| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas 77030
Address all correspondence and requests for reprints to: JoAnne S. Richards, Ph.D., Department of Molecular and Cellular Biology, Baylor College of Medicine, One Baylor Plaza, Houston, Texas 77030. E-mail: joanner{at}bcm.tmc.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
IGF-I via its cognate receptor and the insulin receptor substrates (IRS-1/2) impacts multiple signaling cascades. One cascade that is highly conserved from Caenorhabditis elegans to mammals includes PI3K, phosphoinositide-induced kinases (PDK1/2) and protein kinase B (PKB)/Akt (13, 14, 15). PI3K generates phosphoinositides that activate PDK1/2. PDK1/2 as well as other kinases phosphorylate and activate PKB in a hierarchial and complex manner (16, 17). In contrast, many substrates that are phosphorylated by PKB are inactivated and/or degraded, such as GSK-3ß, BAD, and caspase 9 as well as members of the Forkhead homolog in rhabdomysarcoma (FKHR) winged-helix transcription factor family (14). Three members of the Forkhead family have been identified in the mouse: FKHR (Foxo1), FKHRL1 (Foxo3), and AFX (Foxo4) (18, 19). A current model of IGF-I action indicates that phosphorylation of Forkhead proteins by PKB (and related kinases) restricts the nuclear localization of these factors thereby providing a mechanism to regulate the transcriptional activation of Forkhead target genes (16, 20). Three genes thought to be regulated by FKHR, FKHRL1, and AFX are Fas ligand (FasL), an inducer of apoptosis (21), p27KIP, an inhibitor of cell cycle progression (22) and IGFBP-1, a presumed inhibitor of IGF-I (4, 5, 23). Each of these is hormonally regulated in the ovary; FasL being expressed in follicles (24, 25) and IGFBP-1 and p27KIP being increased in corpora lutea (CL) (4, 26). Lack of IGF-I in mice causes severe growth retardation, including follicular growth in the ovary (1, 2, 27, 28). Mice lacking IRS-2 also exhibit impaired fertility with small anovulatory follicles (3). In contrast, although mice null for GH exhibit reduced growth rates and have low serum levels of IGF-I, ovarian functions are not dramatically altered (29). This is likely a consequence of some GH-independent production of IGF-I by granulosa cells within the ovary (30) (2).
New aspects to FSH-mediated cell signaling cascades have also been documented recently. In the classical pathway, FSH binds its cognate seven-pass membrane receptor to activate adenylyl cyclase thereby increasing intracellular cAMP that activates cAMP-dependent PKA. Downstream targets of PKA phosphorylation include cAMP response element binding protein (CREB), CREB binding protein, and CREB-regulated genes, such as aromatase (12). FSH and cAMP also impact the PI3K pathway in granulosa cells by inducing serum and glucocorticoid-induced kinase (Sgk) (31, 32). Sgk is a PKB-related kinase that like PKB is phosphorylated and activated by the PI3K/PDK1 pathway in response to IGF-I (33, 34, 35) as well as FSH (12, 36). Expression of Sgk has been related to cell proliferation (37, 38) and cell survival pathways in mammalian cells (39) and Yeast (40). The identification of cAMP-regulated guanine nucleotide exchange factors, cAMP-GEFs (41, 42), provides a potential new link between FSH stimulation of adenylyl cyclase and activation of PI3K via Ras-related small GTPases. Functional links between the FSH and IGF-I signal pathways are supported by the observations that IGF-I, IGF-I receptor and FSH receptor colocalize to granulosa cells of small growing follicles and preovulatory follicles (2, 30). In mice null for the FSH receptor or the FSHß subunit follicular growth is impaired beyond the preantral stage (43, 44, 45) and expression of many genes including Sgk is impaired (46).
The steroid hormone E2 acts via two ER subtypes, the classical ER
and the more recently discovered ERß (47), to enhance granulosa cell proliferation (26, 48) and the fate of granulosa cell differentiation (6, 49). The mechanisms by which estrogen alters the response of granulosa cells to FSH has remained elusive but could involve the regulation of components of the IGF-I signaling cascade, including Forkhead transcription factors. Recent studies have shown by two-hybrid screening that FKHRL1 interacts with ER and can act as a bifunctional repressor or activator of nuclear hormone receptor activity (50, 51). Therefore, it is possible that FSH as well as IGF-I regulate the expression and function of ER subtypes in the ovary.
Based on these observations, the following studies were undertaken to determine which of the Forkhead molecules is expressed in the rodent ovary, is hormonally regulated in a cell or stage specific manner during follicular development and is a target of E2, FSH, and/or IGF-I action. For these studies, we have used hormonally primed immature mouse and rat models in which follicles and CL can be analyzed at specific stages of development and in which granulosa cells can be isolated and cultured in the presence of selected hormones. RT-PCR and in situ hybridization were used to provide semiquantitative analyses for the temporal expression and cellular localization patterns, respectively, for FKHR, FKHRL1, and AFX. Hormone-regulated expression of IGF-I, Sgk, and ERß were also analyzed to determine which if any might impact expression of Forkhead molecules. Western blots were used to analyze the relative amounts and temporal changes in levels of FKHR, phospho-FKHR and FKHRL1, PKB, phospho-PKB, Sgk, IGF-I receptor ß (IGF-1Rß), and Glut-1 in granulosa cells of growing follicles. Immunocytochemical studies were done to determine the subcellular localization of FKHR and PKB in granulosa cells under defined conditions.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
, lane 1). PMSG (48 h) stimulated 2- to 2.5-fold increases in FKHR, FKHRL1, and AFX mRNA (lane 2) above that seen in immature mouse ovary (lane 1) (P < 0.01, 0.01 and 0.05, respectively). However, the response to hCG differed among the three genes. hCG caused FKHR mRNA to decrease within 4 h (lane 4; P < 0.01), after which it remained low in ovaries of PMSG-hCG treated mice collected 8–48 h after hCG (lanes 5–9). FKHR transcripts were also present but low in ovaries of pregnant mice (data not shown). In contrast, FKHRL1 mRNA did not change markedly in response to hCG, whereas AFX mRNA was elevated in luteinized ovaries of mice treated with PMSG-hCG 48 h (lane 9) and on d 15 of pregnancy (data not shown) compared with untreated controls (P < 0.02).
|
). In contrast, FKHR mRNA rapidly declined in ovulating follicles, 2–12 h after hCG and was negligible in CL 48 h after hCG (Fig. 1B). FKHR mRNA was also not observed by in situ hybridization in CL of pregnant mice or adult cycling mice (data not shown). Note that FKHR remained expressed abundantly in granulosa cells of growing follicles of PMSG-hCG treated mice, thereby explaining the continued amplification of FKHR in the RT-PCRs of whole ovarian RNA prepared after hCG stimulation (Fig. 1A, FKHR lanes 4–9). In contrast, FKHRL1 and AFX mRNAs were detected in theca cells as well as granulosa cells of some follicles with distinct expression observed in CL (Fig. 1B).
Closer examination of immature mouse ovaries (d 5–21 of age) revealed that FKHR transcripts were readily detectable in follicles that exhibited early signs of atresia, i.e. a distinct beading appearance of the outer layer of granulosa cells and the presence of immune cells in the antral fluid (data not shown). More unexpected and striking was the intense FKHR signal in oocytes of primary follicles (comprised of one or two layers of granulosa cells) in the d 15 and 21 mouse ovaries (Fig. 2). In contrast, nests of oocytes at early stages of follicle formation (d 5 of age) or mature oocytes surrounded by cumulus cells within larger follicles at the proestrous stage of follicular growth did not appear to express detectable levels of FKHR mRNA by this method (Fig. 2). As expected from the results in Fig. 1, granulosa cells and cumulus cells of proestrous follicles were FKHR positive (Fig. 2). Thus FKHR appears to be expressed at specific stages of oocyte development.
|
, upper left). IGF-I mRNA was present in immature ovaries, increased approximately 2-fold with PMSG (P < 0.05) and then decreased (P < 0.05) between 4 and 48 h after hCG as ovulating follicles become CL, confirming IGF-I expression in granulosa cells but not CL (30). Sgk, a kinase that can phosphorylate FKHR, FKHRL1, and AFX (18) and is itself a target of IGF-I activation as well as FSH and LH induction and activation in granulosa cells (33, 34, 36), increased slightly (P < .05) in response to PMSG (Fig. 3, lower left). The most dramatic increases in Sgk mRNA were observed from 2–48 h after hCG (lanes 3–9), when the levels became approximately 8-fold higher than the those observed in the immature mouse ovary (lane 1; P < 0.001). Sgk mRNA was also elevated in ovaries of pregnant mice (data not shown). In situ hybridization showed that Sgk mRNA is present in granulosa cells of growing follicles but is elevated in CL of PMSG-hCG treated mice (Fig. 3), confirming previous observations in the rat (32).
|
RT-PCR analyses show that isolated granulosa cells of H rats express FKHR transcripts (Fig. 4A). E2 induced a >2-fold increase (P < 0.02) in FKHR mRNA (Fig. 4A; HE) in association with the growth of large preantral follicles (32, 53). Subsequent stimulation of HE rats with low doses of FSH for 48 h [Fig. 4A; hypophysectomized, E2 and FSH treated (HEF)] initiated the growth of preovulatory follicles and a transitional stage of differentiation. In these cells, FKHR remains expressed. Injections of hCG that terminate follicular growth and stimulate their differentiation to nondividing luteal cells caused the most dramatic (90%) decrease in FKHR mRNA (Fig. 4A; HEF, hCG; P < 0.001 compared with HE). Levels of IGF-I mRNA exhibited a similar pattern of expression (data not shown) confirming the continued expression of this growth factor in granulosa cells of the H rat ovary and its down-regulation with luteinization (2, 30).
|
). In contrast, intense FKHR labeling of granulosa cells was observed in large E2-stimulated preantral follicles (Fig. 4B; HE). In response to an iv injection of a single high (ovulatory-like) dose of FSH (10 µg), FKHR mRNA decreased rapidly by 2 h and remained low at 8 h in the large preantral follicles (Fig. 4C). However, at 8 h FKHR mRNA was detected in granulosa cells of small preantral follicles. By 24 h, intense staining of FKHR was again observed in many large follicles. In contrast, when HE rats were injected sc with a low dose of FSH for 2 d (HEF, 48 h) to stimulate growth of preovulatory follicles, FKHR remained expressed in the granulosa cells of these follicles (Fig. 4C; HEF, 48 h). FKHR mRNA was not detected in CL of HEF-hCG treated rats or in pregnant rats on d 7 or 22 of gestation. However, FKHR mRNA was detected as a strong signal in follicles present in the pregnant rat ovaries. Note that FKHR message was expressed in granulosa cells and cumulus cells but not the oocyte of the preovulatory follicle present in the d 22 pregnant rat ovary (Fig. 4C). As in the mouse ovary, FKHRL1 and AFX were also detected in CL of pregnant rats (data not shown).
FKHR Protein Is Induced and Phosphorylated by Hormones in Vivo
Because genetic analyses in C. elegans (15) and biochemical analyses in mammalian cells indicate that FKHR is a substrate for both PKB and Sgk (18, 20, 21), we sought to determine if the amount or phosphorylation of FKHR protein was altered by hormones in granulosa cells in vivo. For these studies, protein samples from hormonally primed H rats were analyzed by Western blots using specific antibodies against FKHR, phospho-Thr24-FKHR, phospho-Ser 256-FKHR as well as phospho-PKB. Levels of IGF-1Rß that is an upstream component of the IGF-I pathway and to Glut-1 that is a downstream target of IGF-I were also analyzed (54).
Total FKHR Protein
Granulosa cells of H rats expressed low levels of FKHR protein that increased markedly (>7.5-fold) after in vivo exposure to E2 (Fig. 5; HE), confirming results of RNA analyses (Fig. 4A). The multiple immunopositive bands correspond to intact FKHR (uppermost band) as well as putative proteolytic fragments of the protein (lower MW bands). Stimulation of HE rats with a high dose of FSH caused the amount of immunoreactive FKHR protein in granulosa cells to decrease 50% between 2–8 h. Low doses of FSH given for 48 h (HEF) to stimulate growth of preovulatory follicles also reduced FKHR protein to approximately 50%. A subsequent iv injection of hCG caused the most dramatic decrease in FKHR protein (greater than 90%) between 4 and 24 h when only small amounts of FKHR protein were detected (Fig. 5). Collectively, these results confirm the up-regulation of FKHR mRNA in E2-stimulated follicles and the loss of FKHR transcripts as granulosa cells cease dividing and differentiate to luteal cells (Fig. 4, A–C).
|
). The relative increase in FKHR Thr-24 phosphorylation at 48 h after FSH was 4.9-fold greater than the relative increase in FKHR protein. Thus, low doses of FSH given to HE rats stimulate a selective increase in phospho-Thr-24 FKHR in granulosa cells. hCG further increased to 10-fold the amount of phospho-Thr-24 FKHR relative to total FKHR at 2–8 h suggesting that activation of a specific kinase may be occurring in granulosa cells at this stage of differentiation. Because phosphorylation of Thr-24 by an unknown T24 kinase appears to be essential to restrict nuclear localization, at least in some cells (16), FSH and LH at high doses or in association with differentiation appear to enhance conditions favoring cytoplasmic, localization of FKHR. However, the exact ratio of nonphospho-(nuclear) to phospho-(cytoplasmic) FKHR in granulosa cells at any given time needs to be resolved to determine its functional activity as a transcription factor in vivo.
The phosphorylation pattern of FKHR Ser-256 was more complex. It was not related to the amount of FKHR protein, occurred transiently in response to FSH or hCG and decreased before that of phospho-Thr-24. Specifically, immunoreactive phospho-Ser-256 FKHR was detected in granulosa cells of H rat but increased only slightly in response to E2 (Fig. 5). Thus, the relative amount of phospho-Ser-256 FKHR to total FKHR decreased 80% in granulosa cells of HE rats. The functional significance of this is not yet clear. A high dose of FSH for 2 h stimulated a transient 2-fold increase in phospho-Ser-256 FKHR (relative to total FKHR protein). By 8 h, the amount phospho-Ser-256 FKHR decreased markedly to less than 90% of that in granulosa cells of HE rats. Although serine phosphorylation of FKHR was increased slightly by low doses of FSH for 48 h and by hCG at 2 h, no phospho-Ser-256 FKHR was detected 8–24 h after hCG as granulosa cells luteinize. In fact, immunoreactive phospho-Ser 256-FKHR declined by 90% before the loss of phospho-Thr-24 FKHR and FKHR protein (Fig. 5).
Phosphorylation of PKB
Although FKHR can be a substrate for PKB, the pattern of phosphorylated, activated PKB did not relate directly to that of total or phosphorylated FKHR (Fig. 5). Specifically, phospho-Ser-473 PKB and phospho-Thr-308 PKB were detected in granulosa cells of H and HE rats. Whereas the amount of phospho-Thre-308 was relatively constant, FSH stimulated an increase in phospho-Ser-473 PKB at 2 h, but this effect was transient and levels were reduced at 8 h. In contrast, the hCG stimulated phosphorylation of PKB at Ser-473 was sustained as the granulosa cells luteinize, a pattern previously seen in cultured granulosa cells (36). Because phospho-PKB was present in granulosa cells of H rats, it is clear that factors in addition to the gonadotropins stimulate PKB phosphorylation in vivo. One of these is likely to be endogenous IGF-I.
IGF-1Rß and Glut-1
The expression patterns of IGF-1Rß and Glut-1 protein were similar but less dramatic than that of FKHR (Fig. 5, upper panels). Both proteins were low in granulosa cells of H rats but increased in response to E2 (HE). IGF-1Rß further increased 2 h after acute administration of FSH and hCG but decreased approximately 50% in luteinized ovaries of HEF, hCG-treated rats. In contrast, Glut-1 was still easily detected in ovaries of HEF, hCG-treated rats. Thus, E2 coordinately up-regulates three components of the IGF-I pathway in growing follicles, namely, FKHR (7.5-fold), Glut-1 (
3-fold), and IGF-1Rß (1.5-fold), whereas hCG down-regulates these same components in association with luteinization.
FSH and IGF-I Differentially Regulate Expression of FKHR, IGF-I, Sgk, and ERß in Cultured Granulosa Cells
To determine more directly if the expression of either FKHR and/or IGF-I mRNA was regulated by FSH, a granulosa cell culture system was used (55, 56). Granulosa cells were isolated from E2-primed immature rats and cultured overnight on serum-coated plates in defined medium without serum. At that time, FSH (50 ng/ml) and T (10 ng/ml) were added to the cells for 2 h to examine the acute affects or 48 h to analyze changes associated with granulosa cell differentiation (i.e. cells that express aromatase, LH receptor, and other genes indicative of a preovulatory phenotype). The PKA inhibitor H89 or the PI3K inhibitor LY294002 was added 4 h before the addition of hormone (t=0) or 4 h before the preparation of the extracts at 48 h. RNA was extracted and expression of FKHR and IGF-I mRNA was analyzed by RT-RCR using specific primers as described in Materials and Methods.
FKHR mRNA was elevated in granulosa cells cultured in defined medium alone (Fig. 6, A and C, control), indicating that FKHR expression is not acutely affected by overnight culture in the absence of hormone. FSH/T caused FKHR mRNA to decrease dramatically within 2 h (Fig. 6, solid bars, lane 2; P < 0.01), a response similar to that of HE granulosa cells exposed to high FSH in vivo (Figs. 4 and 5). This rapid FSH-mediated decrease in FKHR was blocked by the addition of the PKA inhibitor H89 or the PI3K inhibitor LY294002 (lanes 3–5) indicating that multiple signaling pathways regulate FKHR expression. After 48 h of culture in the presence of FSH/T, FKHR mRNA was low (25% of control at 0 h; P < 0.02) (Fig. 6; hatched bars, lanes 6 vs. 1, respectively) and was not increased by a 4 h exposure to H89 or LY294002 (hatched bars; lanes 7–10). FSH/T also reduced the expression of IGF-I mRNA at 2 h and 48 h (Fig. 6B; P < 0.01). H89 and LY294002 blocked the effect at 2 h but were less effective at 48 h. Although not shown, the effects of forskolin (10 µM) were similar to those of FSH. Thus FSH/cAMP can acutely repress expression of FKHR and IGF-I by mechanisms that are PKA and PI3K dependent, whereas the loss of FKHR and IGF-I mRNA associated with granulosa cells differentiation appears to involve additional mechanisms.
|
; 2 h, dark bars), FKHR mRNA was elevated in cells cultured in medium alone or with LY294002 alone (lanes 1 and 6, respectively) but was reduced dramatically (P < .01) in cells cultured with FSH/T, IGF-I or the combination of FSH/T/IGF-I for 2 h (lanes 2–5). At 48 h (Fig. 7; open bars), FKHR mRNA was elevated in cells cultured in medium alone (lane 7) or with IGF-I alone (lane 9, solid arrow) but was reduced (P < .001) in cells cultured with FSH/T (lane 8, open arrow). The addition of IGF-I or LY294002 in the presence of FSH/T did not prevent the marked down-regulation of FKHR mRNA at 48 h (lanes 10–12) indicating the actions of FSH/T and IGF-I are distinct at this time.
|
, lane 8, open arrow; P < 0.001) supporting observations that IGF-I is not expressed in luteinized cells (2). IGF-I alone did not alter the expression of IGF-I mRNA (lane 9, solid arrow). In marked contrast, FSH/T increased (P < .01) the expression of Sgk at 2 h (Fig. 7; closed bars, lanes 2, 4) and 48 h (Fig. 7; open bars; lanes 8 and 10; open arrows). The addition of IGF-I alone did not induce expression of Sgk in these granulosa cells (lanes 3 and 9) whereas the addition of LY294002 blocked the actions of FSH/T (lanes 5, 6, 11, 12). Thus, the down-regulation by FSH/T of FKHR and IGF-I mRNAs at 48 h is inversely related to the expression of Sgk and occurs in a temporal manner similar to that observed when granulosa cells luteinize in response to hormones in vivo (Figs. 1–4). Furthermore, the up-regulation of Sgk by FSH/T at 48 h cannot be mimicked by IGF-I alone but appears to involve, at least in part, a PI3K-dependent step (Fig. 7).
To determine if IGF-I alone could regulate transcription factors that impact either the estrogen or FSH pathways, these same samples were analyzed for expression of ERß and ER (57), each of which may mediate specific granulosa functions such as the expression of FKHR, IGF-1Rß, and Glut-1 (Figs. 1, 4, 5). At 2 h, neither IGF-I nor FSH altered ERß mRNA (Fig. 7; closed bars) or ER mRNA (not shown; Ref. 57). IGF-I increased and/or maintained expression of ERß mRNA but had little effect on ER mRNA after 48 h in culture (Fig. 7, open bars; lane 9, solid arrow). In addition, the PI3K inhibitor, LY294002 blocked the effects of IGF-I, suggesting that a PI3K pathway may support expression of ERß at this time. In contrast, control cells cultured in medium alone or with FSH/T for 48 h expressed low levels ERß mRNA (lanes 7 and 8, open arrow) and FSH impaired the effects of IGF-I on ERß (lanes 10 and 11). Thus, FSH, E2 and IGF-I regulate transcription FKHR, IGF-I, Sgk and ERß by mechanisms that appear to be distinct, especially in differentiated granulosa cells.
FKHR and PKB Are Rapidly Phosphorylated in Response to FSH/T and IGF-I in Cultured Rat Granulosa Cells: Nuclear vs. Cytoplasmic Localization
To determine if the different responses of granulosa cells to FSH/T and IGF-I were related to changes in the phosphorylation of FKHR, we next determined if FSH and/or IGF-I stimulated the phosphorylation of FKHR and if FKHR phosphorylation altered its intracellular localization in cultured granulosa cells. Accordingly, granulosa cells were isolated from E2-primed immature rats and cultured overnight on serum-coated plates in defined medium without serum. At that time, decreasing doses of either FSH (50–5 ng/ml in the presence of 10 ng/ml T) or IGF-I (30–5 ng/ml) were added to the cells for different intervals of time (0–90 min, 2, or 48 h). Whole cell extracts (WCE) were prepared and the samples were analyzed by Western blots using antibodies to total FKHR, phospho-specific FKHR or phospho-PKB antibodies as indicated in Materials and Methods. To analyze the subcellular localization of FKHR and PKB, some cells were plated on coverslips, fixed in 4% paraformaldehye, and analyzed by immunocytochemical procedures using the same antibodies as in the Western blots.
Western blots show that the total amount of FKHR protein remained relatively constant in granulosa cells exposed for 0–90 min to FSH/T) but was reduced 50% in the presence of 30 ng/ml IGF-I (Fig. 8A; lanes 1–7 vs. lanes 8–13, respectively). FSH/T initiated rapid phosphorylation of FKHR on Thr-24 and Ser-256 as well as FKHRL1 on Ser-315 (Fig. 8A). Increased levels of phospho-FKHR-Thr-24 and FKHRL1-Ser-315 were sustained for as long as 90 min. In contrast, the amount of FKHR phospho-Ser-256 was elevated 6- to 4-fold between 5–30 min (lanes 2–4) but was nondetectable by 90 min (lane 5) even at lower concentrations (25 and 5 ng/ml) of FSH (lanes 6 and 7). IGF-I (30 ng/ml) also stimulated rapid and sustained increases FKHR phospho-Thr-24 and FKHRL1 phospho-Ser-315. Phosphorylation of Ser-256 FKHR also increased rapidly in response to IGF-I but then declined by 90 min (Fig. 8A, lanes 8–11) although lower concentrations of IGF-I supported phosphorylation at 90 min (lanes 12 and 13). In these same samples (Fig. 8A), FSH/T and IGF-I also stimulated the phosphorylation of PKB. However, despite the remarkably similar phosphorylation patterns of FKHR (and FKHRL1) induced by FSH/T and IGF-I, 30 ng/ml of IGF-I was two to three times more potent than 50 ng/ml of FSH/T in phosphorylating PKB Ser-473. Moreover, the phosphorylation of PKB persisted, whereas that of FKHR (phospho-Ser-356) was transient.
|
; Control, 0 h). In response to FSH/T or IGF-I FKHR protein remained nuclear for 1 min but rapidly localized to a distinct region of the cytoplasm within 5–30 min, as shown more clearly in the enlarged insert (Fig. 8B). This agonist-induced redistribution of FKHR was blocked by the addition of the PI3K inhibitor LY294002 for 1 h before hormone treatment, indicating that some component of the PI3K pathways regulates retention of FKHR in the cytoplasm. The cytoplasmic localization of immunoreactive FKHR was identical to the immunostaining pattern observed for phosph-Thr24 and phospho-Ser256 FKHR, indicating that restriction of FKHR from the nucleus is associated with its phosphorylation (Fig. 8; compare A and B; FSH/T, 30 min). In contrast to FKHR, immunostaining of nonphosphorylated PKB was localized around the nucleus of granulosa cells in the control cultures (Fig. 8, A and C; control 0 h). In hormone-treated cells, PhosphoSer-473 PKB localized as an intense dot adjacent to the nucleus. This pattern was highly similar in the cells stimulated by either FSH or IGF-I for 5–30 min (the latter is not shown) and appears similar to the pattern of phospho-FKHR. Immunostaining of PKB or phospho-PKB was not observed distinctly at the plasma membrane or in the nucleus at these time intervals (Fig. 8C).
Because FKHR is also a substrate for Sgk, we next determined if the hormone induced pattern of Sgk expression was related temporally to that of changes in FKHR phophorylation. Accordingly, granulosa cells were cultured for 48 h in medium alone with or without FSH/T, IGF-I or the combination. After 48 h of culture in medium alone, immunoreactive FKHR was easily detected in granulosa cell extracts, supporting the observations that FKHR expression (mRNA) is also not acutely dependent of hormonal regulation in these cells (Figs. 7 and 9). However, when the cells were cultured in the presence of FSH/T or FSH/T/IGF-I for 48 h, the amount of FKHR protein decreased markedly to a level 40% of that in the control sample, confirming the down-regulation of FKHR message (Figs. 6 and 7). At the same time, the levels of phospho-Ser-256 FKHR increased markedly in response to FSH/T (4-fold relative to total FKHR) or FSH/T combined with IGF-I (7-fold relative to FKHR). Exposing these cells to LY294002 but not H89 for 4 h reduced the amount of phospho-Ser-256 FKHR 80%. The FSH/T- and FSH/T/IGF-I-stimulated increases in phospho-Ser-256 FKHR at 48 h were associated with increased amounts and phosphorylation (multiple bands) of Sgk (Fig. 9, upper panel), supporting previous studies (36). FSH/T and FSH/T/IGF-I also stimulated increased levels of phospho-PKB Thr-308 (Fig. 9, lower panel) and phospho-PKB Ser-473 as shown previously (36). Thus, either or both Sgk and PKB could be the components of the PI3K pathway that mediate FSH (not IGF-I)-dependent FKHR phosphorylation in these differentiated granulosa cells.
|
), supporting previous results (Fig. 7) (36). Furthermore, culturing granulosa cells with IGF-I alone for 48 h had little effect on the amount or phosphorylation of FKHR protein. Addition of forskolin for 2 h to cells cultured with IGF-I alone, decreased the amount of FKHR protein and thereby increased the relative phosphorylation of FKHR at Ser 256 (5-fold) and Thr-24 (6-fold). Thus, FSH/T (cAMP) more than IGF-I appear to be potent regulators of the FKHR expression and phosphorylation in these differentiating granulosa cells. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
We show herein for the first time that the expression of FKHR in granulosa cells is mediated at many levels by the intraovarian regulators E2 and IGF-I as well as external factors FSH and LH (see model, Fig. 10). Most striking is the pronounced ability of estrogen to increase levels of FKHR mRNA and protein in granulosa cells of H rats, indicating that E2 exerts a potent positive effect on this component of the IGF-I signal transduction cascade (Figs. 4, 5, 10). Moreover, estrogen not only increases expression of FKHR mRNA and protein but also up-regulates other notable components of the IGF-I signaling system, including IGF-1Rß subunit and the glucose transporter, Glut-1 (Fig. 5, 10). The coordinated up-regulation of FKHR with IGF-1Rß and Glut-1 indicate further that E2 enhances granulosa cell function in the H rat model by regulating three different targets that control cellular energy flow, glucose metabolism and cell survival. Because IGF-I helps maintain high expression of ERß mRNA, at least in cultured granulosa cells, and because Forkhead proteins can activate or repress ER (50, 51), at least in some situations, E2 and IGF-I appear to comprise an autocrine regulatory system in granulosa cells that promotes cell survival and proliferation (see model, Fig. 10).
|
and 5). Exposure of preovulatory follicles to hCG irreversibly down-regulates FKHR expression in mouse and rat granulosa cells as they terminally differentiate to nondividing luteal cells. The inhibitory effects of FSH and LH in vivo are mimicked in cultured granulosa cells where FSH or forskolin (but not IGF-I) caused a dramatic down-regulation of FKHR mRNA as the granulosa cells differentiated (Figs. 1, 4–7). Thus, at the transcriptional level, E2 and IGF-I appear to support FKHR expression, whereas agonists that stimulate cAMP signaling cascades exert negative regulatory effects both by decreasing transcription of the FKHR gene (Fig. 10). Regulation of Forkhead family members of the Foxo subclass occurs not only at the level of their transcription and expression but also by phosphorylation (16, 20). Specifically, IGF-I-induced phosphorylation of FKHR by PKB (or related kinases) (16) leads to a decrease in apoptotic signals because the movement of phospho-FKHR to the nucleus is restricted (20) and thereby blocks its transcriptional activation of genes such as FasL, p27KIP, and IGFBP-1. Although this model is attractive and supported by an abundance of data in other systems, (primarily cotransfection analysis in cultured cells (21, 22, 23), the role of Forkhead proteins in granulosa cells and the oocyte has not yet been determined. Moreover, each Forkhead gene is expressed in a cell-specific manner at defined stages of follicular growth. Most intriguing are the data that show the levels of FKHR mRNA protein and phosphorylation are not strictly associated with follicles that are undergoing apoptosis. Rather, FKHR is most abundant in granulosa cells that are highly proliferative (26, 48, 59), express high levels of cyclin D2 (59), and ERß (57) and show increased staining for PCNA/BrdU (26). In addition, these cells express Glut-1, a requisite for energy homoestasis. It is important to note that granulosa cells of healthy follicles also express FasL (24, 25) and p27KIP (26, 59). Apoptosis can be triggered in these granulosa cells by specific insults (24) or changes in the hormonal milieu, indicating that neither FasL nor FKHR per se triggers apoptosis but may facilitate the process.
Granulosa cells become resistant to apoptotic insult if they are stimulated with FSH/LH to undergo luteinization (24). At this stage of differentiation, factors that impact proliferation (E2, IGF-I, cyclin D2) and apoptosis (FKHR and FasL) are lost, whereas factors that are expressed in luteal cells and presumed to impact luteinization (FKHRL1, AFX, Sgk, IGFBP-1, p27KIP, and p21CIP) are increased or acquired (Fig. 10). Therefore, it is possible that FKHR, FKHL1, and AFX exert different functions that are dependent on the cell type, stage of cell differentiation, or specific associated proteins. In this regard, one study reports that AFX up-regulates p27KIP (22). In another study, Tanaka et al. (60) show that expression of FKHRL1 does not activate p21CIP or p27KIP promoter activity or regulate Fas L expression. Therefore, the precise relation of FKHR proteins to the regulation of these genes remains uncertain. Recently, FKHR has been shown to interact with and selectively modify the functional activity of other transcription factors, specifically members of the nuclear steroid receptor superfamily (50, 51). FKHRL1 and AFX may have similar or different capabilities. Thus, the function of Forkhead proteins may depend not only their specific transcriptional activities but also on the hormonal milieu, the cell context, and the levels of proapoptotic and antiapoptotic factors (61, 62).
Based on the current model of the IGF-I signal transduction, we initially predicted that PKB and Sgk activation (i.e. phosphorylation) would be related to the phosphorylation of FKHR in granulosa cells. This model now appears to be too simplistic, at least for granulosa cells in vivo. In granulosa cells of H rats, PKB was phosphorylated on Ser-473 and Thr-308 at many stages of follicular development, even in the absence of E2, FSH, and LH. Thus, factors other than the gonadotropins can regulate intrafollicular phosphorylation of PKB in these follicles. This is not surprising because many growth factors can impact the PI3K/PDK1/PKB pathway (14) and because IGF-I remains expressed in granulosa cells of H rats (2). Furthermore, the pattern of PKB phosphorylation at Ser-473 or Thr-308 in granulosa cells of H rats does not relate temporally or hormonally to the expression or phosphorylation of FKHR, a presumed direct target of PKB (Fig. 5) (14, 16). Thus, PKB may be one but not the only regulator of FKHR in granulosa cells in vivo. A specific isoform of PKB or Sgk may be critical (63). Or, other kinases may be important.
In cultured granulosa cells, phosphorylation of FKHR as well as PKB is induced rapidly by FSH and/or IGF-I with remarkably similar kinetics during the initial phase of stimulation. FKHR that is nuclear localized in unstimulated cells cultured in serum-free medium is rapidly redistributed to cytoplasmic structures in response to either FSH or IGF-I within 5 min and is retained for at least 30 min. At later stages of culture when the granulosa cells have become more differentiated, FSH/T and IGF-I exert different effects on the amount and phosphorylation of FKHR (Fig. 9). FSH/T decreases the amount of FKHR protein, whereas IGF-I does not. Nor does IGF-I prevent the FSH-mediated loss of FKHR protein or RNA. At the same time, FSH/T increases the relative amount of phospho-Ser-256 FKHR, indicating this phospho-form is markedly increased relative to the nonphospho-form of FKHR. This effect of FSH/T appears to be related, in part, to the induction and phosphorylation of Sgk at this time (Fig. 9) (36, 38) as well as to increased activation of PKB (Fig. 9) (36). Whether or not these kinases have overlapping or redundant functions in differentiated granulosa cells remains to be determined in vitro as well as in vivo. However, it is tempting to speculate that Sgk may impact other aspects of granulosa cell function to repress expression of FKHR and increase FKHRL1 and AFX. To what extent this apparent switch in Forkhead proteins (or the marked decrease in FKHR alone) impacts luteinization or permits granulosa cells to become resistant to apoptotic insults is not yet known.
Based on these studies our working model (Fig. 10) is that FKHR, FKHRL1, and AFX are expressed in the rodent ovary. FKHR is selectively expressed in granulosa cells of growing follicles and in these cells, is differentially regulated by hormones. First, FSH/PMSG via aromatase and production of E2 enhance the IGF-I signaling cascades by increasing the cellular levels of FKHR mRNA and protein as well as IGF-1Rß and Glut-1 protein. Second, IGF-I supports expression of ERß. Thus, E2 (via FSH and aromatase) and IGF-I appear to comprise an autocrine regulatory system in granulosa cells favoring energy flow, glucose metabolism and cell survival. Conversely, FSH and more specifically LH (via cAMP) markedly down-regulate FKHR (but not FKHRL1 or AFX) expression as granulosa cells differentiate to nondividing luteal cells. Superimposed on the transcriptional regulation of FKHR by FSH, E2, and IGF-I is the ability of FSH and IGF-I to stimulate rapid PI3K-dependent phosphorylation of FKHR. Both FSH and IGF-I can stimulate rapid phosphorylation of FKHR (and FKHRL1) at multiple serine and threonine residues that is PI3K dependent and may be mediated in part by activation of PKB and induction of Sgk in differentiated granulosa cells. Phosphorylation of FKHR is associated with its redistribution from the nucleus to the cytoplasm of granulosa cells. Although this provides an acute mechanism by which to reduce the amount of transcriptionally active FKHR in the cell, the short-term vs. long-term consequences of this on granulosa cell function are not yet clear. The precise ratio of nonphospho (nuclear) FKHR to phospho (cytoplasmic) FKHR in vivo and in vitro in not known and likely depends on the activities of many factors that impact granulosa cell function. Thus, although some of the transcriptional regulators of FKHR in granulosa cells have been identified, the targets of FKHR are not yet known. The elevated expression of FKHR in granulosa cells of growing follicles indicates that it is linked to the proliferative as well as apoptotic pathways in granulosa cells. Expression of FKHRL1 and AFX (but not FKHR) in luteal cells provides an additional layer of regulation. Combined FKHR, FKHRL1, and AFX may help coordinate proliferative, apoptotic, and differentiative events in the ovary.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Cell Culture
Intact immature (23 d of age) rats were primed with E2 (1.5 mg/0.2 ml propylene glycol) for 3 d. Granulosa cells were harvested by needle puncture, pooled, and plated according to routine procedures (55, 66). Cells were cultured on serum-coated plated in defined, DMEM-F12 medium with and without FSH (NIH-FSH-16, the generous gift of Al Parlow), T (from Steraloids, Keene, NH), forskolin, IGF-I and H89 (from Calbiochem, San Diego, CA) or LY294002 (Alexis, San Diego, CA).
In Situ Hybridization
Primers (as described below) were designed to amplify specifically cDNAs encoding mouse and rat FKHR (Foxo1) as well as mouse AFX (Foxo4) and FKHRL1 (Foxo3) mRNAs. The mouse FKHR, FKHL1 and AFX cDNAs were amplified, and subcloned into the pCR4-TOPO vector (Invitrogen,Carlsbad, CA). In situ hybridization was performed as described previously by Wilkenson (67) and as reported by our laboratory (59). Ovaries from mice and rats were fixed immediately in 4% paraformaldehyde in PBS overnight at 4 C before dehydration and paraffin embedding. Sections (6 mm) were baked at 42 C overnight onto 3-amino propyltriethoxysilane coated slides. Slides were prehybridized, hybridized, washed, exposed, and developed as previously described (59). The 35S-labeled riboprobes were also produced as previously described using the Riboprobe In Vitro Transcription Systems Kit from Promega Corp. (Madison, WI). Sgk sense and antisense probes were produced by transcription from the T3 and T7 promoters, respectively, on the NheI digested pBS-sgk vector (32). FKHR, FKHRL1, and AFX sense and antisense probes were produced by transcription from the T3 and T7 promoters using NotI and SpeI digested vector, respectively. Each slide was incubated in 80 µl of hybridization solution containing 5 million counts of the appropriate probe overnight at 55 C in a humid chamber. After washing, slides were exposed overnight to X-OMAT-AR film, dipped in NTB-2 emulsion, and developed (reagents from Eastman Kodak Co., Rochester, NY) according to the intensity of the x-ray film. For most experiments, 3 d of exposure were sufficient to obtain a strong signal. For each in situ hybridization analysis, slides containing ovaries in each treatment group were included to permit direct comparisons of the relative amount of each mRNA signal during follicular development and luteinization. Slides hybridized with the sense probes were also done for each experiment to control for background.
RT-PCRs
RT-PCRs were performed according to established procedure (68). Briefly, total RNA was extracted from tissues or cells with TRIzol (Life Technologies, Inc., Grand Island, NY) and purified by the manufacturer’s specification. Total RNA (300 ng) was reversed transcribed using was reverse transcribed using oligo polydeoxythymidine (Pharmacia Biotech Inc., Piscataway, NJ) and avian myeloblastosis virus reverse transcriptase (Promega Corp., Madison, WI). To determine the linear range of amplification for specific mRNAs, 300 ng RNA was reverse-transcribed and amplified in a range of cycle numbers. Next, increasing amounts of RNA (75–1200 ng) was reverse-transcribed and PCR-amplified at a selected cycle number. Products were amplified using specific primer pairs within the linear range for each gene product. RT-PCR (from 300 ng RNA) for FKHR, IGF-I and Sgk for (r > 0.91) after 25 cycles. Thirty cycles were linear for FKHRL1 (r > 0.95), AFX ER, and ERß. Amplified products were resolved by PAGE. Dried gels were quantitated using a Storm 860 PhosphorImager (Molecular Dynamics, Inc., Sunnyvale, CA) and exposed to autoradiographic film. The authenticity of each product was verified by cloning, sequence analyses or restriction digests.
Primers (from Genosys, The Woodlands, TX):
Mouse Sgk (accession no. AF205855): 1214–1561: forward 5'-gcacttcgatcccgagttta; reverse 5'-ttgagaggagggtgtgctct
Rat Sgk (accession no. NM019232): 1725–2076: forward 5'-ctgcaatgtgccttttctga; reverse 5'-atgcttccctcaagcatctg
Rat/mouse IGF-I (accession no. M15649; J02743): 397–793: forward 5'-gaacagaaaatgccacgtca; reverse 5'-gcagccaaaattcagagagg
Mouse/rat FKHR/Foxo1 (accession no. AF114258): 1399–1798: forward 5'acgtgccattccctggtgtat; reverse 5'-tcattgtggggaggagagtc
Mouse FKHR2/FKHRL1/Foxo3 (accession no. AF114259): forward 5'-gtcatgggccacgataagtt; reverse 5'-gggctgctaacagtctctgc
Mouse AFX/Foxo4 (accession no. Ab032770): forward 5'-cctcctgctgatgtcctcat; reverse 5'-tgctgtgactcagggatctg
Rat ERß’3a forward 5'-ttcccggcagcaccagtaacc; reverse 5'-tccctctttgcgtttggacta (47)
Rat ER: forward 5'-aattctgacaatcgacgccag; reverse 5'-gtgcttcaacattctccctcctc
L19 primers were as described (69).
Western Blot Analyses: Antibodies
Antibodies were obtained from Cell Signaling Technology, Inc. (Beverly, MA) for FKHR (no. 9462), FKHR phospho-Thr-24 (no. 9464), FKHR-phospho-Ser-256 (no. 9461); PKB/Akt (no. 9916), PKB-phospho-Ser-473 (no. 9971), PKB-phospho-Thr-308 (no. 9275), from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) for IGF-1Rß (sc 713), from Chemicon (Temecula, CA) for Glut-1 (no. AB1340), and from Upstate Biochemical (Lake Placid, NY) for IRS-2 (no. 06–506) and FKHRL1 phospho-Ser-253 (no. 06–953). Antibody for FLKHRL1 phospho-Ser-315 was generously provided by Dr. Michael Greenberg (MIT, Boston, MA). Anti-Sgk antibody was generously provided by Dr. Gary L. Firestone (UC Berkeley, Berkeley, CA).
Cell Extracts
Protein was isolated from granulosa cells and luteal cells by homogenization in WCE buffer (10 mM Tris, 1 mM EDTA, 1 mM DTT, 400 mM KCl, 10% glycerol, 1 mM PMSF, 1 mM vanadate, 1 mM diethyl dithiocarbamic acid, 0.1 mg/ml aprotinin) followed by centrifugation (1 min in microfuge) to isolate soluble protein (36, 38). Concentrations of soluble protein in each sample were determined by Bradford assay [Bio-Rad Laboratories, Inc. (Hercules, CA) reagents]. Western blots were run using 30 µg of WCE protein. For in vitro experiments with granulosa cells cultured with or without agonist stimulation, protein extracts were prepared by adding 200 µl of boiling SDS buffer to each well (70), scraping followed by boiling for 5 min. Twenty microliters of each sample in each treatment group were analyzed.
Blotting
One dimensional SDS-PAGE with 4.5% stacking and 10% separating acrylamide gels was used to resolve proteins. Proteins were electrophoretically transferred to 0.45-mm Immobilon membranes and blocked for 1 h in PBS containing 5% milk and 0.1% Tween-20. After one 20-min incubation in wash solution (1% milk in PBS and 0.1% Tween-20), filters were incubated overnight at 4 C with a 1:1000 dilution of all antibodies with the exception of IGF-1Rß (1:500) and PKB phospho-Ser-308 (1:100). After 3 washes (10 min each), blots were incubated with 1:10,000 antirabbit-HRP with the exception of PKB phospho-Ser-308 (1:500). Blots were washed as described above and the immunopositive bands detected using the enhanced chemiluminescence assay system, ECL. Immunoreactive signals were analyzed and quantified using an AlphaImager 2000 (3.3) (Alpha Innotech Corp., San Leandro, CA).
Immunocytochemistry
Immunocytochemistry was performed as previously described (36, 38). Cells were cultured on serum-coated coverslip in defined medium. Agonists were added as described and cells were fixed in 4% paraformaldehyde. Cells were permeabilized with 0.5% NP-40 in saline and processed by routine procedures. Primary antibodies were diluted 1:50 and the fluorescein-labeled IgG (Pierce Chemical Co., Rockford, IL) diluted 1:20.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: AFX, A Forkhead transcription factor; CL, corpora lutea; CREB, cAMP response element binding protein; FasL, Fas ligand; FKHR, Forkhead homolog in rhabdomysarcoma; FKHRL1, Forkhead-like protein-1; H, hypophysectomized; hCG, human CG; HE, E2 primed; HEF, hypophysectomized, E2 and FSH treated; IRS 1-2, insulin receptor substrates; PDK1/2, phosphoinositide-induced kinases; PKB, protein kinase B; Sgk, serum and glucocorticoid-induced kinase; WCE, whole cell extracts.
Received for publication September 4, 2001. Accepted for publication December 21, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
