Minireview: Roles of the Forkhead Transcription Factor FOXL2 in Granulosa Cell Biology and Pathology

The forkhead transcription factor FOXL2 plays key and diverse roles in female sex determination and ovarian development and in the postnatal ovary and follicle maintenance; its mutations can lead to premature ovarian failure and granulosa cell tumor formation.

Abstract

The forkhead transcription factor (FOXL2) is an essential transcription factor in the ovary. It is important in ovarian development and a key factor in female sex determination. In addition, FOXL2 plays a significant role in the postnatal ovary and follicle maintenance. The diverse transcriptional activities of FOXL2 are likely attributable to posttranslational modifications and binding to other key proteins involved in granulosa cell function. Mutations of FOXL2 lead to disorders of ovarian function ranging from premature follicle depletion and ovarian failure to unregulated granulosa cell proliferation leading to tumor formation. Thus, FOXL2 is a key regulator of granulosa cell function and a master transcription factor in these cells.

The forkhead transcription factor (FOXL2) is emerging as a central transcription factor in ovarian development and the growth and maturation of ovarian follicles. FOXL2 is a member of the forkhead/hepatocyte nuclear factor 3 gene family (FKH/HNF3) of transcription factors, the first of which was described in Drosophila (1). Now members have been identified in species ranging from yeast to human (2) and play essential roles in embryogenesis (1, 3), cell differentiation (4–6), and tumorigenesis (7–9). FKH/HNF3 family members are characterized by a conserved winged helix domain that is essential for DNA binding but exhibit divergent transcriptional regulation based on their transactivation or transrepression domains (2, 10, 11).

FOXL2 was first cloned by Crisponi et al. (12) from the Blepharophimosis-Ptosis-Epicanthus Inversus Syndrome (BPES) region on human chromosome 3q23. Heterozygous mutations in FOXL2 lead to a characteristic premature ovarian failure and infertility in females (13). Consistent with this, FOXL2 is expressed in the developing eyelids and ovary in the mouse (12) and is also expressed in adult ovarian follicles (12, 14–16). Unlike the primary ovarian failure from primordial follicle arrest that occurs in mice null for Foxl2 (17, 18), which have complete absence of FOXL2 expression, heterozygous mutations in humans with BPES undergo a complete sequence of follicle development with early depletion of the follicle pool and premature ovarian failure (17, 19). Thus, FOXL2 plays a significant role in early ovarian development and sex determination, as well as a later role in granulosa cell differentiation with subsequent follicle depletion, and mutations of FOXL2 contribute to a variety of conditions and disease states. The functional role of FOXL2 in the ovary will be the focus of this review.

Role of FOXL2 during Early Ovarian Development and Sex Determination

FOXL2 is the earliest known marker of ovarian differentiation in mammals and is expressed in the developing mouse ovary as early as 12.5 d post coitum (12, 17, 20). Several genes have been implicated in early gonadal development and sex determination, including wingless-type MMTV integration site family member 4 (Wnt4) and R-spondin-1 (Rspo1) (21, 22), dosage-sensitive sex reversal, adrenal hypoplasia critical region on chromosome X gene 1 (Dax1) (23), SRY (sex determining region Y)-box 9 (Sox9) (24, 25), and Foxl2. Of these, FOXL2 is restricted to granulosa cells and the early ovarian stroma (17, 20), is excluded from the developing male gonads, and appears to be specifically required for female sex determination (26). Knockout mouse models have shown that ablation of Foxl2 expression blocks ovarian follicle formation and leads to partial ovary-to-testis sex reversal in mice (17, 18), independent of Wnt and Rspo1 (27). Several downstream targets of FOXL2 have been identified through expression profiling (27). These include genes that are down-regulated by FOXL2 and are involved in neuronal or vascular development, such as the transcription factor odd Oz/ten-m homolog 4 (Odz4) (28), plasma transmembrane proteins such as plexin (Plxnc1) (29), the cadherin-domain containing calsyntenin 2 (Clstn2) (30), and the leucine-rich repeat protein (Lrrc4) (31), which may contribute to formation of the ovarian-cortico-medullary axis (28, 32). Genes that are down-regulated by FOXL2 and may contribute to its role in female sex determination were also identified (27), including Grip1, a nuclear repressor required for estrogen receptor (ER)-α activity (33), and the aldo-keto reductase, Akr1c14, which metabolizes the androgen dihydrotestosterone (34). Sox9 and Inhibin B (Inhbb), which are central to testicular development (24, 25, 35, 36), are also down-regulated by FOXL2 (27). Genes that were up-regulated by FOXL2 included P450aromatase (Cyp19) and liver receptor homolog-1 (Lrh-1/Nr5a2) (37), which is related to steroidogenic factor 1 (SF-1), an orphan receptor essential for the formation of female and male gonads (38). Consistent with its role in sexual determination, ablation of Foxl2 in the adult mouse ovary causes immediate induction of the transcription factor SOX9 (39), leading to ovary-to-testis transdifferentiation, with Sertoli cells replacing granulosa cells and seminiferous tubules forming instead of follicles (39).

FOXL2 has been associated with sex reversal in several species, including the frog and several fish species (medaka, Nile tilapia, and Japanese flounder) (40–43), and dysregulated FOXL2 expression has been described in the goat polled intersex syndrome (PIS), in which XX male sex reversal occurs (44); taken together, these demonstrate the highly conserved role of FOXL2 in ovarian development. These effects of FOXL2 may be mediated via control of Cyp19 expression in these species (40–43). For example, during the temperature-sensitive phase of sex differentiation in XX Japanese flounder and Nile tilapia females, high temperatures result in down-regulation of Cyp19 expression, leading to decreased estrogen levels and masculinization. This down-regulation of Cyp19 is mediated by FOXL2 and FSH signaling (43, 45). In the goat, PIS leads to absence of the horns in both sexes in a dominant fashion, and to XX female-to-male sex reversal in the recessive state (44). An 11.7-kb DNA element located upstream of the Foxl2 gene was found to be deleted in PIS, and expression levels of Foxl2 and Cyp19 in the gonads were greatly reduced, resulting in masculinization (46). Thus, FOXL2 plays a highly conserved role in the regulation of early ovarian development and female sex determination across multiple species.

FOXL2 Activity during Folliculogenesis

FOXL2 is expressed in the less differentiated granulosa cells of small and medium follicles (12, 16, 17, 20) and therefore likely also plays a role in granulosa cell differentiation and follicle development and maintenance. The alanine/proline-rich carboxyl (C-) terminus of FOXL2 is characteristic of transcriptional repressors (47) and is found in other forkhead transcription factors that function in repressing cellular differentiation (6, 48–52). FOXL2 was initially found to function as a transcriptional repressor of the steroidogenic acute regulatory (StAR) gene (16), and the entire C terminus of FOXL2 functions as a transrepression domain (16, 53). StAR translocates cholesterol from the outer to the inner membrane of mitochondria, which is the rate-limiting step in steroidogenesis (54–56). StAR is also a marker of granulosa cell differentiation and is expressed in granulosa cells of large preovulatory follicles, but not small and medium immature follicles (57). FOXL2 binds to the human StAR promoter and suppresses its activity (16), suggesting that transcriptional repression by FOXL2 prevents StAR expression in immature follicles. FOXL2 also functions as a transcriptional repressor of Cyp19 (P450aromatase) in rainbow trout gonads (58) and both Cyp19 and Cyp11a (P450scc) in immature mouse ovary follicles (53). P450aromatase converts C-19 androgens to C-18 estrogens, a key product of differentiated granulosa cells, and P450scc cleaves the side chain of cholesterol to produce pregnenolone, the first committed and rate-limiting step in steroid hormone synthesis (56). These results indicate that sex steroids may be important for initiating follicle growth, as previously suggested (59, 60). FOXL2 also represses transcription of cyclin D2, which regulates cyclin-dependent kinases 4 or 6 (Cdk4 or Cdk6) to control G1 phase progression of the cell cycle and is involved in granulosa cell proliferation (61, 62). These results indicate that FOXL2 may function as a suppressor of ovarian follicle progression in small and medium follicles by the prevention of premature differentiation and/or proliferation of granulosa cells, thus preventing the premature depletion of ovarian follicles which is seen when FOXL2 function is altered because of mutations.

In postnatal ovaries of Foxl2+/− heterozygous mice (which have been suggested as a mouse model for the human phenotype of premature ovarian failure) (17, 18), a number of somatic cell genes were found to be repressed to a lesser degree but similar to Foxl2−/− null mice (27), including InhbB, Lrh-1, Cyp11a, and another steroidogenic enzyme and cytochrome P450 family member, 17-hydroxylase (Cyp17a1). This may be the result of Foxl2 dose dependency or, because Foxl2 is expressed in granulosa cells of heterozygotes, these genes may be differentially regulated in heterozygous mice.

The KGN cell line (63) has also been widely used as a model of granulosa cells of the postnatal ovary. These adult granulosa cells are derived from granulosa cell tumors (GCTs) which contain a C402G mutation in FOXL2 (discussed below). Therefore, they may actually reflect changes that occur in the pathological state (see below) rather than normal granulosa cells. However important information has been revealed by studies of this cell line. In these cells, a number of genes have been shown to be regulated by FOXL2 (64). These included immunomodulators such as interferon β 1 (IFNB1) (65), interleukin-12 subunit α (IL12A) (66), and interleukin-29 (IL29) (67); intercellular adhesion molecule 1 (ICAM1) (68); proteins involved in cell proliferation, differentiation, and transformation such as the cellular proto-oncogene c-FOS (69); and antiapoptotic factors such as BCL2(B-cell lymphoma 2)-related protein A1 (BCL2A1) (70) and immediate early response 3 (IER3) (71).

A number of genes that lead to fertility defects also appear to be regulated by FOXL2 in the postnatal ovary. These include prostaglandin-endoperoxide synthase 2/cyclooxygenase-2 (PTGS2/COX-2). PTGS2/COX-2 null mice exhibit ovulation and fertilization defects (72). Although FOXL2 up-regulates PTGS2/COX-2 in the KGN granulosa cell model (64), PTGS2/COX-2 is repressed by FOXL2 through a nonclassical estrogen signaling pathway in vitro (73). Thus, depending on the available binding partners, FOXL2 likely has differential effects on a number of genes imperative for ovarian function and female fertility. FOXL2 up-regulates the SF-1–related gene, Lrh-1, in both the mouse ovary, as noted above (27), and in the KGN cell line (64), and also up-regulates peroxisome proliferator-activated receptor-coactivator 1α (Ppargc1A) in KGN cells (64). Ppargc1A binds to both Lrh-1 and SF-1 and markedly enhances their effects on the transcription of genes involved in progesterone production in granulosa cells (74). However, at the protein level, FOXL2 binds to SF-1 and negatively regulates transcriptional activation of other steroidogenic enzymes in human granulosa cells (75). Thus, FOXL2 plays critical roles in ovarian development, folliculogenesis, and maintenance, and mutations of FOXL2 lead to ovarian failure as well as malignancy, effects which are mediated through the differential regulation of a diverse set of genes in granulosa cells. This differential regulation is likely dependent on the posttranslational modifications and available FOXL2 binding partners that make up the transcriptional complex leading to gene regulation.

Posttranslational Modifications of FOXL2

In keeping with a central role for FOXL2 in the ovary, several posttranslational modifications have shown to be involved in modulating its activity, as summarized below.

Sumoylation

Sumoylation is a key mechanism in transcriptional regulation and acts to influence protein stability and subcellular localization (76, 77). Small ubiquitin-related modifier (SUMO) proteins, structurally related to ubiquitin, covalently bind to substrate proteins at lysine residues, in a three-step conjugation pathway similar to that involved in ubiquitination (78). FOXL2 is sumoylated (79), and we recently showed that it is sumoylated by SUMO1 but not SUMO2/3, and that this sumoylation is mediated by ubiquitin-conjugating enzyme-9 (Ubc9) (80). Further, we found that sumoylation enhances FOXL2's activity as a transcriptional repressor of the StAR promoter (80). Marongiu et al., have recently identified four additional sumoylation sites, all of which may be involved in FOXL2 function (81) because they affect localization, sumoylation, and transcriptional activity. Because sumoylation is a dynamic process, the differences in sumoylation sites identified by the two groups may be attributable to different culture conditions and perhaps other upstream regulators that may influence sumoylation of FOXL2. Nevertheless, the data clearly demonstrate a key role for sumoylation in regulating the functional activity of FOXL2.

Phosphorylation

FOXL2 is also phosphorylated, and this phosphorylation regulates the functional activity of FOXL2 (79, 82). We have shown that FOXL2 is phosphorylated by the serine/threonine kinase large tumor suppressor 1 (Lats1), and that this enhances FOXL2's activity as a repressor (82). Similar to the ovarian failure phenotype in mice null for Foxl2 (17, 18), deletion of Lats1 in mice also result in an ovarian phenotype similar to POF (83). Therefore, FOXL2 and LATS1 likely function through a similar pathway during follicle maintenance, and phosphorylation may be another control mechanism regulating FOXL2 activity in granulosa cell differentiation and hence, follicle maturation.

Acetylation and Cell Stress

Veitia and coworkers (79) demonstrated that cell stress significantly influences the posttranslational modification of FOXL2. They found that oxidative stress and heat shock both increase FOXL2 expression levels in the KGN cell line, and that the FOXL2 protein appears to be hyper-acetylated upon oxidative stress. This resulted in increased recruitment of FOXL2 to target promoters, including several genes involved in stress response or regulation of apoptosis (79). Of these, Manganese Superoxide Dismutase (MnSOD), a key mediator of the oxidative stress response, was found to be specifically and directly up-regulated by FOXL2. The authors also found that both the expression level and transcriptional activity of FOXL2 are down-regulated by the NAD-dependent deacetylase sirtuin 1 (SIRT1), and FOXL2 itself induces SIRT1 transcription, demonstrating the importance of acetylation as a regulator of FOXL2 activity and the existence of a feedback loop controlling FOXL2 activity in these cells (79). This response to oxidative stress in the GCT KGN cell line may actually be a mechanism defining the uncontrolled cell proliferation in the pathological state and warrants further investigation.

Potential FOXL2 Binding Partners

DEAD box-containing protein DP103

The DEAD box-containing protein DP103 is known to bind to SF-1 and more recently identified to bind to FOXL2 (84, 85). Co-expression of DP103 and FOXL2 potentiates cell death. DP103 represses the transcriptional activity of SF-1 (84), a key regulator of steroidogenesis, during fetal ovarian differentiation and reproduction (38, 86). The region of DP103 that interacts with FOXL2 is also necessary for SF-1 binding to DP103 (84), suggesting that DP103, FOXL2, and SF-1 may be components of a complex regulatory mechanism in the ovary.

Steroidogenic factor-1

FOXL2 binds directly to SF-1, and recently the nature of the interaction between FOXL2 and SF-1 has been further elucidated (75). FOXL2 and SF-1 proteins interact in granulosa cells, and FOXL2 negatively regulates the transcriptional activation of CYP17a1 by SF-1, by inhibiting binding of SF-1 to the CYP17 promoter (75). These findings illustrate another potential mechanism by which FOXL2 regulates ovarian steroidogenesis and normal ovarian follicle development.

ERα and SOX9

The transcription factors FOXL2 and SOX9 are required for female and male mammalian gonadal development (see above), and alterations in their expression have been implicated in various disorders of sex development (DSD), discussed below (26). Uhlenhaut et al. (39) demonstrated that FOXL2 represses the testis differentiation program mainly through repression of Sox9-regulatory sequences that are required for its testis-specific expression, and found that FOXL2 and ERα cooperate, through protein–protein interactions, in Sox9 repression in vivo. Conditional loss of FOXL2 in the adult ovary results in ovary-to-testis differentiation (17).

ERα – nonclassical activity

An additional role for FOXL2 in modulating ER activity has also been investigated (73). Although FOXL2 has not been shown to be involved in the classical ERα-mediated transcriptional pathways, in which binding of ligand-activated receptors canonical estrogen response elements (EREs) results in the up- or down-regulation of target genes, a role for FOXL2 in mediating nonclassical tethered transcriptional pathways has been demonstrated (73). Specifically, FOXL2 selectively repressed stimulation of an AP1 reporter through ERα by tamoxifen, likely through FOXL2 binding to ERα (73). PTGS2/COX-2, which is required for ovulation (72) and is induced by the nonclassical ER pathway, is suppressed by FOXL2 (73). These findings illustrate that FOXL2 is likely involved in regulating ovarian function through a number of important binding partners, and the complete transcriptional complex likely alternates depending on the upstream signaling pathways.

SMAD

In addition to its functions in the ovary, FOXL2 localizes to α-glycoprotein subunit- and FSH β-positive cells of the adult mouse pituitary and is present in αT3–1 and LβT2 cells (87), but its role remains largely unknown. Activin is a key regulator of follicle development and initiates granulosa cell proliferation (88). Follistatin is a transcriptional target of activin (89) and also modulates activin action (90). In gonadotropic αT3–1 cells, activin induces follistatin transcription via SMAD3 action at an intronic Smad-binding element (SBE1) (87), and FOXL2 functions as a SMAD3 partner in this SBE1-mediated transcription, binding to a forkhead-binding element (FKHB) downstream of the follistatin gene SBE1 site (87). Activin is also a major physiological regulator of FSH (91), and Corpuz et al. recently demonstrated that FOXL2 is also required for activin induction of both mouse and human FSHβ (92). Interestingly, this induction is SMAD-dependent for the mouse gene, but SMAD-independent for human FSHβ (92). We recently demonstrated expression of all SMADs in human granulosa cells (Kuo et al., submitted), suggesting that FOXL2 may also function as a transcriptional regulator and coordinator of SMAD3 targets in the ovary.

FOXL2 in BPES

BPES is an autosomal dominant disorder and is associated with heterozygous mutations of FOXL2 (12). Patients with BPES type 1 exhibit a characteristic eyelid dysplasia together with premature ovarian failure and infertility in affected females (13). Ovaries from BPES type 1 patients are variable in their histological appearance, ranging from the presence of some primordial follicles with atretic follicles to the complete absence of follicles (19, 93). In contrast, BPES type 2 patients present with the eyelid defects only, and both males and females are fertile (13). In BPES type 1, nonsense mutations in the FOXL2 gene create premature stop codons, which are predicted to result in truncated proteins lacking the C-terminal alanine/proline-rich transrepression domain (12, 16, 94–96), whereas FOXL2 mutations in BPES type 2 are predicted to result in expansion of the polyalanine tract (94). These differences in the FOXL2 mutations associated with BPES type 1 and type 2, and loss of the transrepression domain vs. the polyalanine tract, likely underlie the differences in the resulting ovarian phenotypes.

Potential Mechanisms of FOXL2 Mutants in Premature Ovarian Failure in BPES Type 1

Mice deficient for Foxl2 exhibit follicle arrest between the primordial and primary stages, followed by follicle degeneration (17, 18). They fail to undergo sexual maturation and undergo primary ovarian failure (17, 18). However, this contrasts with the majority of human BPES patients. Unlike the complete loss of Foxl2 alleles in null mice, BPES patients typically have heterozygous FOXL2 mutations and do not exhibit primordial follicle arrest, but rather undergo a complete sequence of follicle development, menstrual cyclicity, and ovulation, followed by premature follicle depletion and subsequent ovarian failure (13, 19, 93). As FOXL2 is expressed in primordial, primary, and larger secondary follicles (12, 14–16), it may play important roles during separate stages of follicle development leading to premature ovarian failure.

Patients with BPES type 1 carry heterozygous mutations of FOXL2 that are predicted to result in truncated proteins lacking the C-terminal transrepression domain. The human Q219 × mutation, common in BPES type 1 (12), produces a truncated FOXL2 protein, FOXL2 (a.a. 1–218), that lacks the entire alanine/proline rich region, however it retains the forkhead binding domain. This mutant fails to repress transcription of the StAR (16), and P450aromatase (53) promoters and functions as a dominant negative for wild-type FOXL2's activity as a transcriptional repressor of these promoters (16, 97), likely attributable to hetero-dimerization of the wild-type and mutant FOXL2 (Kuo et al., in preparation). This effect, leading to loss of transcriptional repression of key genes involved in granulosa cell proliferation, differentiation, and steroidogenesis, may result in accelerated granulosa cell differentiation and premature follicle depletion and thus contribute to the ovarian failure seen in BPES type 1.

In addition to potential heterodimer formation, missense mutations of FOXL2 have also been shown to lead to altered subcellular localization, ranging from a diffuse nuclear distribution to nuclear aggregation with cytoplasmic mislocalization (98). In the KGN cell line, this has been shown to lead to differences in transactivation capacity of these mutants, such as loss-of-function and dominant negative effects (98). Although these results should be viewed with caution, as KGN cells carry a mutation of FOXL2, they nonetheless illustrate that altered subcellular localization of mutant FOXL2 proteins may contribute to the development of BPES phenotypes in some patients.

Role of FOXL2 in GCT

In addition to mutations in FOXL2 leading to granulosa cell and follicle depletion, other mutations of FOXL2 are associated with unregulated granulosa cell proliferation, as in the case of adult onset GCTs. GCTs represent 5% to 10% of all ovarian cancers (99, 100) and are associated with a 20% mortality rate and a 5-year survival rate for advanced stage patients of less than 50% (101). GCTs exhibit some features that are similar to normal granulosa cells, including expression of the FSH receptor (102, 103). They bind FSH (104, 105), synthesize estradiol, inhibin, and anti-müllerian hormone (AMH) (106–110), and exhibit up-regulation of cyclin D2, which is necessary for FSH-stimulated granulosa cell proliferation (103). Shah et al. found that a recurrent heterozygous somatic mutation (C402G) in FOXL2 is present in 97% of adult GCTs (111). The C402G mutation was absent in all other tumors tested, including other ovarian stromal sex cord tumors, epithelial ovarian tumors, and breast cancers (111–114), and is also absent from the COV434 human GCT-derived cell line (115), which may in fact represent a juvenile GCT. This mutation is predicted to cause an amino acid change from cysteine to tryptophan (C134W) in the DNA binding domain of FOXL2 (111) and thus may affect its activity as a transcriptional regulator. However, Benayoun et al. (116) found that the transactivation ability of the mutant protein was similar to that of the wild-type, except for a promoter known to be coregulated by FOXL2 and SMAD3, and they did find that homodimer formation might be affected by the presence of the mutant protein. In juvenile GCTs, loss of protein expression is associated with an aggressive pattern of progression and may be a factor regulating unregulated granulosa cell proliferation.

Role of FOXL2 in DSDs

FOXL2 has been described as a key determinant of sexual differentiation in multiple species, including those that undergo sex differentiation secondary to environmental cues such as temperature (43, 45). In addition, genetic aberrations associated with Foxl2, including mutations/deletions in the mouse (24, 39) and deletions of the upstream region in goat (46), lead to ovary-to-testis transdifferentiation. In humans, SOX9 is expressed in (pre-) Sertoli cells, whereas FOXL2 is expressed in granulosa cells and ovarian stroma. In patients with DSDs and intersex states, when both ovarian and testicular development is present, expression of both FOXL2 and SOX9 can be detected, and if an ovotestis is present (i.e., hermaphrodite), expression of both genes can be detected in the same gonad (26). Interestingly, in one patient with DSD, FOXL2 expression was identified in well-developed seminiferous tubules, but it was never strongly coexpressed with SOX9 in the same cell. In a unique rare juvenile GCT of the testis, expression of FOXL2, along with the extinction of SOX9, was coupled with the transdifferentiation of a testicular cell into a GCT (117). Therefore, although different gonadal cell lineages may exist in these patients, FOXL2 and SOX9 expression is not present in the same cell, likely the result of FOXL2 repression of SOX9 similar to that seen in the mouse model.

Conclusions

Similar to other forkhead family members, FOXL2 is essential for embryogenesis, cell differentiation, and tumorigenesis; however, unlike other members with individual roles at these stages of regulation, FOXL2 plays an important role in all of these stages (Fig. 1). The highly conserved nature of this gene, and limited expression predominantly in the ovary, suggest that it is a key factor throughout ovarian development, both pre- and postnatally. Because of the vast processes FOXL2 regulates, it is not surprising that its function is modified at the posttranslational level. Further, various binding partners of FOXL2 can make up the transcriptional complex, leading to gene regulation attributable to as yet undefined upstream signaling (Fig. 1). Although we are beginning to develop an understanding of this recently discovered gene, there continues to be a great deal to discover, particularly the upstream cues determining its regulation and the modifications that take place. A better understanding of the functional role of the mutations associated with FOXL2 in pathological states ranging from aberrant sex determination to follicle depletion and cancer may help us develop treatments for these conditions.

Fig. 1.

Activity of FOXL2 as a transcriptional repressor. FOXL2 likely functions as a dimer (118) and requires sumoylation and phosphorylation for its activity. Other factors that may make up the transcriptional complex (Erα, SF-1, and DP103) are indicated, as are the genes known to be regulated by FOXL2. The effects of FOXL2 on gene transcription may differ depending on the stage of ovarian development, including sexual differentiation, granulosa cell development, and pathological states such as GCT formation.