For qualitative and quantitative normal spermatogenesis an intact hypothalamic-pituitary-gonadal (HPG) axis is essential. Gonadotropin releasing hormone (GnRH) is released by the hypothalamus. In turn, GnRH stimulates the pituitary to secrete LH and FSH. LH stimulates production of testosterone in Leydig cells, which negatively feeds back to the pituitary as well as the hypothalamus in order to modulate the production of GnRH and by this gonadotropin levels (13). FSH synthesis and secretion depends on slow GnRH pulses (every 2–4 h), while rapid GnRH pulses (every 30 min) lead to preferential secretion of LH (14).

During the “mini puberty” (12–18 months of age) the HPG axis is activated for normal genital development, later on during development the HPG axis is again activated with the onset of puberty. During prenatal and prepubertal stage, FSH stimulates Sertoli cell proliferation and by this determines their final number and subsequently testicular size. The proliferation and functional maturation of Sertoli cells is controlled and terminated by thyroid-stimulating hormone (TSH) (15).

Sertoli cells (SCs) are part of the seminiferous tubules of the testes and play a key role in spermatogenesis. They are “nurse cells” as they provide nutritional support for germ cell development. Moreover, they contribute to the spermatogonial stem cell (SSC) niche and are this way indispensable for functional spermatogenesis (16). Sertoli cells and their metabolism are regulated by hormones (17). Some hormone receptors are solely expressed in Sertoli cells which underlines the importance of hormonal signaling for spermatogenesis (17). Thus, Sertoli cells transduce endocrine signals into the paracrine regulation of germ cells (16).

Busch et al. showed that boys with the genotype FSHB c.-211GT/TT and FSHR c.-29AA entered puberty later, which indicates that the overall endocrine network as well as FSH action might be affected in early phases by SNPs (18). In the adult stage, proliferation is ceased in mature Sertoli cell and FSH stimulates the proliferation of spermatogonia (19). In humans FSH mainly regulates sperm output of the seminiferous epithelium by controlling the expansion of premeiotic germ cells (20). FSH also influences the proliferation of type A spermatogonia upregulating nerve growth factor inducible gene B (NGFI-B, also known as Nur77) which increases the expression of glial cell line-derived neurotrophic factor (GDNF) in SCs (21). GDNF in turn supports the proliferation of germinal stem cells (GSCs) and other undifferentiated spermatogonia (22).

The FSH receptor (FSHR) belongs to the 7 transmembrane domains receptor (7TMR) family of G-protein coupled receptors and is only expressed in Sertoli cells (23). FSH binding induces a conformational change of the FSHR especially within the TM domains 5 and 6 which cause intracellularly the dissociation of α- and βγ- subunits of G protein heterotrimer inside the cell. Subsequently, the α-subunit binds to and triggers adenylyl cyclase, which leads to an increase of cAMP levels (24). The main signal transduction pathway for the FSHR is the cAMP-PKA pathway. Its activation leads to a release of the catalytic subunit of protein kinase A (PKA); followed by phosphorylation of enzymes and proteins. Moreover, it targets the cAMP response element binding protein (CREB) which activates transcription of FSH-dependent genes (25). The MAP kinase cascade and extracellular-signal regulated kinase (ERK) get activated most likely via cAMP interactions with guanine nucleotide exchange factors (GEFs) and activation of Ras-like G proteins. By GEFs the phosphatidylinositol 3-kinase (PI3-K) pathway gets activated which leads to an activation of protein kinase B (PBK) (26). The PI3-K pathway plays an important role as it regulates several biological processes e.g., glucose uptake, oxidative burst and mitogenesis (17). Moreover, FSH causes an increase in intracellular calcium mediated by cAMP. Elevated calcium concentrations cause an activation of calmodulin and CaM kinases which result in downstream effects including the phosphorylation of CREB. FSH inducts phospholipase A2 (PLA2) and the release of arachidonic acid (AA) and the activation of eicosanoids (26). These different pathways activate different transcription factors thereby stimulating the transcription of FSH-targeted genes (27). Consequently, Sertoli cells transduce signals from FSH into production of necessary factors for germ cell nutrition and differentiation.

FSH is a pituitary derived heterodimeric glycoprotein which consists of an alpha-subunit and a unique beta-subunit that determines biological specificity and provides specificity for receptor binding (14). The human FSHB gene (National Center for Biotechnology Information (NCBI) database https://www.ncbi.nlm.nih.gov/gene/2488; GeneID:2488; Locus tag:HGNC:3964) is located on chromosome 11p13 and consist of 3 exons (Figure 2A) (30). It encodes the FSH beta-subunit consisting of an 18-amino acid (aa) signal peptide and the 111-aa mature protein (31). The NCBI SNP database (https://www.ncbi.nlm.nih.gov/SNP/) lists 1380 SNPs in the gene region of FSHB, 114 SNPs in coding regions. Only very few SNPs have proven clinical relevance. One of them, SNP rs10835638 (c.-211G>T) is located in the 5′untranslated region within an evolutionary conserved element of the FSHB promotor (Table 1; Figure 2A), which leads to an influence on gene transcription (33). This SNP impairs LHX3 binding and induction of FSHB transcription. Thus, the SNP rs10835638 reveals significant functional importance (33). The regulation of FSHB transcription in gonadotropic cells is essential as the amount of the beta-subunit being transcribed is the rate limiting step in FSHB synthesis and defines how much mature hormone will eventually be secreted, as the alpha subunit is shared with LH, Thyroid-stimulating hormone (TSH) and human Chorionic Gonadotropin (hCG), and produced in excess. Hence, FSHB transcription is directly associated with its translation, secretion and serum levels (14).

The 76 kDa FSH receptor (FSHR) is a G protein-coupled receptor, which belongs to the rhodopsin-like receptor subfamily, and consists of 695 amino acids (34). The FSHR gene (NCBI database https://www.ncbi.nlm.nih.gov/gene/2492; GeneID:2492; Locus tag:HGNC:3969) is located on chromosome 2 p21-p16 and consists of 10 exons of which the first nine exons encode the extracellular amino-terminal domain of the receptor (34, 35). Exon 10 encodes the transmembrane and intracellular portions of the protein (Figure 2B) (36). Two human isoforms are known, one containing all exons (NM_000145) and the second one lacking exon 6 (NM_181446) (13). The extracellular domain contains a stretch of leucine-rich repeats essential for FSH binding and is encoded by exon 2-9 (Figure 2C) (36). The NCBI SNP database (https://www.ncbi.nlm.nih.gov/SNP/) lists 51.677 SNPs in the gene region of FSHR, 779 SNPs are located in coding regions. The following SNPs: rs6166 (c.2039A > G, p.N680S) and rs6165 (c.919A > G, p.T307A) (Table 1), which are located in exon 10 have been analyzed more thoroughly and are in linkage disequilibrium (LD). The SNP c.919A>G results in an amino acid exchange removing a potential O-linked glycosylation site in the hinge region of the receptor, the SNP c.2039A>G also results in an amino acid exchange causing a potential phosphorylation site in the intracellular domain of the receptor (37, 38). Other known SNPs rs1394205 (c.-29G>A) and rs115357990 [c.-114T>C; MAF 0.0126 (1000 Genomes Project)] are located in the promotor region of the FSHR gene (39).

Interestingly, data on the putative impact of the several SNPs within the FSHB and FSHR genes have been mainly and firstly obtained from clinical studies and only later in part paralleled by experimental studies. There are only few in vitro studies available concerning the effect of FSHB and FSHR SNPs on FSH action in the male. This is mainly due to the fact that appropriate read-out systems for studying FSH function and corresponding SNPs functions do not exist. There are neither human gonadotropic nor Sertoli cell lines available. While it is generally believed that mouse gonadotropic cell lines such as the LßT2 cell line are useful and informative for the human as well, the situation for Sertoli cells is worse. The commercial available “Sertoli” cell lines are closer to peritubular cells than to Sertoli cells and therefore of only limited usefulness. Moreover, immortalized Sertoli cells tend to lose their intrinsic FSH receptor expression, making it difficult to study FSH action (40). Some groups have therefore stably transfected Sertoli cell lines with the FSH receptor to regain FSH sensitivity (41). The characteristic feature of losing FSH receptor during immortalization of testicular somatic cells can also be observed for the female pendant of Sertoli cells, the granulosa cells. One of the reasons for the shut-down of FSH receptor expression might be aberrant methylation of regulatory elements controlling receptor expression (42).

Usage of primary cells as a substitute for lacking immortalized cell lines is at least for adult human Sertoli cells also not an option, since there is currently no protocol available which allows to isolate intact Sertoli cells in sufficient amounts. This is mainly due to the fact that the Sertoli germ cell niche is so tightly interlinked, that most cell separation protocols lead to the destruction of these cells. Thus, there is a great need for cell model systems which would allow studying the impact of SNPs and mutations in the male.

Nevertheless, in the mouse gonadotropic cell line (LßT2) it could be shown that the FSHB c.-211G>T polymorphism, located on the promoter region has an effect on the binding of the LHX3 homeodomain transcription factor, which leads to an impaired binding and this way to a 50% decrease in transcriptional promotor activity (33).

The SNP in the promotor region of FSHR c.-29G>A, currently used in a number of FSH studies, was analyzed in murine Sertoli cells SK11, but a significant effect could not be shown (39). Conflicting to this result, another group showed that the FSHR c.-29G>A decreases transcriptional promoter activity by 56% for the A allele in vitro in Chinese hamster ovary (CHO) cells (43). While these results await further confirmation by additional experiments, one should note that there are more SNPs allocated in the core promotor region of the FSHR gene such as a highly variable oligo-A stretch and a SNP at position at c.-114T>C. Therefore, a valid investigation on the impact of the c.-29G>A SNP should include the other polymorphic sites as well (39).

The impact of the SNPs in exon 10 in FSHR was analyzed in vivo showing that a decreased activity of the FSHR has a clinical implication for female infertility (44–46) and in vitro (38, 47): Nordhoff and colleagues analyzed the increase of cAMP and Estradiol after FSH stimulation between the variants NN and SS in human granulosa cells and could not show a difference in the mentioned parameters (47). Simoni and Casarini also used human granulosa-lutein cells (hGLC), but analyzed the kinetics of different signal transduction pathways. The p.Asn680Ser FSHR variant led to decreased ERK1/2 activation and the FSHR seemed to be less active (38). Again, no data on the impact of these SNPs on Sertoli cells are available. Hence, there are some in vitro studies demonstrating the effect of SNP in FSHB/FSHR granulosa cells/cell lines, but more in vitro studies targeting the molecular impact of the SNPs on FSH signaling using Sertoli cells are needed.

The essential role of FSH for spermatogenesis is underlined by the identification and clinical characterization of inactivating FSHB mutations. Until now five men with such FSHB mutations have been described which all showed azoospermia and either very low or absence of FSH serum levels (31). Similar to this, several mutations and variants of the FSHR have already been described in humans (48–50). The first description of a human inactivating FSHR mutation (p.Ala189 Val) showed elevated FSH levels and abnormal sperm parameters, but no azoospermia in affected men (51). An activating FSHR mutation (p.Asp567Gly) was discovered by Gromoll et al. describing a hypophysectomized man who fathered children being under Testosterone treatment only (52). So far, there have been 11 inactivating - and 7 activating mutations of the FSHR described in men and women (53).

Along the findings of completely abolishing FSH action by inactivating mutation, minor genetic changes such as SNPs affecting FSHB transcription, FSH binding properties or FSH receptor sensitivity could impact male fertility too. In the FSHB gene, the SNP c.-211G>T (rs10835638) has a major effect on serum FSH concentration in men (30). This effect as well as reduced testis size, reduced sperm concentration and lower serum Inhibin B and Testosterone was also shown in a cohort of Italian (54), Baltic (55), and German men (56).

Tüttelmann et al. analyzed the combined effect of a SNP in FSHB and FSHR on male reproductive parameters. They found a marked dominant effect of FSHB c.-211G>T in combination with FSHR c.2039A>G on serum FSH and testicular volume. The T allele carriers of the SNP c.-211G>T showed reduced FSH, increased LH, lower testicular volume, lower sperm count and concentrations in comparison to GG homozygotes men (Figure 3) (56). A recent study on the impact of FSHB c.-211G>T in Italian men from Tamburino et al. described decreased FSH, LH, Testosterone, sperm count and testicular volumes in men with GT or TT in comparison to GG (57). Additionally, two population-based studies analyzed the effect of FSHR c.-29G>A in a Baltic and an Italian men cohort, respectively. Grigorova et al. showed an association between this SNP and FSH levels and Tamburino et al. showed that the FSHR c.-29G>A SNP is associated with higher FSH and LH in normozoospermic men. This effect could only be observed in normozoospermic men but not in men with alterations in conventional sperm parameters (58, 59).

The impact of the total phenotypic variance (SNP) was evaluated for the first time in a cohort of young Baltic men by Grigorova et al. (58). The FSHB –c.211G>T in combination with the FSHR c.-29G>A and the FSHR c.2039A>G explained 2,3% of serum FSH variance in young men as well as 1,4% serum Inhibin B, 1% Testosterone and 1,1% total testes volume. Additionally, in a cohort of infertile Estonian men the SNP combination affected 2,3% of serum FSH variance, 2,6% serum Inhibin B and 2% total testes volume (58). In a recent meta-analysis Wu et al. investigated the effect of four SNPs (FSHB c.-211G>T, FSHR c.-29G>A, FSHR c.919A>G, FSHR c.2039A>G) on male infertility. It seems that the combination of three SNP genotypes of FSHR (FSHR c.-29G, c.919A, c.2039A) results in protection against male sterility than either one alone (60).

In the past multiple clinical studies were carried out on idiopathic infertile men receiving FSH therapy to increase birth and pregnancy rates in the couple. However, results were conflicting. Therefore, a Cochrane Review and a recent meta-analysis were carried out to further elucidate these diverse results (7, 61).

Attia and colleagues analyzed 6 randomized controlled trials (RCTs) in which application of FSH was compared to placebo or no treatment at all. The main results comprise a significant increase in spontaneous pregnancy rate and in live birth rate in couples with FSH treatment in the male. The authors critically conclude that these results are promising, but since number of RCTs and participants is small and the evidence is low, no final clinical conclusions can be drawn (7).

In the most recent meta-analysis Santi et al. evaluated similar endpoints of FSH treatment outcome in idiopathic infertile men in 15 studies. The key finding of this analysis, supporting the prior Cochrane analysis, was the significant improvement in pregnancy rate in FSH-treated men (61). Additionally, an increased pregnancy rate was observed after ART when applying FSH. The pregnancies achieved were independent of FSH preparation and duration of therapy. Interestingly, in the sub-analysis considering the treatment response in terms of sperm parameters, a significant increase was only seen in sperm concentration, whereas other semen parameters did not change.

These are in principle promising results for FSH treatment of idiopathic infertile males, however the most susceptible parameters like FSH serum level or basal sperm count do not contribute to distinguish patients who will benefit from treatment from those who won't; predictive markers are therefore eagerly warranted (61) and a pharmacogenetic approach was suggested by several groups (3, 38).

To our knowledge, there are currently only 4 studies (54, 62–64), that chose a FSH-based pharmacogenetic approach (Table 2). Our literature search was conducted in MEDLINE (PubMed), the Cochrane Library, Scopus and UpToDate. Search was last updated in October 2018.

Records for: FSH/male infertility (n = 2245); FSH/male infertility/gene (n = 292); FSHB gene/male infertility/SNP (n = 23); FSH/male infertility/pharmacogenetics (n = 9). The four pharmacogenetics studies identified via the literature search are being discussed in terms of study design, selection criteria, treatment, endpoints and results.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.