FSH is a glycoprotein hormone having disulfide-rich heterodimers, a common α subunit (sharing with TSH and LH), and a unique β subunit. Evolving pieces of evidence suggest that pituitary-derived activins are the primary stimulators of FSH generation by gonadotrope cells. Activins control transcription of the FSH component gene (Fshβ) in vitro via SMAD3, SMAD4, and FOXL2 (22–25). FSH acts on Sc via FSH-R ( Figure 2 ), a G protein-coupled receptor (GPCR), which transmits its signal by recruiting the intracellular GTP binding proteins (G-proteins, either stimulatory Gα_s or inhibitory Gα_i) associated with it (26). Dual coupling of Gα_s or Gα_i to FSH-R differentially modulates the activity of adenylyl cyclase (AC) to regulate FSH-induced cAMP production within Sc (26). The concentration of cAMP subsequently directs the multiple downstream signaling cascades such as canonical Protein Kinase A (PKA) or other (PKC, PI3K, Akt/PKB, and ERK1/ERK2) pathways highlighting the pleiotropic effects of FSH in Sc (26). The robust cAMP response in Sc results in the activation of PKA which in turn phosphorylates cAMP Response Element Binding protein (CREB) to induce the transcription of genes such as Stem cell factor (SCF), Glial cell line-derived neurotrophic factor (Gdnf), Androgen binding protein (Abp), Kruppel-like factor 4 (Klf4), Transferrin etc, that play a critical role in Gc differentiation (6, 26–30).

In rats, FSH-R is first detected at E14.5 [embryonic age in days (E)], whereas the fetal plasma FSH concentration rises from E 19.5- 21, peaks at P5 [post-natal age in days (P)], then substantially drops during P15-20, finally recovered to a steady state by P40-50 (31, 32); similar events occur in mice ( Figures 3A, C ). On the other hand, FSH is uniformly detectable in human fetal circulation from 12-18 week of gestation (WG), peaks during 20-22 WG and then gradually declines in term pregnancy ( Figures 3B, D ) (33, 34), whereas specific binding of FSH is observed in human and rhesus monkey (Macaca mulata) testes during 8–16 and 19–22 WG, respectively (35, 36). In post-natal life, FSH concentration first raises upto the adult range within a week of parturition and stays stable till 4-6 months, then declines and gets undetectable during the juvenile period prior to its re-elevation at puberty (4, 5). Although circulatory FSH levels remain relatively constant in adult men and rats (4, 5), the expression pattern of FSH-R cyclically changes in a stage-specific manner, maximal during stages XIII–II and minimal at VII–VIII (37). FSH has been shown to suppress FSH-R transcription at 6-8 hr (38) in cultured Sc and subsequently gets recovered by FSH at 24-48 hr (39).

In utero life, FSH has been shown to induce Sc proliferation and augments AMH (Anti Müllerian Hormone) production in both rodents (40) and primates (41) and this fetal expansion of the Sc population critically regulates the maximal spermatogenic output in adult testes (42–45). Such FSH-driven Sc proliferation gets continued in neonatal (upto P15) rats and infant primates (upto 3-6 months) and ceases with functional maturation of Sc during pubertal development (27–30). It is interesting to note here that unlike puberty, FSH induced cAMP production is limited during infancy in both rats (27, 28) and rhesus monkeys (29, 30) and therefore Sc fails to support robust Gc differentiation at younger ages despite being exposed to sufficiently high levels of FSH and FSH-R (27–29). Unlike pubertal cells, diminished plasma membrane localization of FSH-R protein in rats (27) and limited expression of Gαs protein in monkeys are considered to be the underlie causes of such poor cAMP response by FSH in infant Sc (29).

In hypophysectomised or GnRH depleted (via pharmacological or immunological inhibition) rats, administrations of FSH alone show partial spermatogenic restoration (46, 47). For example, FSH replacement in GnRH antagonist-treated rats significantly rescues spermatogonia B and early spermatocytes (48). Immuno-neutralization of FSH in post-natal rats indicates FSH promotes Sc proliferation and Gc survival in neonatal age, whereas pre-meiotic Gc differentiation in pubertal age (49). Exogenous administration of FSH alone in pre-pubertal hpg mice fails to induce sperm production (50). Similarly, pituitary independent transgenic expression of human (h) FSH (51) or mutated [at Asp567Gly and constitutively active (capable of FSH independent cAMP production)] h-FSH-R (h-FSH-R*) (52) in male hpg mouse leads to incomplete meiotic progression. Furthermore, although h-FSH-R* over-expression augments proliferation/development of Sc/pre or early meiotic Gc in wild-type testes (53) this hyper-active receptor fails to maintain normal spermatogenesis during experimental deprivation of gonadotropins (54). However, over-expression of h-FSH-R* shows LH-independent steroidogenic activity (55). Notably, over-expression of FSH-Rs [either h-FSH-R* (along with normal h-FSH-R) or another hyper-mutated (at Asp-580-His, constitutively active (capable of FSH independent cAMP productive) mouse (m) FSH-R (m-FSH-R*)] do not affect normal spermatogenic maintenance (55). Finally, both FSH or FSH-R Knock-out (KO) mice demonstrate reduced testis size with reduced numbers of Sc and Gc (spermatogonia, spermatocytes and round spermatids) leading to sub-fertility (56–58) concluding dispensable role of FSH in rodents. However, this dogma has recently been challenged as the expression of hyper-active m-FSH-R* shown to rescue male fertility in LH-Receptor (LH-R) KO mice with a complete absence of testicular androgens (due to exogenous flutamide treatment) (59).

FSH has been shown to be mitogenic for Sc and induce early differentiation in spermatogonia A in rhesus and cynomolgus monkeys (long-tailed macaque; Macaca fascicularis) (15–17). However, five finish men with an inactivating mutation in FSH-R have been reported to have variable degrees of spermatogenic failure without complete loss of fertility (60). In multiple hypogonadotropic hypogonadal clinical studies (61–64) and/or experimentally induced and/or gonadotropin deficient non-human primates (65–68), supplementations of FSH alone (independent of LH/T) results to limited spermatogenic recovery without appearance of either elongated spermatid or spermatozoa. FSH has been shown to regulate the number of pachytene spermatocytes in adult men (69). These reports suggest that like rodents, FSH plays only a supportive role in regulating male fertility in men. However, there are substantial contradictory reports available in men indicating an absolute requirement of FSH for sperm production. For example, hCG-mediated suppression of circulatory FSH in adult men results into poor sperm counts, with one individual developing complete azoospermia, which later gets recovered by FSH supplementation alone (70). Similarly, a hypophysectomized man with complete gonadotropin deficiency fathered three children having h-FSH-R* (71). Finally, complete infertility has been observed in men lacking normal circulating FSH due to mutated FSH-β (72–74). Furthermore, two cases of isolated FSH deficiency with normal FSH-β gene and usual LH/T levels [first, two young men having moderate testicular hypotrophy (75, 76), second, a 19 years old boy being homozygous for a novel silent polymorphism (G/T substitution) in FSH-β promoter (77),] show severe sperm abnormalities to complete azoospermia respectively. Intriguingly, immuno-neutralization of circulatory FSH shows acute spermatogenic abnormalities in both bonnet monkeys (Macaca radiata) (78) and men (79) suggesting FSH vaccination as a promising male contraceptive strategy (80). Taken together, the critical contribution of FSH in regulating primate spermatogenesis is still currently disputed (15, 17, 81, 82).

LH binds to LH-R expressed by interstitial Lc and indirectly exerts its actions on spermatogenesis through T–AR interaction via regulating Sc functions ( Figure 2 ) (6, 82). In rats, fetal plasma LH concentration gets elevated from E 18- 21, then rises at P5-7, further gets reduced during P 20-25, rises again by P35 to peak at P60 and remains constant thereafter throughout adulthood prior to aging (P 400-500) (31, 32). In humans, pituitary LH is measurable from 12-18 WG (which is around 10-fold lower than placental hCG), peaks during 20-22 WG and then gradually decline in term pregnancy ( Figures 3B, D ) (33, 34). However, such a pattern remains inconsistent with the corresponding T profile which peaks during 12-14 WG and then drops during the second trimester corroborating with placental hCG (83). In post-natal life, LH concentration first raises upto the adult range within a week of parturition and then stays stable till 4-6 months, subsequently gets undetectable during the juvenile period, and finally shows the pubertal elevation by reaching its maximal range (4, 5).

Classical histological studies have identified two developmentally diverse populations of Lc e.g.- fetal (FLc) and adult (ALc) (83). FLc originate from coelomic epithelium and notch active Nestin-positive perivascular cells located at the gonad–mesonephros borders, and get specified as Nr5a1 or Ad4BP/SF-1 expressing cells by E 12.5 in fetal mouse testes (84). These cells produce androstenedione (precursor of T, due to lack of HSD17β3 enzyme) and play a critical role in initial virilization and patterning of the male external genitalia (84). However, in neonatal (P 5-15) testis, FLc undergo massive dedifferentiation and during puberty (P 15-21) gradually get replaced by T producing ALc (85, 86). FLc also secretes INSL3, a member of the insulin-relaxin family of peptides that acts on the body through the G-protein-coupled receptor relaxin/insulin-like family peptide receptor 2 (RXFP2). Missense mutations or ablation of Insl3 or Rxfp2 causes cryptorchidism leading to azoospermia (87, 88). However, unlike rodents, primate Lc shows a triphasic developmental pattern (83–86). In human, FLc peak during 12-14 WG (83) and subsequently get dedifferentiated by the end of the second trimester and is replaced by a unique population of neonatal-Lc (NLc) just during/after birth which persist for first 4-6 months of infantile age, when the HHT axis remains active (89). During the onset of juvenile period (inactivation of the HHT axis) massive involution occurs in the NLc population and finally ALc population originates from the dedifferentiating NLc population during puberty (83).

Like FSH-R, LH-R/LHCG-R is also a GPCR that recruits cAMP-dependent PKA pathway to induce the expression and activation of steroidogenic acute regulatory protein (STAR) at the outer mitochondrial membrane of ALc leading to cholesterol trafficking for initiation of steroidogenesis and eventually biosynthesize T (90). However, despite being responsive towards LH signal, FLc of both rodents and primates are independent of fetal LH action (83). FLc number or external genitalia remain unaffected in hpg (13), LH-RKO (91), LH-βKO (92) and ARKO (93, 94) adult male mice suggesting murine FLc are functionally independent of LH or T. In contrast, although patients having LH-β mutations show normal masculinized development (95–99), LHCG-R mutations lead to pseudo-hermaphroditism (100) indicating definite role of hCG on FLc functioning in men. However, in both the species LH is absolutely required for ALc function (83) as evident from various mouse models [hpg (13), LH-RKO (91), LH-βKO (92) and ARKO (93, 94)], etc and mutations in human LH-β/LHCGR genes resulting masculinized fetus but compromised pubertal development and complete azoospermia due to total absence of functional pituitary LH and testicular T (100). It is interesting to note here that fertility can be restored in men with isolated LH deficiency due to mutations in the LHβ gene by long-term hCG supplementations within the critical “window of testicular susceptibility” during pubertal development (101).

Stimulation of LH (resulting T) in rhesus and cynomolgus monkeys leads to spermatogonial differentiation and initiation of Gc meiosis without insignificant rise in Sc number (15, 17, 102–105). LH/hCG (or T) mediated absolute recovery of spermatogenesis has been demonstrated in gonadotropin withdrawal models (either by hypophysectomy or treatment of GnRH receptor antagonist or active immunization against GnRH) in adult rodents (106–111), men (64, 112, 113) and non-human primates (114–118). Exogenous supplementations of T or LH/hCG alone have been shown to induce complete spermatogenesis in immature hpg mice (119, 120) or natural or induced hypogonadal men (121, 122). Genetic ablations of LH-β or LH-R in mice further show cryptorchid testes with spermatogenic arrest and male infertility (91, 92). Human patients having inactivated LHCG-R or LH-β frequently show pseudohermaphroditism and cryptorchidism with Lc hypoplasia and spermatogenic arrest (123–132). Interestingly, a unique homozygous deletion on exon 10 in LHCG-R has been reported in an azoospermic man having normal phenotype with diminished LH signaling (but not towards hCG) indicating higher potency of hCG on ALc (123). In contrast, activating mutations in LH-β or LHCG-R were shown to be associated with precocious puberty and Lc hyperplasia (133–148). Such precocious puberty with Lc hyperplasia followed by infertility has been observed in mice over-expressing hyper-active (Asp582Gly) LH-R (149). However, spermatogenesis has been reported in a man with a splice-mutation (homozygous point mutation G to A at -1 position of intron-10 to exon-11 junction) in LHCG-R with severe loss of T production (150). A more surprising study has been reported in a 43 years old man with a homozygous deletion of nine bases in LHβ gene generating a deletion of amino acids from 10 to 12 (His, Pro, Ile) in the amino-terminal critical for conformational changes leading to undetectable LH (high FSH) with very low T (151). Paradoxically, this isolated LH deficiency case eventually shows sub-optimal but spontaneous spermatogenesis (151). It is important here to note that, despite high (20-100 fold) intra-testicular T (IIT) concentration has been considered to be critical for spermatogenic initiation (152, 153), low levels of T are sufficient to drive spermatogenic maintenance as evident by spontaneous spermatogenesis in LH-RKO mice at 12 months of age (154).

LH operates spermatogenic regulations through testicular androgen T and AR (155). T is essential for suppression of AMH (156, 157), pubertal maturation of testicular somatic cells (e.g.- PTc, Sc, Lc in developmental order) (2), the establishment of Blood-testis barrier (BTB) (158), meiotic progression of Gc and spermiation (159). The free titer of T depends upon the extent of the presence of sex hormone-binding globulin (SHBG) which binds to T with strong affinity; thus, SBHG regulates the process of spermatogenesis by controlling the serum concentration of biologically active T (160, 161). The absolute requirement of T on male fertility has been confirmed from ARKO (ubiquitously lacking AR) mice (93, 94). Despite most of the somatic testicular cells (Sc, PTc, Lc etc) express AR, Gc do not have functional AR (2, 3). Cell-specific selective ablation of AR [Sc specific i.e. SCARKO (162–164), Lc specific i.e. LcARKO (165, 166), PTc specific i.e. PTARKO (167, 168) or Gc specific i.e. GcARKO (169, 170)] demonstrated that AR expressed by Sc plays a pivotal role in the progression of Gc meiosis (20, 21, 155). Furthermore, the crossing of hpg mice with ARKO or SCARKO mice followed by T/5α- dihydrotestosterone (DHT) supplementation confirmed the critical significance of Sc-mediated AR signaling in spermatogenesis (171). The transition of round to elongated spermatid is fully dependent on T action transmitted via Sc (159).

In Sc, AR signals via both classical and non-classical manner (155). In the classical pathway, T (or 5α-DHT) activated AR binds to specific DNA sequences having Androgen Response Elements (ARE) and initiates the androgen-dependent transcriptional events e.g. Rhox5 expression (155). However, in a non-classical pathway, T gets coupled with membrane-bound AR and triggers the binding of the proline-rich region of AR with the SH₃ domain of membrane bound SRC kinase leading to stimulation of EGF receptor and subsequently activates MAP (RAF, MEK, ERK) kinase or CREB cascade inducing several genes which lack typical AREs on their promoters e.g. Ldha, Claudin11, etc (155). In vitro studies show that T regulates spermiation via a non-classical pathway (155), however, in vivo studies suggest that classical pathway is most crucial for meiotic completion of Gc and fertility (159).