Follicle-stimulating hormone activation of hypoxia-inducible factor-1 by the phosphatidylinositol 3-kinase/AKT/Ras homolog enriched in brain (Rheb)/mammalian target of rapamycin (mTOR) pathway is necessary for induction of select protein markers of follicular differentiation.

We sought to elucidate the role of AKT in follicle-stimulating hormone (FSH)-mediated granulosa cell (GC) differentiation. Our results define a signaling pathway in GCs whereby the inactivating phosphorylation of tuberin downstream of phosphatidylinositol (PI) 3-kinase/AKT activity leads to Rheb (Ras homolog enriched in brain) and subsequent mTOR (mammalian target of rapamycin) activation. mTOR then stimulates translation by phosphorylating p70 S6 kinase and, consequently, the 40 S ribosomal protein S6. Activation of this pathway is required for FSH-mediated induction of several follicular differentiation markers, including luteinizing-hormone receptor (LHR), inhibin-alpha, microtubule-associated protein 2D, and the PKA type IIbeta regulatory subunit. FSH also promotes activation of the transcription factor hypoxia-inducible factor-1 (HIF-1). FSH-stimulated HIF-1 activity is inhibited by the PI 3-kinase inhibitor LY294002, the Rheb inhibitor FTI-277 (farnesyltransferase inhibitor-277), and the mTOR inhibitor rapamycin. Finally, we find that the FSH-mediated up-regulation of reporter activities for LHR, inhibin-alpha, and vascular endothelial growth factor is dependent upon HIF-1 activity, because a dominant negative form of HIF-1alpha interferes with the up-regulation of these genes. These results show that FSH enhances HIF-1 activity downstream of the PI 3-kinase/AKT/Rheb/mTOR pathway in GCs and that HIF-1 activity is necessary for FSH to induce multiple follicular differentiation markers.

The anterior pituitary hormone follicle-stimulating hormone (FSH) 1 provides the key stimulus that promotes proliferation and differentiation of the ovarian follicle to the pre-ovulatory phenotype (1,2). FSH binding to the FSH receptor, a G proteincoupled receptor found in females exclusively on the surface of granulosa cells (GCs), leads to G s -stimulated activation of adenylyl cyclase and increased production of the second messenger, cAMP (3). FSH has been shown to activate multiple signaling pathways in GCs. We have demonstrated that FSH stimulates the activation of protein kinase A (PKA) and have identified the histones H1 and H3 as PKA targets (4,5). Additionally, we have recently shown that FSH signals via PKA to increase extracellular regulated kinase (ERK) signaling in response to FSH by turning off an inhibitory ERK phosphatase (6). Furthermore, we and others have shown that FSH activates AKT downstream of phosphatidylinositol 3 (PI 3)-kinase (6 -11).
FSH signaling results in increased transcription of a number of characteristic follicular differentiation markers, including the steroidogenic enzymes side-chain cleavage cytochrome P-450 and aromatase cytochrome P-450 (12), the luteinizing hormone receptor (LHR) (3), inhibin-␣ (13), the signaling intermediates microtubule-associated protein 2D (MAP2D) (14), and the PKA type II␤ regulatory subunit (RII␤) (15). FSH also stimulates antrum formation and angiogenesis in the peripheral follicle thecal cells mediated in part by the vascular endothelial growth factor (VEGF) (16,17). Our investigation into the mechanisms by which two of these FSH targets, MAP2D and RII␤, are up-regulated in response to FSH revealed that their up-regulation at the protein level is blocked by the PI 3-kinase inhibitors LY294002 and wortmannin. FSH-mediated PI 3-kinase activation is also necessary for the expression of the cartilage link protein (8) in GCs and for optimal transferrin secretion and lactate production by Sertoli cells (9). In support of a role for PI 3-kinase/AKT activity in GC differentiation, insulin-like growth factor-1 (IGF-1), which activates the PI 3-kinase/AKT pathway via the IGF-1 receptor (18), synergizes with FSH in GCs to increase the expression of LHR (19), inhibin-␣ (20), and side-chain cleavage cytochrome P-450 and is sufficient for the expression of 3␤-hydroxysteroid dehydrogenase/isomerase (21). The importance of AKT activation by PI 3-kinase is further demonstrated in recent studies showing that constitutively active AKT synergizes with FSH to upregulate follicular differentiation markers such as 3␤-hydroxysteroid dehydrogenase/isomerase, LHR, inhibin-␣, and aromatase cytochrome P-450, whereas the addition of a dominant negative AKT inhibits their up-regulation (22). These data indicate that signaling downstream of AKT is crucial for the up-regulation of GC differentiation markers.
In light of the evidence that the PI 3-kinase/AKT signaling pathway is necessary for the process of follicular maturation, we explored signaling events downstream of AKT to ascertain their contribution to the transcriptional up-regulation of follicular differentiation markers. Our results show that FSH-stimulation of PI 3-kinase/AKT activity culminates in the activation of signaling proteins that control mRNA translation, including mTOR, p70 S6 kinase (p70 S6k ), and the S6 ribosomal protein (Fig. 1). We also show that this mTOR-mediated pathway is required for FSH to induce MAP2D, RII␤, LHR, and inhibin-␣. Finally, we identify one target protein, hypoxia-inducible factor-1␣ (HIF-1␣), whose translation can be enhanced by growth factors or hormones (23)(24)(25). HIF-1 is a transcription factor of the basic-helix-loop-helix Per/ARNT/Sim family consisting of a regulated ␣ subunit and a constitutively expressed ␤ subunit (26). HIF-1␣ is rapidly degraded by ubiquitin-mediated proteosomal degradation under normal oxygen tension (27). Although hypoxia stabilizes the HIF-1␣ protein by preventing its degradation, certain growth factors have been shown to enhance HIF-1 activity under normoxic conditions by increasing HIF-1␣ translation (23)(24)(25). HIF-1 recognizes hypoxia response elements (HREs) containing the core ACGTG sequence in the promoters of target genes (28). HIF-1 targets include genes involved in angiogenesis such as VEGF and plasminogen activator inhibitor 1, genes that control cellular growth and metabolism such as glucose transporters and glycolytic enzymes, and genes that regulate proliferation such as insulin-like growth factor-2 (IGF-2) and p21 (29). Our results show that FSH enhances HIF-1 activity by the PI 3-kinase/AKT-dependent activation of mTOR and that HIF-1 activity is necessary for the up-regulation of such FSH target genes as VEGF, inhibin-␣, and the LHR.
Animals-Sprague-Dawley rats (Charles River Laboratories, Inc.) were housed at Northwestern University animal care facilities and maintained in accordance with guidelines for the care and use of laboratory animals by protocols approved by the Northwestern University Animal Care and Use Committee.
Pregnant Mare Serum Gonadotropin (PMSG) Treatment and Tissue Extract Preparation-Immature female rats (26 -27 days old) were injected subcutaneously with 25 IU of PMSG. Ovaries were harvested at the indicated times post PMSG injection and subjected to tissue extract preparations. Ovarian extracts were prepared as described previously (34). Protein concentrations were measured by the method of Lowry et al. (35) using crystalline bovine serum albumin as a standard. The samples were prepared for SDS-PAGE by suspension in SDS-containing sample buffer followed by heat denaturation (100°C, 5 min). Western blotting was performed as described below.
Granulosa Cell Culture and Western Blotting-GCs were isolated from ovaries of 26-day-old Sprague-Dawley rats primed with subcutaneous injections of 1.5 mg of estradiol-17␤ on days 23-25 to promote the growth of pre-antral follicles. Cells were plated on fibronectin-coated (BD Biosciences) 33-mm plastic dishes at a density of ϳ3 ϫ 10 6 cells/ dish in Dulbecco's modified Eagle's/F12 serum-free medium supplemented with 1 nM estradiol-17␤, 100 units/ml penicillin, and 100 g/ml streptomycin and treated with the indicated additions ϳ20 h after plating (5). Treatments were terminated by aspirating medium and rinsing cells once with phosphate-buffered saline. Total cell extracts were collected by scraping cells in SDS sample buffer (36) followed by heat denaturation. Protein concentrations were controlled by plating identical cell numbers per plate in each experiment and then loading equal volumes of total cell extract per gel lane. Equal protein loading was confirmed by AKT, mTOR, or CREB Western blots, as indicated. Granulosa cell proteins were separated by SDS-PAGE and transferred to Hybond C-extra nitrocellulose (Amersham Biosciences) (5). Blots were incubated with primary antibody overnight at 4°C, and antigenantibody complexes were detected by enhanced chemiluminescence (Amersham Biosciences). Western signals were quantitated with Molecular Analyst TM /PC Image Analysis software program, divided by the densitometric signal for control protein load, and expressed relative to the maximal signal.
Transfection and Luciferase Assays-GCs were plated in 12-well plates at 3 ϫ 10 5 cells/well in Dulbecco's modified Eagle's/F12 medium, 1 nM estradiol-17␤, 100 units/ml penicillin, and 100 g/ml streptomycin. Cells were washed with PBS and transfected with various promoter-luciferase constructs (0.5 g DNA/well) and the indicated expression constructs (0.05 g of DNA per well) using LipofectAMINE 2000 (Invitrogen) as Results support the schematic model in which FSH via cAMP stimulates the activation of PI 3-kinase/ AKT leading to the inactivation of tuberin and subsequent Rheb activation. Rheb then activates mTOR to stimulate translation by phosphorylating p70 S6k , which activates the S6 ribosomal protein. The HIF-1␣ protein and, thus, HIF-1 activity are consequently up-regulated via increased translation. An increase in HIF-1␣ leads to the induction of VEGF, inhibin-␣, and LHR in GCs, leading to follicular maturation. described previously (37). Briefly, cells were incubated with the transfection mixture at 37°C for 4 -5 h, after which Dulbecco's modified Eagle's/ F12 serum-free medium supplemented with 1 nM estradiol-17␤, 100 units/ml penicillin, and 100 g/ml streptomycin was added to the cells. Approximately 20 h after transfection, cells were treated as indicated. Cells were lysed and analyzed for luciferase activity using a luminometer as described previously (37). Luciferase activity for each measurement was normalized to protein content using 280-nm absorbance measurements (38). Data are presented as the mean Ϯ S.E. of triplicate samples. All transfection experiments were repeated three times with similar results unless otherwise indicated. Results were analyzed using Student's t test (p Ͻ 0.05 denoted by **) (39). The percentage of inhibition in the presence of various inhibitors was calculated as (1 Ϫ (fold induction FSH ϩ inhibitor /fold induction FSH )) ϫ 100.

FSH Stimulates Activation of AKT, Inactivation of Tuberin,
and Activation of p70 S6k and S6 Ribosomal Protein-AKT kinase activation requires the sequential phosphorylation of Thr-308 by phosphoinositide-dependent kinase 1 and Ser-473 by an unidentified kinase (40). It has been previously established that FSH stimulation of rat ovarian GCs leads to the activation of AKT phosphorylation (6,7). We confirmed that AKT is activated in GCs under our culture conditions in response to FSH using an antibody specific for AKT phosphorylated at Ser-473. As shown in Fig. 2A, AKT phosphorylation was readily detected at 10 min post FSH treatment of GCs, had further increased by 1 h, and had subsided to undetectable levels by 4 h post FSH treatment. Total AKT is shown as a protein loading control.
Tuberin, a component of the tuberous sclerosis complex, has been identified in various cellular models as a target of AKT kinase (41,42). In its active, unphosphorylated state, tuberin prevents signaling of the downstream kinase mammalian target of rapamycin, mTOR (43). AKT phosphorylation of tuberin at Thr-1462 inhibits its activity, thus preventing it from impeding mTOR signaling (44). We therefore determined whether FSH stimulated the phosphorylation of tuberin in GCs. Using a Thr(P)-1462-specific anti-tuberin antibody, our results demonstrate that FSH treatment of GCs promoted phosphorylation of tuberin by 10 min, which was further increased at 1 h post FSH treatment and waned moderately by 4 and 8 h, but not, however, to the extent observed with AKT phosphorylation ( Fig. 2A).
Because FSH promotes the phosphorylation of tuberin and thus the expected activation of mTOR, we explored whether previously identified targets of mTOR kinase activity, namely p70 S6k (45) and 4E-binding protein-1 (4EBP-1) (46,47) were phosphorylated in GCs in response to FSH. Phosphorylation of p70 S6k at Thr-389 downstream of mTOR activity is indicative of p70 S6k activation (48,49). When active, p70 S6k phosphorylates the S6 ribosomal protein. S6 is a component of the 40 S ribosome that promotes the translation of a subset of mRNAs containing a 5Ј-oligopyrimidine tract (50). Furthermore, inhibitory phosphorylation of 4EBP-1 at multiple sites downstream of mTOR, including Ser-65 (51), prevents 4EBP-1 from sequestering the eukaryotic translation initiation factor 4E, which promotes cap-dependent translation (52). Thus, phosphorylation of both p70 S6k and 4EBP-1 leads to the activation of the translational machinery.
Western blotting with an antibody recognizing phosphorylated p70 S6k (Thr(P)-389) demonstrated that FSH treatment of GCs resulted in p70 S6k phosphorylation that was maximal at 1 h and remained elevated 4 h post FSH treatment ( Fig. 2A). The phosphorylation of S6 by p70 S6k in GCs treated with FSH was monitored with a Ser(P)-235/Ser(P)-236-specific anti-S6 antibody. S6 phosphorylation was observed at 10 min, elevated by 1 h, and remained detectable at 4 h after FSH treatment (Fig 2A). Additionally, the inhibitory phosphorylation of 4EBP-1 at Ser-65 was detected by 1 h and increased by 2 h post FSH treatment (Fig. 2B). Taken together, these data demonstrate that FSH stimulation of GCs leads to the inactivating phosphorylation of tuberin, an inhibitor of mTOR, as well as to the phosphorylation of p70 S6k , S6, and 4EBP-1, previously characterized targets downstream of mTOR kinase activity. These results suggest that FSH stimulates the translational machinery in GCs by activating S6 ribosomal protein and eukaryotic translation initiation factor 4E.
FSHR-dependent Phosphorylation of AKT, Tuberin, p70 S6k , and S6 Occurs via cAMP-mediated Stimulation of PI 3-Kinase Activity-FSH signals to activate G s and, consequently, adenylyl cyclase, promoting the synthesis of cAMP (3). In the following experiments, we investigated whether FSH-dependent signaling to AKT, tuberin, p70 S6k , and S6 could be mimicked by the direct pharmacological adenylyl cyclase activator forskolin or the cell permeable cAMP analogue, 8-chlorophenylthio (8-CPT)-cAMP. The results demonstrated that the treatment of GCs for 1 h with forskolin or 8-CPT-cAMP stimulated the phosphorylation of AKT, tuberin, p70 S6k , and S6 (Fig. 3A, lanes  1-4). Therefore, cAMP is sufficient to bring about the phosphorylation of AKT, tuberin, p70 S6k , and S6. FSH-stimulated AKT activation in GCs has previously been demonstrated to require PI 3-kinase activity (7). We therefore used the PI 3-kinase inhibitor, wortmannin, to determine whether the phosphorylation of tuberin, p70 S6k , and S6, also required activation of PI 3-kinase. Wortmannin treatment of GCs prevented FSH from eliciting maximal phosphorylation of AKT, tuberin, p70 S6k , and S6 ( abrogated by wortmannin. These results thus show that in GCs, PI 3-kinase activity downstream of cAMP is required for the phosphorylation of AKT, tuberin, p70 S6k , and S6. Rheb Activation of mTOR Downstream of AKT Leads to p70 S6k and S6 Activation-In the following experiments, we used the mTOR inhibitor rapamycin, a macrolide that inhibits mTOR in a selective manner (53), to verify that phosphorylation of p70 S6k at Thr-389 and the subsequent activation of S6 occur downstream of mTOR activation in FSH-stimulated GCs. Treatment of GCs with rapamycin did not affect FSH-mediated phosphorylation of either AKT or tuberin but abrogated the phosphorylation of p70 S6k and its target S6 (Fig 3B, compare   lanes 1 and 2 versus lanes 3 and 4). These data place mTOR activation downstream of AKT phosphorylation/activation and tuberin phosphorylation/inactivation, but upstream of p70 S6k and S6 phosphorylation/activation.
Recent studies have identified the mechanism by which the AKT substrate tuberin regulates mTOR activity. The small G-protein Rheb has been identified as an activator of mTOR, while tuberin has been identified as a negative regulator of Rheb by acting as its GTP-ase activating protein (54 -56). In the following experiments, we sought to determine whether Rheb activity was necessary for mTOR signaling downstream of FSH in GCs. To ascertain the effects of inhibiting Rheb activity, we treated GCs with the farnesyltransferase inhibitor, FTI-277, because optimal Rheb function requires C-terminal farnesylation (56,57). Treating GCs with the Rheb inhibitor FTI-277 did not affect the FSH-mediated phosphorylation of AKT or tuberin; however, FTI-277 reduced phosphorylation of p70 S6k and S6 by 54 and 63%, respectively (Fig. 3C, compare  lanes 1 and 2 versus lanes 3 and 4). This places Rheb downstream of AKT phosphorylation/activation and tuberin phosphorylation/inactivation, but upstream of mTOR signaling to p70 S6k and S6.
The protein kinase inhibitor H89 is reported to inhibit p70 S6k with an IC 50 of 80 nM, and other kinases including PKA at higher concentrations (53). We therefore used H89 to confirm that the signaling pathway initiated in GCs by FSH and resulting in S6 phosphorylation requires p70 S6k activity. As shown in Fig. 2A, 10 M H89 inhibited S6 phosphorylation but did not block the phosphorylation of AKT, tuberin, or p70 S6k (Fig. 3A, compare lane 1-4 versus lanes 9 -12). The inability of H89 to inhibit AKT phosphorylation in FSH-stimulated GCs has been reported previously (7). These results confirm that the S6 protein is phosphorylated downstream of p70 S6k in GCs treated with FSH, forskolin, or cAMP.
Because IGF-1 signaling via the IGF receptor has been shown in other cellular models to activate the PI 3-kinase/AKT/ mTOR pathway and not PKA (58), we confirmed the site of H89 inhibition at p70 S6k in IGF-1-stimulated GCs. As seen in Fig.  2D, IGF-1 treatment of GCs for 1 h led to the phosphorylation of AKT, tuberin, p70 S6k , and S6. Consistent with the results seen with FSH-treated GCs, pretreatment of GCs with H89 did not affect IGF-stimulated phosphorylation of AKT, tuberin, or p70 S6k , but did abrogate S6 phosphorylation (Fig. 3D, compare  lanes 1 and 2 versus lanes 3 and 4). Thus, at 10 M, H89 is an effective inhibitor of p70 S6k activity in GCs.
Taken together, these data define a novel signaling pathway in GCs whereby AKT-stimulated inactivation of tuberin results in increased Rheb and, consequently, mTOR activities leading to p70 S6k and S6 activation.
Tuberin Inactivation and p70 S6k and S6 Activation Occur in Response to PMSG in Vivo-We also tested whether this path-way was activated in vivo in response to PMSG (an FSH source). Rats were injected subcutaneously with PMSG, and the phosphorylation of key intermediates was evaluated at indicated times post PMSG treatment by Western blotting of detergent-soluble ovarian extracts. Results demonstrated that AKT, tuberin, p70 S6k , and S6 were phosphorylated at 8 h post PMSG treatment and, with the exception of p70 S6k , remained phosphorylated at 48 h after PMSG treatment (Fig. 4). Thus the AKT/mTOR pathway leading to S6 phosphorylation is also activated in vivo in ovarian cells by gonadotropin stimulation. (15), MAP2D (14), LHR (3), and inhibin-␣ (13) are transcriptionally up-regulated in response to FSH and comprise markers of follicular differentiation. The dependence of LHR and inhibin-␣ up-regulation on AKT signaling has recently been demonstrated using a dominant negative AKT adenovirus (22). To test whether the activation of the Rheb/ mTOR pathway downstream of PI 3-kinase/AKT is involved in the up-regulation of these genes, we used either reporter as- says or Western blotting to assess the effect of various inhibitors on the FSH-mediated up-regulation of these genes. Both MAP2D and RII␤ induction is seen by Western blotting at 72 h post FSH treatment of GCs (Fig. 5, A and B, lanes 1 and 2). Pretreating GCs with the PI 3-kinase inhibitor wortmannin (Fig. 5A, compare lanes 1 and 2 versus 3 and 4) or the mTOR inhibitor rapamycin (Fig. 5B, compare lanes 1 and 2 versus 3  and 4) inhibited FSH-mediated up-regulation of both of these proteins. The effect of these inhibitors on the up-regulation of the LHR and inhibin-␣ was assessed using a LHR-luciferase reporter containing 2082 base pairs upstream of the transcription initiation site of the LHR gene, LHR-Luc, and an inhibin-␣ luciferase reporter consisting of 2021 base pairs upstream of the transcription initiation site of the inhibin-␣ gene (31), inhibin-␣-Luc. Treatment of GCs with FSH for 6 h stimulated LHR-Luc and inhibin-␣-Luc activity ϳ3and ϳ20-fold, respectively (Fig. 5, C-F). Pretreating GCs with the PI 3-kinase inhibitor LY294002 (Fig. 5, C and D) or the mTOR inhibitor rapamycin (Fig. 5, E and F) significantly inhibited the FSHmediated increase in the activity of both the LHR and inhibin-␣ reporters. These data establish that signaling downstream of mTOR is necessary for the transcriptional up-regulation of these FSH differentiation markers.

PI 3-Kinase/AKT/Rheb/mTOR Pathway Is Necessary for the Up-regulation of Genes Transcriptionally Up-regulated in Response to FSH-RII␤
HIF-1␣ Protein Levels and HIF-1 Activity are Up-regulated by FSH in GCs-HIF-1 is a heterodimeric transcription factor whose activity is up-regulated in other cellular models downstream of the PI 3-kinase/AKT/mTOR pathway (24, 25, 59 -62). HIF-1 consists of a typically regulated subunit, HIF-1␣, and an unregulated subunit, HIF-1␤ (63). In the following experiments, we examined whether FSH stimulation of GCs affected HIF-1␣ and HIF-1␤ protein levels. Because the half-life of HIF-1␣ is ϳ5 min under normoxic conditions (64), we used either the proteosome inhibitor MG115 (65) or CoCl 2 (66), both of which inhibit degradation of HIF-1␣, to investigate whether FSH stimulation of GCs promoted the accumulation of HIF-1␣ relative to untreated cells. GCs were treated with FSH for 4 h in the presence of either one of these inhibitors. In Fig. 6A, Western blotting of total cell extracts demonstrated that FSH induced the accumulation of an ϳ120-kDa protein reactive with an anti-HIF-1␣ antibody, relative to untreated cells. The levels of HIF-1␤ did not change relative to control levels in response to 2 or 4 h of FSH treatment (Fig. 6B) even in the presence of reagents that prevent HIF-1␣ degradation (data not shown).
In contrast to hypoxia, which stabilizes HIF-1␣, growth factor/hormone stimulation may increase HIF-1␣ translation (23)(24)(25). We therefore ascertained in GCs whether the FSH-stimulated increase in the HIF-1␣ protein, detected when HIF-1␣ degradation is blocked, is sensitive to transcriptional or translation inhibitors. Results show that the accumulation of HIF-1␣ induced by 4 h of FSH treatment of GCs was unaffected when the cells were pretreated for 1 h with the transcriptional inhibitor actinomycin D (Fig. 6C). However, when the cells were pretreated for 1 h with the translational inhibitor cycloheximide we were unable to detect any protein in either untreated or FSH-treated cells (Fig. 6C). These data suggest that the FSH-stimulated increase in the HIF-1␣ protein is largely dependent upon de novo translation and not transcription.
For the increase in HIF-1␣ protein in FSH-stimulated GCs to be functionally relevant, FSH should also increase HIF-1 activity. Therefore, in the following experiments, we tested whether HIF-1 activity was up-regulated in GCs in response to FSH treatment. To ascertain whether FSH stimulated HIF-1 transcriptional activity, GCs were transfected with a luciferase reporter containing three copies of the HRE from the phosphoglycerate kinase promoter (HRE (3)-TK-Luc). FSH stimulation of GCs for 6 h resulted in a ϳ12-fold increase in the activity of the HRE (3)-TK-Luc reporter relative to the untreated control cells (Fig. 7A). Accumulating evidence suggests that VEGF, which is an established HIF-1 target in other systems (33), is up-regulated in GCs in response to FSH (16,17). Using a VEGF-luciferase construct (VEGF-Luc) (33), we found that stimulation of GCs for 6 h resulted in a ϳ 6-fold stimulation of VEGF-Luc activity relative to untreated cells (Fig. 7B). These results demonstrate that FSH enhances HIF-1 activity in GCs.
Because the PI 3-kinase pathway has been implicated in the up-regulation of HIF-1 activity, we investigated whether HIF-1 activation in GCs is dependent on the PI 3-kinase/AKT/mTOR pathway. Addition of the PI 3-kinase inhibitor LY294002 significantly inhibited the FSH-dependent activation of both HRE (3)-TK-Luc and VEGF-Luc reporter activities by 72 and 85%, respectively (Fig. 7, A and B). These data are in accord with reports that growth factor or hormone stimulation of HIF-1 activity requires the activation of the PI 3-kinase pathway (24, 25, 59 -62).
Next, the Rheb inhibitor FTI-277 and the mTOR inhibitor rapamycin were used to determine whether Rheb and mTOR activities are necessary for FSH activation of HIF-1 transcriptional activity. Reporter assays using the HRE (3)-TK-Luc as well as the VEGF-Luc were performed in the presence and absence of the inhibitors. Treatment with FTI-277 inhibited FSH-mediated up-regulation of HRE (3)-TK-Luc and VEGF-Luc activities 45 and 48%, respectively (Fig. 8, A and B). Similarly, rapamycin inhibited the FSH-mediated up-regulation of HRE (3)-TK-Luc and VEGF-Luc activities by 44 and 60%, respectively (Fig. 8, C and D). These data demonstrate that both Rheb and mTOR activation downstream of PI 3-kinase/AKT are necessary to up-regulate HIF-1 activity.
Expression of a Dominant Negative HIF-1␣ Interferes with FSH-mediated Up-regulation of Follicular Differentiation Markers-Finally, we sought to determine whether HIF-1 transcriptional activity was necessary for FSH-stimulated induction of VEGF in GCs. A dominant negative construct of HIF-1␣ lacking the DNA binding domain, as well as the transactivation domain of HIF-1␣ (33), was transfected in GCs along with VEGF-Luc. Dominant negative HIF-1␣ inhibited the FSH-mediated up-regulation of VEGF by 54% (Fig. 9A). We then sought to determine whether inhibiting HIF-1 activity prevented the up-regulation of the follicular differentiation markers inhibin-␣ and LHR. Co-transfection of the HIF-1␣ dominant negative construct inhibited the FSH-mediated up-regulation of the inhibin-␣ and LHR reporter activities 55 and 69%, respectively (Fig. 9, B and C). These data demonstrate that FSH-stimulated HIF-1 activity is necessary for the up-regulation of VEGF, LHR, and inhibin-␣. DISCUSSION FSH-stimulated differentiation of GCs in ovarian follicles is obligatory for ovulation and the resumption of oocyte meiosis (1,2). The differentiation response consists, in part, of the induction of a number of proteins including the LHR (3), inhibin-␣ (13), MAP2D (14), and RII␤ (15), as well as the proteins necessary for steroidogenesis, gap junction, and antrum formation (67,68). In addition to the important role for PKA in chromatin remodeling (5) and CREB phosphorylation (69) as well as ERK activation (6), recent studies from a number of laboratories have implicated the PI 3-kinase/AKT pathway in FSH-stimulated GC differentiation (7,22).
In this report, we demonstrate that FSH promotes the PI 3-kinase-dependent activation of p70 S6k by mTOR. mTOR is emerging in many cellular models as a key regulator of protein synthesis, cell growth, and cell survival (70). mTOR integrates signals from growth factor receptors and G-protein coupled receptors to stimulate mRNA translation by activating p70 S6k and, consequently, activating the 40 S ribosomal protein S6 and inhibiting the translational repressor 4EBP-1 (71,72). mTOR is activated upon FSH treatment by a pathway that is mediated by cAMP and requires the activation of PI 3-kinase/ AKT, the inactivation of tuberin, and the activation of Rheb, as summarized in Fig. 1. Our results in GCs show that mTOR activation in response to FSH is indeed necessary for the induction of LHR, inhibin-␣, MAP2D, and RII␤, proteins that characterize GCs of the preovulatory phenotype. Taken together, these results indicate that in GCs, FSH-stimulated mTOR activity is obligatory for GC differentiation.
We show that FSH stimulation of the PI 3-kinase/AKT/ mTOR pathway results in the stimulation of HIF-1 activity. We also detect induction of HIF-1␣ protein in FSH-treated GCs when its degradation is inhibited, consistent with reports that HIF-1␣ is rapidly degraded under normoxic conditions (64). The binding partner for HIF-1␣, HIF-1␤, is constitutively expressed in GCs, and its levels are not affected by FSH. HIF-1␣ and HIF-1␤ dimerize to regulate the transcription of a number of genes (29). Our results show that FSH enhances HIF-1 activity, as detected by using a minimal HRE reporter (30) as well as a reporter for VEGF, a well known HIF-1 target (33). To our knowledge, these data are the first report that HIF-1 activity is up-regulated by FSH in GCs. To further assess the role of HIF-1 transcriptional activity in the FSH-mediated induction of follicular differentiation markers, we used a HIF-1␣ dominant negative that has been shown previously to inhibit, in a dose dependent manner, the HIF-1-mediated activity of reporters containing the VEGF (33) and erythropoietin (73) promoter sequences. The dominant negative HIF-1␣ blocked the FSH-dependent induction of VEGF, inhibin-␣, and LHR in GCs, indicating that HIF-1 activity is critical for the up-regu- lation of these genes. The importance of HIF-1 activity in the follicular maturation process is further supported by the observation that the conditional knockout of HIF-1␤ in mice results in a sub-fertile phenotype (74).
Although our results demonstrate that the expression of a dominant negative HIF-1␣ reduces the FSH-mediated increase of LHR and inhibin-␣ reporter activities, in this report we do not address the mechanism(s) of HIF-1 action in GCs. There are several possible mechanisms by which HIF-1␣ may promote transcription of LHR and inhibin-␣. For example, HIF-1 could directly bind to HREs in LHR or inhibin-␣ promoters to increase transcription of these genes. Searching the promoter sequences of both LHR and inhibin-␣ revealed the presence of putative HREs in each of their promoters (75). Functional analyses of these sites are currently underway in our laboratory. HIF-1 may also be an important component of a multiprotein complex necessary for the transcriptional regulation of LHR and inhibin-␣. Indeed, direct interactions of steroidogenic factor-1 (SF-1), CREB, and associated coactivators such as the CREB-binding protein induce transactivation of the inhibin-␣ gene (31). In Leydig cells, Sp1 and Sp3 have been shown to transactivate the LHR gene (76). In various cell types, HIF-1 has been shown to interact with other transcription factors such as Sp1 (77) and Smad (78) and to recruit the CREBbinding protein as part of a multi-protein complex to transactivate target genes (79). Studies are underway to elucidate possible interactions between HIF-1 and transcription factors known to regulate FSH-responsive genes in GCs. Alternatively, the role of HIF-1 in LHR and inhibin-␣ induction may be to promote the expression of transcriptional activators that, in turn, act at specific sites on the LHR and inhibin-␣ promoters.
In addition to these possible mechanisms by which HIF-1 could regulate the expression of inhibin-␣ and LHR, it is also well known that inhibin-␣ and LHR are not coincidentally regulated by FSH in GCs. An association between phosphorylated histone H3 and the inhibin-␣ promoter is detected by 1 h post FSH (5), and increased inhibin-␣ mRNA is detected by 10 h post FSH (80). Although we do not know whether FSH stimulates a similarly rapid association of phospho-histone H3 and the LHR promoter, increased LHR mRNA is not detected until 36 h post FSH (81). These results indicate, as discussed above, that distinct factors must be involved in producing the discrete chronological expression pattern observed for these two genes. The relatively rapid induction of inhibin-␣ most likely can be attributed to the rapid regulation of both CREB and SF-1 activities by FSH-stimulated phosphorylation events (5,82), followed by their interactions with the CREB-binding protein and, in some manner, with HIF-1. However, the mechanism by which FSH regulates the activities of Sp1 and Sp3 and perhaps those of newly synthesized transcriptional activators, in association with or in response to HIF-1 and leading to the delayed regulation of LHR expression, is not understood. Recruitment of various cofactors to HIF-1 may also contribute to the different timing of inhibin-␣ versus LHR expression (83).
Although the expression of RII␤ and MAP2D requires mTOR activity, we do not yet know if they are regulated by HIF-1 in GCs. The RII␤ promoter contains a putative HRE (84), but the MAP2D promoter region has not yet been identified. Studies are underway to determine whether HIF-1 regulates expression of these two FSH targets.
HIF-1 activity is also necessary for the up-regulation of VEGF reporter activity. VEGF, a recently identified FSH target (17), has two well characterized HREs in its promoter (33), and it is possible that the up-regulation of VEGF by FSH occurs through HIF-1 interaction directly with these elements. Recent reports emphasize the importance of VEGF to follicle develop-ment, as substances that block VEGF signaling prevent PMSGstimulated follicular development by reducing angiogenesis in the theca cell layer, antrum formation within the follicle, granulosa cell proliferation, and estrogen production (16).
HIF-1 has been identified as a transcriptional regulator of many genes in addition to VEGF, including Glut-1, transferrin, and glycolytic enzymes (85). Because HIF-1 activity is upregulated in GCs, a subset of these genes may also be upregulated in response to FSH. Consistent with this hypothesis, FSH up-regulates Glut1 (86) in GCs and transferrin secretion and glycolysis in Sertoli cells in males (9). HIF-1 transactivation may facilitate follicle growth and cell survival by promoting glucose metabolism. However, it has yet to be determined whether the up-regulation of these genes by FSH in GCs is dependent upon HIF-1 activity or on an alternate pathway.
It is likely that expression of each of the FSH-responsive genes in GCs is regulated by a unique combination of transcription factors and coactivators. Our results show that, like SF-1 and CREB, HIF-1 also serves as an important transcription factor that is necessary to regulate at least a subset of FSH-responsive genes. Investigation of protein-protein interactions of HIF-1 with SF-1, CREB, and other transcription factors is needed to better understand the regulation of FSH-responsive genes.
In summary, in this report we have demonstrated that mTOR is a critical target of the FSH-stimulated PI 3-kinase/ AKT pathway in GCs. mTOR activity is obligatory for FSHstimulated up-regulation of LHR, inhibin-␣, MAP2D, VEGF, and RII␤, all of which characterize follicular differentiation to a preovulatory phenotype. FSH stimulation of mTOR is also necessary for the induction of HIF-1 activity in GCs, and HIF-1 activation is, in turn, necessary for the FSH-dependent upregulation of inhibin-␣, LHR, and VEGF in GCs. These results categorize mTOR as a key regulator of GC growth and differentiation and identify HIF-1␣ as a key translational product downstream of FSH. Our results further suggest that FSH induces an early translational response in addition to immediate-early transcriptional responses to bring about the up-regulation of targets genes that characterize the preovulatory phenotype. Future studies will focus on the mechanism by which FSH-stimulated HIF-1 activity contributes to follicular differentiation.