Commentary on the Recent FSH Collection: Known Knowns and Known Unknowns

Abstract

Follicle-stimulating hormone (FSH) is a dimeric glycoprotein secreted by the anterior pituitary gonadotrope that is necessary for reproductive function in mammals. FSH primarily regulates granulosa cells and follicular growth in females, and Sertoli cell function in males. Since its identification in the 1930s and sequencing in the 1970s, significant progress has been made in elucidating its regulation and downstream function. Recent advances provide deeper insight into FSH synthesis, and effects in the gonads suggest potential roles in extragonadal tissues and examine pharmacological approaches and clinical applications in infertility treatment that now affect 18% of couples. These advances were discussed in detail in a number of reviews published in the last 2 years in Endocrinology. In this brief commentary, we summarize these reviews and point to the outstanding questions that should be answered in the near future to bridge a gap in our understanding of this hormone.

Follicle-stimulating hormone (FSH) is a gonadotropin hormone critical for reproductive function. It is synthesized and secreted from the anterior pituitary gonadotrope cells, primarily under the influence of hypothalamic gonadotropin-releasing hormone (GnRH) neuropeptide, and ovarian or the locally produced activin-inhibin-follistatin system (reviewed recently in (1)). Structurally, FSH is a dimeric glycoprotein composed of a unique β-subunit and a common α-subunit shared with luteinizing hormone (LH) and thyroid-stimulating hormone. Contrary to LH, FSH secretion is not entirely regulated, and most of FSH is constitutively released. FSH in males stimulates Sertoli cell proliferation during development and maintains Sertoli cell function in adults by inducing androgen-binding protein that contributes to spermatogenesis through interaction with testosterone. In females, FSH stimulates growth of ovarian follicles and aromatase expression in follicle granulosa cells, which synthesize estrogen. FSH deficiency in humans results in the absent or incomplete pubertal development in women, relatively normal pubertal development but azoospermia in men, and infertility in both women and men. Female mice deficient in FSH have a block in folliculogenesis prior to the antral stage resulting in infertility, while males have impaired reproductive function due to lower sperm count, but they are not infertile (variety of FSH mouse models reviewed recently in (2, 3)).

FSH concentration displays 2 temporally separate increases during the female reproductive cycle as opposed to LH, which exhibits a single rise. During the rodent estrous cycle, a surge of GnRH during the afternoon of proestrus triggers a surge in both LH and FSH, resulting in ovulation of the mature follicle in response to LH. Several hours later, during the morning of estrus, a secondary FSH increase occurs without a corresponding rise in LH. This secondary FSH rise is essential for follicular development for the subsequent cycle in rodents. In humans, the FSH level increases in the late luteal phase through the midfollicular phase of the menstrual cycle in addition to the preovulatory rise, corresponding to the recruitment of a new cohort of follicles to the growing pool. During reproductive aging in females, a rise in FSH is considered the hallmark of the reduction in the follicle reserve. The fall in inhibin B, caused by the loss of follicles, coincides with the rise in FSH, supporting the primary role of inhibins in the negative feedback of FSH.

Normal physiological function of FSH in women follows the Goldilocks principle: low FSH impedes follicular growth, while high levels are associated with premature ovarian failure (POF). Mouse models of FSH overexpression exhibit female infertility of unknown ovarian etiology with follicles at various stages of development, lack of corpora lutea, and the presence of large hemorrhagic and fluid-filled cysts (3). POF is a disorder affecting 1% of reproductive-aged women who lose ovarian function before the age of 40 years, and 70% of these women experience this loss from idiopathic causes. POF can occur due to an accelerated loss of follicles, an inability of the remaining follicles to respond to ovulatory signals, an initially diminished ovarian reserve, or any combination thereof. Although it is possible that a high FSH level in POF is solely a consequence of reduced ovarian inhibin level, a high FSH level from any cause can potentially lead to the recruitment of a larger number of follicles into a growing pool in every cycle in younger women, resulting in early depletion (1, 4). In fact, increased levels of FSH in the circulation due to re-trafficking of FSH to the GnRH-regulated secretory pathway or to ectopic overexpression, resulted in significantly more corpora lutea, implicating higher FSH levels in greater follicle recruitment in these mouse models (3). Therefore, studies addressed in recent reviews all demonstrate that proper regulation of FSH concentration and function is critical for female fertility. In this brief commentary, we will summarize recent reviews published in Endocrinology that discuss FSH synthesis and function and point to outstanding questions, answers to which will provide further insight into the physiology and pathophysiology of the reproductive system.

Transcriptional Regulation

The unique FSH β-subunit provides biological specificity and is the limiting factor in the mature FSH synthesis since α-subunit is minimally regulated by hormones and is expressed at the sufficiently high basal level. While concentration of LH in the circulation is regulated at both the level of β-subunit transcription and GnRH-stimulated secretion, FSH concentration is regulated primarily at the level of β-subunit transcription since FSH is constitutively secreted. FSH increases together with LH during the preovulatory surge and exhibits a separate, second increase that is necessary for folliculogenesis in humans and rodents. Changes in FSHβ messenger ribonucleic acid (mRNA) (Fshb) levels precede changes in FSH concentration in the circulation. Transcription of Fshb is induced primarily by activin and GnRH. In mice lacking GnRH or GnRH receptor, FSH levels, as well as LH levels, are 60% to 90% lower in males and females. GnRH injection in rats with low endogenous GnRH increased FSHβ transcription 4-fold, comparable to the changes in FSH concentration in the circulation throughout the cycle. Signaling pathways and molecular mechanisms of this induction are reviewed by Stamatiades et al. (5). Ex vivo studies using primary cells or studies using gonadotrope-derived model cell lines determined roles for calcium-activated pathways, calcineurin and calmodulin-dependent protein kinase II, as well as for extracellularly regulated kinase 1/2 (ERK1/2), mitogen-activated protein kinase (MAPK), and protein kinase A (PKA) pathways in Fshb induction by GnRH. These pathways activate activating protein 1 (AP-1) or cAMP response element-binding protein (CREB) transcription factors that replace repressors Jun Dimerization Protein 2 (JDP-2) or inducible cAMP early repressor (ICER) on the Fshb promoter to activate Fshb transcription. Although whole-body null mice of several of these players confirm their roles in maintaining FSH levels, studies using gonadotrope-specific deletions are either lacking or do not support findings in vitro. Future studies are needed to accurately determine the roles of identified signaling molecules and transcription factors in vivo and to establish mechanisms of GnRH regulation.

Activin is a potent regulator of FSHβ gene expression and was originally identified as a component of ovarian follicular fluid that increased FSHβ synthesis and FSH secretion from pituitary gonadotrope cells. Expression of both intrapituitary follistatin and ovarian inhibin, inhibitors of activin, fluctuate during the estrous cycle in opposition to the levels of FSHβ mRNA, suggesting that bioavailability of activin, through changes in follistatin and/or inhibin levels, is a critical regulatory component of FSHβ synthesis. In particular, the secondary rise of FSH that is necessary for follicular development is dependent on activin since both FSHβ mRNA and FSH levels in the blood can be reduced during the secondary rise in female rats infused with follistatin. Fshb regulation by activin/transforming growth factor-β family of proteins was reviewed by Fortin et al (6), and Ongaro et al (7).

As with GnRH, studies using primary pituitary cells ex vivo or cell lines demonstrated that activins activate Fshb expression through canonical type I and II receptors and the Smad pathway. Smad3 is phosphorylated by the type I receptor, primarily ALK4; dimerizes with Smad4; and translocates to the nucleus where it binds deoxyribonucleic acid (DNA) either directly or via its interactions with Smad4 and/or Foxl2 Forkhead factor. Smad2, which can also be phosphorylated by activin signaling, is not critical for Fshb induction by activin. Although Smad3 links activin receptor to transcriptional activation, pituitary Fshb expression and fertility are unaffected in mice with a targeted deletion of Smad3 in gonadotropes, either alone or in combination with Smad2. Since a truncated Smad3 protein that lacks the DNA-binding domain is expressed in the gonadotrope following gonadotrope-specific Cre cross, Smad3 may regulate Fshb transcription without direct DNA binding. Mice with Smad4 deletion in gonadotropes, on the other hand, are hypogonadal, and females are subfertile with impaired ovarian follicle development. Both male and female Smad4 conditional knockout mice exhibit diminished FSH levels and pituitary Fshb mRNA expression. Deletion of Foxl2 in gonadotropes using a Cre/lox approach also causes reductions in Fshb and subfertility, with females producing smaller and fewer litters. The reproductive phenotype of a double Smad4/Foxl2 conditional knockout mice is more dramatic and resembles FSHβ knockouts: females are infertile and lack estrous cyclicity, with arrested ovarian follicle development at an early antral stage. Their FSH deficiency is also more pronounced than in the single knockout mice. The same signaling pathway may be shared with the bone morphogenetic protein (BMP) subfamily, which stimulates Fshb expression in vitro. In particular, a role for BMP2, BMP4, BMP6, and BMP7 was proposed using cell lines (reviewed in (7)). Although specific receptors for the BMPs, ALK1, ALK2, ALK3, and ALK6 most often phosphorylate Smad1/5/8 instead of Smad2/3, Smad4 is a common dimerization partner. Roles of ALK2 and ALK3 have been examined with gonadotrope-specific deletion, and results demonstrate that they are dispensable for Fshb expression. Since ALK1 and ALK6 expression is limited in gonadotropes, these results brought into question a role for BMPs in FSH regulation. In conclusion, studies demonstrate the necessity for Smad4 and Foxl2 is Fshb expression; however, the roles of receptors (except for ALK2 and ALK3) and specific ligands belonging to activin/transforming growth factor-β family have not been determined. The variable effects of different ligands on Fshb may be a result of the complex interplay and crosstalk with the variety of type I and type II receptors.

Posttranscriptional Regulation, Translation, Glycosylation, and Secretion

With an exception of glycosylation (reviewed in (8)), little is known about posttranslational regulation, sorting, and secretion of FSH. A couple of studies identified several microRNAs, miR-132 and miR-212 in particular, that regulate GnRH-stimulated FSH synthesis and secretion (reviewed in (5)). miR-125b, on the contrary, prevents GnRH-induced FSH synthesis and serves as a repressor. Importantly, in vivo studies are remaining to confirm their roles. Further, potential roles of other noncoding RNA species regulating FSH levels need to be addressed. Additionally, translational control of 2 other gonadotrope-specific genes, Lhb and Gnrhr, is beginning to emerge (reader is referred to the recent commentary by MacNicol et al. (9)), but it is completely unexplored in regard to Fshb.

FSH is an N-linked glycosylated protein, containing 4 N-acetylglucosamine residues (2 on each subunit) linked to an amide group of an asparagine amino acid in the peptide chain (reviewed in (8)). Thus, FSH exhibits macroheterogeneity, resulting from the absence of 1 or more of the N-glycans and microheterogeneity, reflecting different branching of glycans and variable sialic acid content, resulting in size and charge variants. As opposed to LH, where sulfated glycans are more abundant, sialic acid predominates in FSH, although sulfate and phosphate may provide alternative negatively charged residues. Glycosylation is critical for bioactivity due to the necessity of glycosylation for dimerization of α- and β-subunits, subsequent glycohormone secretion, regulation of half-life in the circulation, and binding to its receptor in the gonads. In particular, mutation of either 1 of 2 FSHβ N-glycans significantly increases clearance and reduces in vivo biological activity, whereas mutation of both sites eliminates biological function due to dramatically accelerated clearance. FSH glycosylation changes during puberty, reproductive cycles, and aging imply hormonal regulation of glycosylation. Accordingly, the biological activity of circulating FSH can vary under changing physiological conditions. Interestingly, during reproductive aging in women, the ratio of partially glycosylated, lower molecular weight FSH^21/18 to fully glycosylated FSH²⁴ decreases. Several studies demonstrated that hypoglycosylated FSH^21/18 is more potent in activating the FSH signaling pathway than FSH²⁴ (reviewed in (2) and (10)). The mechanisms underlying carbohydrate modulation of FSH activity are poorly understood. Although advances in mass spectrometry permit characterization of FSH glycan populations, positions, abundance, or linkages remain difficult to assess. However, understanding glycohormone regulation may be critical since variants affect the biological activity of FSH, and glycosylations may vary in recombinant FSH preparations that are used in the clinic.

FSH is largely released constitutively, while LH is released in a pulsatile manner via a regulated secretory pathway following GnRH stimulation. LH contains a carboxy terminal heptapeptide that directs its secretion via the regulated pathway. LH and FSH appear to diverge early in the secretory pathway rather than later in the trans-Golgi network, where most of protein sorting occurs. For that reason, FSH-only granules are relatively rare in the pituitary, while LH-only dense core granules can be observed, although the majority of secretory granules contain both gonadotropins. However, it is still not clear if FSH associates with any chaperone or vesicular proteins in order to be sorted to the constitutive pathway. Questions also remain whether there is any regulation of FSH secretion by hormones such as steroids or activins, separately from their effects on β-subunit transcription.

Receptor Binding and Signaling

FSH binds and activates a 7-transmembrane domain, G-protein–coupled receptor (FSHR) expressed in granulosa cells in ovaries and Sertoli cells in testes (reviewed in Ulloa-Aguirre et al. (10) and Law et al. (11)). In the granulosa cells, FSH regulates the expression of about 3800 genes that mediate proliferation and differentiation and controls the folliculogenesis from the preantral to the preovulatory stage. Such diverse functions are accomplished with complicated signaling pathways that are still not completely elucidated. FSHR can interact, directly or indirectly, with other receptors such as insulin-like growth factor-1 receptor and the epidermal growth factor receptor, or heterodimerize with the LH receptor. Interactions with these other receptors alter or augment signaling pathways, but conditions under which these interactions occur in vivo are not clear. For example, FSHR-activated PKA promotes inactivation of the MAPK phosphatase dual specificity phosphatase DUSP6, which in turn elicits higher ERK phosphorylation by the epidermal growth factor receptor. Crosstalk between LH and FSH receptors may be important for reproductive function, given that during the last stages of follicle development prior to ovulation, both receptors coexist in granulosa cells. FSHR primarily activates the Gs/3',5'-cyclic adenosine monophosphate (cAMP)/PKA pathway leading to FSH-responsive gene induction and sex steroid production. However, although this pathway canonically terminates in the activation of CREB transcription factor, FSH-responsive genes in humans are enriched in binding motifs for the GATA factors and not CREB. FSHR also activates several other signaling pathways via interaction with other G proteins, at least under certain conditions. FSHR stimulates Ca²⁺ mobilization in both granulosa and Sertoli cells via a variety of mechanisms. Besides possible interaction with Gq, since GPCRs can be promiscuous, APPL1 adaptor protein interacting with FSHR leads to calcium mobilization and to the activation of the protein kinase B (PKB)/AKT anti-apoptotic pathway. Alternatively, cAMP activates not only PKA, but exchange protein activated by cAMP (EPAC) as well, which leads to the activation of PKB/AKT and MAPK pathways. Activation of PKB/AKT is involved in proliferative pathways in both Sertoli and granulosa cells and in FSH-mediated protection of granulosa cells from atresia. Subsequent to G protein activation, FSHR itself is phosphorylated by G protein-coupled receptor kinases (GRKs) at the intracellular tail, triggering association with β-arrestins that also leads to ERK1/2–MAPK activation. The ERK-activated mammalian target of rapamycin pathway can induce the expression of FSH-differentiative target genes, such as LH receptor, inhibin-α, and Cyp19 aromatase. Most of these pathways were identified using ex vivo cells, cell lines, or reconstituted heterologous cells, while in vivo confirmation of their roles is unresolved, primarily owing to the lack of mouse models allowing cell-specific deletions. Greater understanding of the FSH-mediated signaling network is of high physiological and clinical importance due to the crucial role for FSH in regulating mammalian reproduction.

Elucidation of FSH signaling pathways could also lead to designing drug candidates that may activate selective signaling pathways in target cells. Several selective, nonpeptide, small molecules have been identified as FSH agonists (reviewed in Nataraja et al. (12),). In addition, selective, nonpeptide antagonists to FSH receptors have been generated that inhibit ovulation in rats. Antagonists have the potential to offer highly selective, nonsteroidal contraception methods with fewer side effects than the currently available steroid-based contraceptives. Of particular interest is the potential for allosteric regulation of the receptor activity, meaning that the regulator interacts at one site to modulate interactions at a spatially distinct site of the same molecule. FSHRs can be modulated at allosteric sites that are away from the orthosteric, ligand-binding sites. This feature of FSHR signaling may allow development of ligands that specifically modulate effectors with desired effects.

FSH Function in Extragonadal Tissues

FSHR is expressed in extragonadal tissues as well, especially in osteoclasts, adipocytes, chondrocytes, benign prostatic hyperplasia and prostate cancer, ovarian cancer, and the placenta. These findings imply a role for FSH in these tissues that has not been studied before, and more importantly, potential FSH effects in pathophysiology in instances with high or low FSH levels.

During reproductive aging, women experience profound reductions in bone mineral density due to increased resorption rates. This phenomenon was considered to be a result of decreased estrogen; however, new results point to the possible effects of increased FSH, which as mentioned above, is elevated in menopause. This phase of a woman’s life is associated with weight increase and visceral adiposity as well as dysregulated energy homeostasis. Since FSHR is expressed in osteoclasts and adipocytes, it was postulated that increased FSH may be responsible for these effects of aging (review in (13) and (14)). In bone, FSH may increase osteoclast formation, function, and survival and thus, increased FSH in peri- and postmenopausal women may contribute to bone loss since FSH can stimulate resorption by osteoclasts. Women with POF exhibit elevated FSH earlier, and regardless of causes of increased FSH as discussed above, they may face not only infertility, but increased risk of osteoporosis and heart disease, which are associated with increased visceral adiposity. Therefore, it is critical to address these outstanding questions.

Furthermore, FSHR overexpression is observed in a variety of endocrine tissue cancers (reviewed in (15)). FSHR was expressed primarily on the endothelial cells in the peripheral tumor blood vessels. The available evidence indicates that epithelial FSHR promotes proliferation, migration, and invasion of ovarian, prostate, and breast cancer cells. Given the propensity for distant metastases, it is important to address a possible role of FSH in angiogenesis or the metastatic potential of these tumors.

Finally, in addition to tumor endothelial cells, FSHR protein was observed in normal vascular endothelium of the fetal vasculature within the chorionic villi, in glandular epithelium of the cervix, in proliferative and secretory endometrium, in muscle fibers and stroma of the myometrium, and in the placenta in pregnancy (reviewed in (16)). FSH in umbilical vein endothelial cells and in osteoclasts acts through a shorter FSHR isoform that lacks the C-terminal domain and does not stimulate cAMP production. This may explain the inability to detect FSHR in these tissues in earlier studies. Analyses of the wildtype and FSHR null fetoplacental units revealed that the FSHR null placentas have fewer fetal vessels, revealing an important role for the FSHR in fetal vessel angiogenesis. This is reminiscent of its role in cancer angiogenesis. Normal placental development is necessary for the healthy pregnancy, and impairment in placental function may result in fetal growth retardation and fetal death.

Although these reports of extragonadal function of FSH are still controversial, they warrant examination using a variety of approaches, potentially tissue-specific deletion of FSHR, to answer emerging questions. Studies reviewed here, moreover, lend support for development of FSH antagonists for a treatment of osteoporosis, obesity, or endocrine tumors, in addition to new agonists for infertility discussed above. Further investigations into functions of FSH beyond gonadal tissues are also critical to evaluate the consequence of dysregulated FSH levels.

Conclusion

Undoubtedly, we came a long way since the first evidence for the pituitary role in gonadal regulation was published in 1910. The pituitary–gonadal relationship as we know it today was described in 1930 (albeit authors called the corresponding hormones different names), and a year later, gonadotropin hormones were extracted from the pituitary. Afterwards, in the mid-seventies, gonadotropin hormone amino acid sequences were determined and 2 decades later genes cloned. Nonetheless, this commentary points to the outstanding questions that need to be answered to gain further insight in FSH function and regulation. Answers to these questions are crucial from the basic science perspective to increase our understanding of the regulation of reproductive function. Moreover, the future of infertility treatment, or other pathologies mentioned above, relies on our ability to fully elucidate the function of this hormone.

Glossary

Abbreviations

BMP

bone morphogenetic protein

cAMP

3',5'-cyclic adenosine monophosphate

DNA

deoxyribonucleic acid

ERK

extracellularly regulated kinase

FSH

follicle-stimulating hormone

FSHR

FSH receptor

GnRH

gonadotropin-releasing hormone

LH

luteinizing hormone

MAPK

mitogen-activated protein kinase

mRNA

messenger ribonucleic acid

PKA

protein kinase A

PKB

protein kinase B

POF

premature ovarian failure

Additional Information

Disclosure Summary: The authors have no conflicts of interest to declare and nothing to disclose.

Data Availability. Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.