependently inhibited progesterone secretion (P Ͻ 0.001 at 5 and 10 m). U0126 also dose-dependently inhibited progesterone secretion and 3␤-HSD protein production in response to FSH and IGF-I treatment (P Ͻ 0.001 at 5 and 10 m). However, even with the highest dose of U0126 (10 m), this inhibition was not total, because there remained a 2-fold increase of progesterone secretion in response to FSH or IGF-I, despite a total inhibition of ERK1/2 phosphorylation (Fig. 7E), suggesting the involvement of other signaling pathways. AICAR (1 mm) treatment alone approximately halved ERK1/2 phosphorylation, progesterone secretion, and 3␤-HSD production. However, when AICAR and U0126 (at the lower dose) were combined, phosphorylation of ERK1/2 was totally inhibited, and progesterone secretion and 3␤-HSD production were reduced to the same level as that obtained with the higher dose of U0126. These results suggest an additive effect of AICAR and U0126 (0.5 m) treatments on the inhibition of ERK1/2 phosphorylation, 3␤-HSD expression, and progesterone production (Fig. 7E).
Quantitative analysis of Western blots also indicated that AICAR (1 mm) significantly increased phosphorylation of p38 MAPK in a time-dependent manner in the basal state when cells were in the presence of serum (5% FBS) (Fig. 8A). However, it strongly reduced p38 MAPK phosphorylation in the basal and FSH-and/or IGF-I-stimulated conditions used for assaying progesterone and estradiol production (48 h of stimulation in serum-free medium) (Fig. 8B). We observed similar results when the cells were stimulated or not with FSH or IGF-I for 24 h without serum (Fig. 8C). Furthermore, in these latter conditions, AICARdecreased MAPK p38 phosphorylation was significantly restored by expression of the ␣1-DN adenovirus (Fig. 8C). However, no such effect was seen on infecting the cells with a control GFP virus (data not shown). We showed, using the p38-specific inhibitor SB202190 that the MAPK p38 signaling pathway was not involved in progesterone secretion and 3␤-HSD production in basal (Fig. 8D) and FSH-and IGF-I-stimulated conditions (Fig. 8E). Unlike the MAPK ERK1/2 and p38 signaling pathways, Akt phosphorylation on Ser473 was not affected by AICAR treatment in the basal state in the presence (Fig. 9A) or absence FIG. 5. AICAR-reduced progesterone secretion in rat granulosa cells infected with adenovirus-expressing AMPK mutant constructs. A, The expression and phosphorylation of AMPK in rat granulosa cells infected with adenovirus constructs expressing mutant AMPK or adenovirus expressing GFP. Granulosa cells were infected with 20 or 2 pfu/cell of the viruses indicated, and the cells were simultaneously stimulated with FSH or IGF-I as indicated in Materials and Methods. After 24 h, cells were prepared and the lysates were resolved by SDS-PAGE, transferred to nitrocellulose, and probed with anti-GFP, anti-AMPK␣1, and anti-phospho-AMPK-Thr172 antibodies. The immunoblots shown are representative of repeated experiments on two different samples of lysates with similar results. B and C, Progesterone and estradiol secretion and 3␤-HSD protein in granulosa cells expressing ␣1-DN in the presence or absence of 1 mM AICAR (24 h) and FSH (B) or IGF-I (C). Granulosa cells were stimulated with FSH (10 Ϫ8 M) or IGF-I (10 Ϫ8 M) in the presence or absence of 1 mM AICAR and infected or not with 20 pfu/cell of ␣1-DN virus for 24 h. The culture medium was then collected and analyzed for progesterone and estradiol content by RIA, and protein extracts were prepared from cells for immunoblotting with anti-3␤-HSD, anti-tubulin (as loading control), anti-AMPK␣1, and anti-phospho-AMPK-Thr172 antibodies. The data shown represent means Ϯ SEM from three independent experiments of the concentrations of progesterone and estradiol and the induction of the 3␤-HSD protein obtained from the ratio 3␤-HSD/tubulin-␣. ***, P Ͻ 0.001, *P Ͻ 0.05 relative to the value in absence of FSH, AICAR, and infection with ␣1-DN. (Fig. 9, B and C) of serum and in response to FSH and/or IGF-I (Fig. 9, B and C).

Discussion
Our results demonstrate that AMPK activation induced by AICAR treatment strongly reduces progesterone secretion both in the basal state and in response to FSH or IGF-I without affecting estradiol production in rat granulosa cells in culture. Our findings also indicate that AICAR-induced AMPK activation reduces only 3␤-HSD production but not P450scc or StAR production. Moreover, we showed that AICAR treatment for 48 h (the condition used to assay progesterone secretion) reduced ERK1/2 and p38 phosphorylation. We also show that inhibition of ERK1/2 phosphorylation and not p38 phosphorylation markedly reduces progesterone secretion and 3␤-HSD protein expression in rat granulosa cells. Together, these findings suggest that AMPK activation induced by AICAR treatment decreases progesterone secretion through the MAPK ERK1/2 and not p38 signaling pathway in rat granulosa cells (Fig. 10).
The potential involvement of ERK1/2 in the regulation of steroidogenesis in different steroid-producing cells appears to be contradictory. Some investigations show stimulatory effects of ERK1/2 on steroidogenesis (47)(48)(49), whereas others demonstrate inhibitory effects (50). For example, it has been shown that LH and FSH activate ERK1/2 and enhance steroid production in ovarian cells (48,49), whereas in cell lines derived from granulosa cells, stimulation of the ERK cascade by these same gonadotropins reduces steroid production (50). However, all these studies used different cell lines that do not necessarily have the same number of signaling molecules. Moreover, specific inhibitors are used to demonstrate that MAPK ERK1/2 is involved in the steroidogenesis, and the incubation time (long or short) of these inhibitors may affect ERK1/2 inhibition and consequently the different molecules involved in the regulation of steroidogenesis. In our study, we showed that a long-term inhibition of ERK1/2 reduces progesterone secretion associated with decreases in the 3␤-HSD production in primary rat granulosa cells. Moreover, these data were observed without modifying the amount of the P450scc or StAR produced (data not shown). In mouse adrenocortical Y1 cells, ERK1/2 activation increased steroid production through increased transcription of the StAR gene (47). Thus, MAPK ERK1/2 may play a key role in regulating the expression of the different molecules involved in progesterone synthesis in different cell types. MAPK ERK1/2 regulates target gene expression by activating downstream transcription factors. For example, it has been implicated in the regulation of steroidogenic factor 1 (SF-1) in the human breast cancer cells MCF-7 (51) and mouse adrenocortical Y1 cells (47). The 3␤-HSD type 2 promoter contains a consensus sequence for SF-1 (52). Thus, MAPK ERK1/2 may regulate 3␤-HSD production by phosphorylation of SF-1 and then SF-1 binding to regions of the 3␤-HSD promoter.
We showed that in culture medium with serum, short AICAR stimulation (from 1-60 min) increased p38 phosphorylation, whereas it decreased ERK1/2 phosphorylation. In serum-free medium, long AICAR stimulation decreased both p38 and ERK1/2 phosphorylation. Thus, the effect of AICAR on the MAPK p38 phosphorylation appears to be dependent on the time of stimulation and/or the presence of the serum in the culture medium. Short and long AICAR stimulation may lead to different physiological responses in rat granulosa cells. In these cells, the p38 signaling pathway is involved in cell rounding/aggregation (53). AMPK activation and p38 MAPK activity have been associated in several systems (54). For example, the inhibition of p38 MAPK with SB203580 inhibits the effect of AICAR on glucose uptake in clone 9 cells (55). In our study, short (5-60 min) or long (24 or 48 h) AICAR treatment had no effect on Akt phosphorylation. AICAR also had no effect on Akt activity in 3T3-L1 adipocytes (56). However, cross-talk between AMPK and Akt has been demonstrated for several cell systems, includ- ing endothelial cells (57) and the myocardium (58). In endothelial cells, it appears AMPK and Akt are both activated, whereas in the myocardium, Akt acts as a negative regulator of AMPK activity.
Several mechanisms may explain the inhibitory effect of AMPK activation on IGF-I-induced progesterone secretion. First, Sprenkle et al. (59) demonstrated that AMPK phosphorylated Ser621 of Raf-1, expressed in Escherichia coli or SF9 insect cells. Furthermore, it has been claimed that phosphorylation of Raf-1 on Ser621 by cAMP-dependent protein kinase (protein kinase A) negatively regulates progesterone secretion (60). Also, in NIH-3T3 cells, an antisense RNA for the AMPK catalytic subunit decreased AMPK activity and significantly diminished the effect of AICAR on IGF-I-induced Ras activation, and the subsequent ERK activation, indicating that its effect is indeed mediated by AMPK (34). Cells were then prepared, and the lysates were resolved by SDS-PAGE, transferred to nitrocellulose, and probed with anti-phospho-ERK1/2 antibodies. The blots were then stripped and reprobed with anti-ERK2 antibodies. Representative immunoblots are shown, and the experiment was repeated with similar results on three different samples of lysates. Blots were quantified, and the induction of the phosphorylation state of phospho-ERK1/2 was determined. The results are expressed as means Ϯ SEM. *, P Ͻ 0.05 relative to the value in the absence of AICAR. D, Cells were cultured in McCoy's medium without serum, in the presence or absence of different doses of U0126, and in the presence or absence of AICAR (1 mM) for 48 h. The culture medium was then collected and analyzed for progesterone and estradiol content by RIA. Results are means Ϯ SEM of three groups of granulosa cells. E, Granulosa cells were cultured for 48 h in serum-free medium in the presence or absence of IGF-I (10 Ϫ8 M), FSH (10 Ϫ8 M), or a combination of both and 1 mM AICAR or different doses of U0126. The culture medium was then collected and analyzed for progesterone and estradiol content by RIA, and cells were lysed and the lysates were resolved by SDS-PAGE, transferred to nitrocellulose, and probed with anti-phospho-ERK1/2 antibodies, anti-3␤-HSD antibodies, and ERK2 total. Representative blots are shown. Results are means Ϯ SEM of three groups of granulosa cells.
Second, in cell-free assays, as well as mouse C2C12 myotubes, AMPK rapidly phosphorylates IRS-1 on Ser789 in response to AICAR (35). Therefore, phosphorylation of IRS-1 on tyrosine residues in response to IGF-I, and thus IGF-I receptor signaling, may be reduced. Third, the activation of the AMPK pathway may be partly responsible for the increases in IGF-binding protein 1 (IGFBP-1) seen during the in vitro stresses of hypoxia (61) and glucose deprivation (62). Furthermore, in H4-II-E rat hepatoma cells, AICAR (150 m) stimulated IGFBP-1 secretion (63). IGFBP-1 is one of a family of proteins that regulate the activity of the IGFs as well as having IGF-independent effects on a variety of cellular functions. Thus, AMPK activation may increase the level of IGFBP-1, thus inhibiting the effects of IGF-I in granulosa cells. However, this third suggestion is unlikely because IGFBP-1 is only poorly expressed in granulosa cells.
In our study, we did not see an effect of AICAR on estradiol production. In our culture conditions, we directly add into the medium p450 aromatase substrate because granulosa cells are not able to synthesize it. Thus, it is likely that in the presence or absence of 1 mM AICAR and infected or not with 20 pfu/cell of the ␣1-DN virus as indicated in Materials and Methods. Cells were then prepared and the lysates were resolved by SDS-PAGE, transferred to nitrocellulose, and probed with anti-phospho-p38 antibodies. The blots were then stripped and reprobed with anti-p38 antibodies. Representative immunoblots are shown, and the experiment was repeated with similar results on three different samples of lysates. Blots were quantified, and the induction of the phosphorylation state of phospho-p38 was determined. The results are expressed as means Ϯ SEM. *, P Ͻ 0.05 relative to the value in the absence of AICAR. D, Cells were cultured in McCoy's medium without serum, in the presence or absence of different doses of SB202190, and in the presence or absence of AICAR (1 mM) for 48 h. The culture medium was then collected and analyzed for progesterone and estradiol content by RIA. Results are means Ϯ SEM of three groups of granulosa cells. E, Granulosa cells were cultured for 48 h in serum-free medium in the presence or absence of IGF-I (10 Ϫ8 M), FSH (10 Ϫ8 M), or a combination of both and 1 mM AICAR or different doses of SB202190. The culture medium was then collected and analyzed for progesterone and estradiol content by RIA, and cells were lysed and the lysates were resolved by SDS-PAGE, transferred to nitrocellulose, and probed with anti-phospho-p38 antibodies, anti-3␤-HSD antibodies, and p38 total. Representative blots are shown. Results are means Ϯ SEM of three groups of granulosa cells.
AICAR treatment does not affect p450 production and activity or that the effects on production are compensated by the effects on the activity or vice versa. We also did not see any effect of AICAR on granulosa cell proliferation induced by either IGF-I or FSH. Similarly, AICAR had no effect on apoptosis of rat granulosa cells. However, several studies have described AICAR treatment as either a positive or negative regulator of cell proliferation or cell death (64 -66). Thus, the FIG. 9. Effect of AICAR on Akt phosphorylation in rat granulosa cells. Granulosa cells were cultured in the presence of 5% FBS and incubated with 1 mM AICAR for various times (A), for 48 h in the presence or absence of IGF-I, FSH, or a combination of both and 1 mM AICAR (B), or for 24 h with FSH (C, left) or IGF-I (C, right) in the presence or absence of 1 mM AICAR and infected or not with 20 pfu/cell of the ␣1-DN virus as indicated in Materials and Methods. Cells were then prepared and the lysates were resolved by SDS-PAGE, transferred to nitrocellulose, and probed with anti-phospho-Akt Ser473 antibodies. The blots were then stripped and reprobed with anti-Akt antibodies. Representative immunoblots are shown, and the experiment was repeated with similar results on three different samples of lysates. Blots were quantified and the induction of the phosphorylation state of phospho-Akt was determined. The results are expressed as means Ϯ SEM. *, P Ͻ 0.05 relative to the value in the absence of AICAR.
FIG. 10. Schematic representation of AMPK-mediated AICAR effect and its mechanism involved in the progesterone secretion in rat granulosa cells. A, FSH and/or IGF-I receptor (FSH-R and IGF-I-R) activation lead to stimulation of the MAPK ERK1/2 and p38 signaling pathways. However, only the MAPK ERK1/2 is involved in the 3␤-HSD expression and progesterone production. B, AMPK activation induced by AICAR treatment (1 mM) for 24 or 48 h reduces markedly the MAPK ERK1/2 and p38 phosphorylation in response to FSH and/or IGF-I stimulation. The reduction in MAPK ERK1/2 phosphorylation leads to an inhibition in the 3␤-HSD expression and progesterone secretion, whereas the consequences of the inhibition in the p38 MAPK phosphorylation are yet unknown. effects of AICAR on the cell proliferation and apoptosis seem to be dependent on the cell type, the dose, and probably the duration of the treatment.
In this study, we tested the phosphorylation state of ACC at Ser79 to assess AMPK activity in rat granulosa cells. Indeed, ACC is an important substrate for AMPK, and its phosphorylation serves as an indirect assay for AMPK activation (67). AICAR treatment (1 mm) increased the phosphorylation of ACC on Ser79 in a time-dependent manner, consistent with activation of AMPK in rat granulosa cells. Although other protein kinases can phosphorylate ACC, the increase in ACC phosphorylation at Ser79 caused by AICAR in our study was inhibited by the overexpression of DN-AMPK (data not shown), which is consistent with AMPK being the mediator. AMPK inhibits ACC by phosphorylation at Ser79 (68 -70). Through its control of ACC, AMPK may make it easier to use fatty acids in granulosa cells. Indeed, as ACC catalyzes the conversion of acetyl-CoA to malonyl-CoA, AMPK, by inhibiting ACC, is able to decrease malonyl-CoA and minimize its inhibition of fatty acid oxidation (71). Thus, AMPK in rat granulosa cells may be involved in the interactions between metabolism and reproduction.
In summary, our present investigation has revealed that AMPK activation induced by AICAR treatment reduces progesterone secretion and 3␤-HSD production through the MAPK ERK1/2 signaling pathway in rat granulosa cells. We suggest that variation of AMPK activity in granulosa cells in response to different stimuli may play a role in the regulation of progesterone secretion.