Steroidogenic factor 1 (SF-1), also called Adrenal 4 binding protein (Ad4BP), was originally discovered in the early 1990s by Drs. Parker and Morohashi as a regulator of steroid hydroxylases in adrenal glands (Lala et al. 1992; Morohashi et al. 1992). It is a transcription factor, member of the nuclear receptors superfamily. Official designation of the gene encoding for SF-1 is NR5a1, although in the literature the name SF-1 is still predominantly used, and therefore, we will use this designation throughout this chapter.
Numerous studies have shown wide role of SF-1 in the endocrine development and function beyond regulation of steroid hydroxylases. In particularly, studies with SF-1 knockout (SF-1 KO) mice have shown the importance of SF-1 for adrenal and gonadal development, development of the ventromedial nucleus of the hypothalamus (VMH), and spleen and pituitary gonadotropin-producing (gonadotrope) cells. Different studies have revealed that SF-1 does not regulate only the expression of genes encoding for steroid hydroxylases, but many other genes such as antimullerian hormone (AMH), gonadotropin-releasing hormone (GnRH) receptor, common alpha subunit of luteinizing hormone (LH) and follicle-stimulating hormone (FSH), brain-derived neurotrophic factor (BDNF), and many others (reviewed in Schimmer and White 2010).
Regulation of SF-1 Expression
In adult tissues, SF-1 is primarily expressed in steroidogenic cells such as adrenal cortex, testicular Leydig cells, ovarian theca, and granulosa cells and at lower levels in corpus luteum. However, SF-1 is also expressed at some other, non-steroidogenic cells and tissues such as testicular Sertoli cells, VMH, pituitary gonadotrope cells, and spleen. In the fetal tissues, SF-1 is initially expressed in the adrenal and gonadal primordia and serves as an early marker for developing gonads and adrenal glands. SF-1 is also expressed in the fetal spleen and in the prosencephalon, a region of developing brain that includes the hypothalamic primordium (Morohashi et al. 1995; Majdic and Saunders 1996; Hanley et al. 2000a).
Transcriptional Regulation of Expression
Regulation of SF-1 expression is not completely understood, yet. Proximal promoter of SF-1 gene contains SOX binding site, E-box, Sp1/Sp3 site, and CCAAT box (Nomura et al. 1995; Scherrer et al. 2002; Shen and Ingraham 2002). Further upstream and downstream, there are binding sites for other transcriptional regulators such as GATA-4, WT1, and Lhx9, and studies in vitro have shown a potential role of SOX15, SOX30, TEAD-4, and CBX2 proteins in the regulation of SF-1 expression. It is believed that variety of binding sites and transcriptional factors binding to the SF-1 promoter regions are needed to ensure tissue specificity of SF-1 expression (Schimmer and White 2010). However, in vivo studies have shown that these promoter regions are not sufficient to drive the full expression of SF-1 in all tissues that physiologically express this gene. A very large fragment of genomic DNA, flanking the Sf-1 gene is needed to recapitulate endogenous expression of SF-1 in GFP transgenic mice. This, about 5 KB long fragment extending from exon 2 of the Sf-1 gene into the upstream gene NR6A1 (germ cell nuclear factor, GCNF), is able to induce gene expression in Leydig cells, theca cells, VMH, and spleen, but not in the pituitary and corpus luteum (Stallings et al. 2002). Only large rat genomic fragment extending well into the NR6A1 and PSMB17 genes on either side of the Sf-1 gene duplicates patterns of Sf-1 endogenous expression (Karpova et al. 2005). Within this fragment, two conserved enhancer regions were identified that are required for the expression of Sf-1 in the VMH (Shima et al. 2005) and pituitary (Shima et al. 2008). In addition to specific enhancers for VMH and pituitary, special enhancer has been identified that drives the expression of SF-1 in the fetal adrenal gland. This enhancer is found within the intron 4 of the SF-1 gene and is necessary to sustain the expression of SF-1 in the fetal adrenal, but is insufficient to drive the expression of SF-1 in adult adrenal, suggesting that regulation in adult and fetal adrenal gland is differentially regulated (Zubair et al. 2006).
Epigenetic Regulation of SF-1 Expression
Several studies have shown the role of epigenetic regulation through the methylation of different sites within SF-1 regulatory regions playing a role in the expression of SF-1. The basal promoter of SF-1 contains 14 and 10 CpG islands in human and mouse, respectively. These CpG sites are spread across 250 bp and thus do not represent a classical CpG islands, but rather a weak CpG island, such that are frequently found in genes with high tissue specificity in their expression. Methylation pattern of SF-1 basal promoter is in nearly perfect correlation with the levels of expression with the exception of the pituitary, where basal promoter is hypermethylated, what is in agreement with the finding that expression in the pituitary is driven from alternative promoter.
Enhancers that drive tissue-specific expression of SF-1 in different tissues such as VMH, pituitary gonadotrope cells, and fetal adrenal all contain CpG sites, and methylation of these CpGs does follow expression patterns of SF-1 gene. Pituitary enhancer is thus hypomethylated in the pituitary gonadotrope cells and hypermethylated in the VMH neurons and adrenal cortex. Similarly, VMH enhancer is hypomethylated in VMH neurons but hypermethylated in adrenal glands and other tissues that do not use this enhancer for tissue-specific expression of SF-1. However, fetal adrenal specific enhancer is hypomethylated in fetal adrenal cells and remains hypomethylated in adult adrenals, despite solid evidence that this enhancer is not needed for adult adrenal expression (Hoivik et al. 2013).
Fetal Development of Gonads
During development, SF-1 expression is first detected in urogenital ridges of both sexes in murine fetuses at E9.5 (Ikeda et al. 1994; Morohashi et al. 1995) and in human fetuses at 32 days post-ovulation (Hanley et al. 2000b). In mice, total ablation of Sf-1 gene causes regression of the gonadal ridge by E11.5–12.0. The formation of the gonadal ridge is regulated also by other genes such as Emx2, Gata4, Wt1, Lhx9, Cbx, Pod1, Six1, and Six4 most of which are believed to act upstream of the Sf-1. Therefore it is not surprising that promotor region of Sf-1 gene contains binding sites for many of these genes and that impaired gonadal formation in mutant mouse fetuses for several of these genes is associated with downregulation or ectopic upregulation of the SF-1 (reviewed by Tanaka and Nishinakamura 2014).
Shift from the bipotential gonadal ridge to the testis happens in male mice fetuses around E10.5, triggered by the expression of Sry, which induces differentiation of somatic cells into Sertoli cells. SF-1 is proposed to be one of the initial upstream regulators of the Sry gene. A few hours later the control of the differentiation process is switched from Sry to Sox9. In Sertoli cells SF-1, together with Sox9, induces activation of the gene for AMH (Arango et al. 1999) and activates Cyp26b1, an enzyme that catalyzes degradation of the retinoic acid. AMH promotes the regression of Mullerian ducts, and the absence of retinoic acid prevents male germ cells from entering meiosis prematurely. In mice fetal Leydig cells appear in the interstitium around E12.5 and persist in the testis until approximately one week after birth, when adult fetal Leydig cells develop (O’Shaughnessy et al. 2006). SF-1 is expressed in fetal Leydig cells and remains expressed in adult Leydig cells, as it is needed for the production of androgens both during fetal development and in adulthood by regulating the expression of several steroidogenic enzymes (Parker and Schimmer 1997; Ikeda et al. 2001). It is still not known to what extend SF-1 regulates Leydig cell differentiation. While loss of one Sf-1 allele causes delayed Sertoli and Leydig cell differentiation, it does not impact the final number of adult Leydig cells (Luo et al. 1994; Park et al. 2005). However, conditional deletion of Sf-1 in testicular somatic cells result in the absence of fetal and adult Leydig and Sertoli cells in testes, but the absence of Leydig cells in this model could also be due to the absence of Sertoli cells (Jeyasuria et al. 2004). However, an ectopic upregulation of SF-1 in Pod1 lacZ/lacZ mice leads to remarkable increase in the number of fetal Leydig cells (Cui et al. 2004), suggesting direct involvement of SF-1 in the process of fetal Leydig cell formation.
In female mice fetuses, expression of SF-1 follows the same pattern as in males until E11.5–12.0. At E12.0 SF-1 transcripts are no longer detected in mouse developing ovaries, and its expression is switch on again just before birth (Ikeda et al. 1994; Morohashi et al. 1995; Majdic and Saunders 1996). In human ovaries downregulation of SF-1 is less striking. At 52 days post-ovulation, SF-1 remains broadly distributed through the ovaries, with stronger localization in the germinal epithelial layers, and SF-1 expression in the human fetus persists through the entire fetal development (Hanley et al. 2000b) (Fig. 2).
Fetal Development of Adrenals
Gene dosage seems to play important role in the regulation of adrenal development by SF-1. In mice, heterozygous for disrupted Sf-1 allele (SF-1+/− mice), adrenals show a selective decrease in number of adrenal precursor cells at E10.0. Although by E13.5 cell proliferation increases, adrenals do not develop normally and are reduced in size in newborn mice (Bland et al. 2000a, 2004). While SF-1 haploinsufficiency results in the adrenal hypoplasia, overexpression of SF-1 in mice causes increased adrenal size and formation of ectopic adrenal cortical tissues in the thorax (Zubair et al. 2009).
Adult Steroidogenic Organs; Gonads and Adrenal Glands
In gonads and adrenal glands, SF-1 is a key regulator of steroid hormone biosynthesis. Control of steroid production already starts at the level of cholesterol transport. SF-1 increases the expression of the scavenger receptor B1 (SR-B1), important for cellular import of high-density lipoprotein cholesterol, and steroidogenic acute regulatory protein (StAR), which regulates transport of cholesterol from outer to inner mitochondrial membrane. SF-1 also regulates the expression of different cytochrome P450 steroid hydroxylases and other steroidogenic enzymes such as 3β-Hydroxysteroid dehydrogenase (HSD3B2 and HSD3B1) (Schimmer and White 2010). Transcription of enzymes involved in the steroid production in the adrenal cortex is primarily regulated by adrenocorticotropic hormone (ACTH) from the pituitary. After ACTH binds to its cognate receptor, the melanocortin 2 receptor (MC2-R), numerous signaling pathways are activated in adrenal cells and at least some of these pathways are believed to be regulated by SF-1 (Sewer and Waterman 2002). For example, activation of cAMP/PKA pathway, activated by ACTH, decreases the amount of sphingosine and lysosphingomyelin bound to SF-1. Both phospholipids act on SF-1 as inhibitory ligands, and reduced amount of both ligands increases the expression of P45017a -hydroxylase/17,20-lyase (CYP17A1), mediated by SF-1 (Urs et al. 2006).
In testicular Leydig cells, SF-1 regulates both basal and LH-stimulated production of testosterone through regulation of cytochrome P450 cholesterol side chain cleavage (CYP11A1), CYP17A1, HSD3B2, and also 17β-hydroxysteroid dehydrogenase (HSD17B) enzymes (Schimmer and White 2010).
In ovaries, SF-1 is expressed in theca, granulosa, and luteal cells and has a role in production of estrogens during follicular phase, as well as in production of progesterone during luteal phase (Saxena et al. 2007).
Although initial studies reported that there is no deleterious phenotype in SF-1 +/− mice, subsequent studies did show the importance of gene dosage and that both alleles of Sf-1 gene are needed for normal development and function of adrenals (Bland et al. 2000a, b, 2004). Adrenal glands in SF-1 +/− mice are smaller, with normal zonation but marked cellular hypertrophy within the zona fasciculata and striking dilatation of the cortical vasculature. In SF-1 +/− mice stimulated plasma corticosterone levels are lower in comparison to WT mice, evening levels of ACTH are higher, and expression of StAR and MC2-R is increased, while, interestingly, expression of CYP11A1 expression is unchanged (Bland et al. 2000b).
SF-1 is expressed also in the anterior pituitary, specifically in gonadotrope cells (Ingraham et al. 1994; Shinoda et al. 1995). Global and pituitary-specific SF-1 KO mice have markedly diminished expression of LH, FSH, and receptor for GnRH suggesting that SF-1 regulates expression of these pituitary genes (Ingraham et al. 1994; Shinoda et al. 1995). This was confirmed by in vitro studies, demonstrating that SF-1 indeed regulates the expression of common α-subunit (αGSU) of LH and FSH, as well as β-subunits of both FSH and LH (βFSH and βLH), and GnRH receptor (Barnhart and Mellon 1994; Ingraham et al. 1994; Halvorson et al. 1996; Jacobs et al. 2003). However, it is not yet clear whether reduced expression of both LH and FSH in SF-1 KO mice is due to direct effect of missing Sf-1 gene or perhaps secondary due to diminished expression of GnRH receptor since expression of both LH and FSH is strongly influenced by GnRH, secreted from the hypothalamus. This is supported by finding that the treatment of SF-1 KO mice with supraphysiologic doses of GnRH stimulates LH and FSH production in the pituitary in SF-1 KO mice (Ikeda et al. 1995) indicating that gonadotropin cells are capable of gonadotropin synthesis in the absence of SF-1, although this was done only qualitatively (by immunohistochemistry and levels of expression were not determined).
Ventromedial Nucleus of the Hypothalamus
In the central nervous system, the VMH is the only part where SF-1 is expressed. In the mouse its expression has been detected in the hypothalamic primordium as early as embryonic day E9.0. Initial studies showed that target disruption of the gene encoding SF-1 impaired the cytoarchitecture of the VMH (Ikeda et al. 1995; Luo et al. 1995; Shinoda et al. 1995). Subsequent studies on SF-1 KO fetuses (E15) and neonates revealed that distribution of cells is altered significantly in the region of the developing VMH, but without significantly affecting cell numbers up until birth, suggesting that disrupted neuronal migration could be the main cause of disorganized VMH in SF-1 KO mice (Davis et al. 2004). SF-1 could be also required for terminal differentiation of the VMH and for expression of BDNF since SF-1 KO neonates show complete loss of projections to the bed nucleus of stria terminalis (BNST) and amygdala, and neuronal precursors do not express BDNF (Tran et al. 2003). Similar findings were reported in CNS-specific SF-1 KO mice, which have also diminished hypothalamic expression of BDNF, CRH type 2 receptor, and cannabinoid receptor 1 (CBR1) (Davis et al. 2004; Kim et al. 2008; Zhao et al. 2008).
In adult life, SF-1 KO mice develop obesity, caused mainly by hypoactivity rather than hyperphagia, suggesting the role of VMH in the regulation of spontaneous physical activity (Majdic et al. 2002). Expression of several orexigenic and anorexigenic peptides was studied in these mice, and although changes in their expression pattern were observed, this was most likely due to disrupted neuronal migration in the VMH region and might not be functionally connected with obese phenotype (Budefeld et al. 2011). Diminished running wheel activity was observed also in the CNS-specific SF-1 KO mice, although these mice only develop obesity when on high-fat diet, probably reflecting the difference between early and later disruption of the VMH development. In addition to the obesity when on high-fat diet, CNS-specific SF-1 KO mice have impaired thermogenesis after acute exposure to high-fat diet and have decreased expression of leptin receptors (LepR) in the VMH (Zhao et al. 2008; Kim et al. 2011).
VMH is also involved in the regulation of different social behaviors such as anxiety, aggression, defensive, maternal and female sexual behavior. CNS-specific SF-1 KO mice showed elevated anxiety (Zhao et al. 2008), impaired female sexual behavior – lordosis and subfertility with the decreased number of corpora lutea (Kim et al. 2010). Lordosis is also impaired in hormone primed agonadal SF-1 KO mice (Grgurevic et al. 2012), while in the absence of any hormones, these mice show elevated aggression toward receptive females (Grgurevic et al. 2008). In our laboratory we also noted elevated anxiety-like behavior and poorer parental behavior in global SF-1 KO mice when compared to WT mice (unpublished observations – Grgurevic), and our recent study shows that SF-1+/− females have impaired maternal behavior (Spanic et al. 2016). Study by DaSilva et al. (Silva et al. 2013) has also shown the involvement of SF-1 neurons in the dorsolateral VMH in fear response, although in this study SF-1 was only used as a marker for the VMH neurons and direct role of SF-1 protein in the regulation of fear response was not studied.
Although numerous behavioral effects have been observed in SF-1 KO mice, it is not known to what extent SF-1 directly influence these behavioral traits as many of those could arise due to impaired VMH structure, and SF-1 role might only be important during development of this nucleus. However, SF-1 might also have a direct role in the regulation of some of these behaviors as studies have shown the role of SF-1 in the regulation of expression of genes encoding for BDNF, CBR1, CRH type 2 receptor, and LepR (Schimmer and White 2010).
SF-1 in Human Disease
Based on in vitro and in vivo studies of SF-1, the first attempts to expose the potential roles of SF-1 in human pathology were focused on adrenal insufficiency and complete gonadal dysgenesis, a syndrome similar to SF-1 KO mice phenotype. This phenotype is very rare in humans; nevertheless, in 1999 Achermann et al. identified mutations in SF-1 in 46,XY girl with complete gonadal agenesis, male to female sex reversal, and primary adrenal failure. In this patient, de novo mutation (p. G35E) was found in the P-box of the first zinc finger domain, affecting key amino acid in this domain, responsible for DNA binding. The second patient had similar phenotype but was found to carry homozygous recessive mutation (p. R92Q) affecting key amino acid in the A box within the Ftz-F1 box. Despite these two mutations, it turned out that mutations in SF-1 are rare in patients with adrenal phenotype (Suntharalingham et al. 2015). In contrast, SF-1 seems to be more often involved in disorders of sexual development (DSD). Heterozygous mutations in SF-1 in DSD patients were found to be remarkably common, present in 10–20% of all DSD patients (Suntharalingham et al. 2015). The most often observed phenotype is a chromosomally male patient (46,XY DSD) with ambiguous genitalia and with, or more often without, Mullerian structures. The testes are often well developed, and main phenotype characteristics seem to be arising due to androgen deficiency. Therefore, in some cases, the patient may be misdiagnosed for androgen insensitivity syndrome (Coutant et al. 2007). In addition to 46,XY DSDs, SF-1 mutations have been identified in patients with hypospadias and undescended testes (Köhler et al. 2009) and bilateral anorchia (Philibert et al. 2007), and some studies suggested that mild mutations/variations in SF-1 gene might be involved in some cases of male factor infertility (Bashamboo et al. 2010; Röpke et al. 2013; Ferlin et al. 2015).
Similarly to male patients, primary ovarian insufficiency has been reported in sisters or mothers of children with DSD due to SF-1 mutations. The clinical phenotype in these patients is very variable, from rare reports of complete ovarian insufficiency and the absence of menarche to more common early menopause and early depletion of ovarian reserves (Janse et al. 2012; Voican et al. 2013).
Hyperactivity and/or overexpression of SF-1 has also been linked to pathologies in human patients. In high proportion of childhood adrenal carcinomas, duplication of the gene locus containing SF-1 gene has been reported in somatic cells, and overexpression of SF-1 has been detected in many adult adrenal tumors. The degree of overexpression of SF-1 might even be an important predictor of adrenal tumor outcome, as suggested by some studies (Almeida et al. 2010; Sbiera et al. 2010).
Interestingly, although studies in mice have shown importance of SF-1 also for the development of hypothalamus, no reports about SF-1 role in obesity in humans have been reported to date. Interestingly, there is one report suggesting a possible link between SF-1 mutations and psychiatric symptoms of anxiety and/or depression (Suwanai et al. 2013) although the direct connection between SF-1 mutations and psychiatric symptoms has not been unequivocally established.
SF-1, first identified over 20 years ago, has been shown conclusively to be one of the most important genes involved in the regulation of gonadal and adrenal development, with important functions in other tissues such as pituitary, hypothalamus, and spleen. Although some roles of SF-1 such as regulation of expression of steroidogenic enzymes in gonads and adrenal glands, regulation of expression of AMH and Sox9 in fetal Sertoli cells, regulation of GnRH receptor expression in the pituitary, and regulation of spleen development through Tlx1 and retinoic acid signaling have been clearly established, there are many questions still unanswered. Precise temporal and tissue-specific regulation of SF-1, functional role of the SF-1 in the hypothalamus, role of SF-1 in infertile patients, and some other human diseases are among the questions that are still not fully understood and will have to be studied in future years.
- Ferlin A, Santa Rocca M, Vinanzi C, Ghezzi M, Di Nisio A, Foresta C. Mutational screening of NR5A1 gene encoding steroidogenic factor 1 in cryptorchidism and male factor infertility and functional analysis of seven undescribed mutations. Fertil Steril. 2015;104(1):163–9. e1.Google Scholar
- Janse F, Larissa M, Duran KJ, Kloosterman WP, Goverde AJ, Lambalk CB, et al. Limited contribution of NR5A1 (SF-1) mutations in women with primary ovarian insufficiency (POI). Fertil Steri. 2012;97(1):141–6. e2.Google Scholar
- Köhler B, Lin L, Mazen I, Cetindag C, Biebermann H, Akkurt I, et al. The spectrum of phenotypes associated with mutations in steroidogenic factor 1 (SF-1, NR5A1, Ad4BP) includes severe penoscrotal hypospadias in 46, XY males without adrenal insufficiency. Eur J Endocrinol. 2009;161(2):237–42.PubMedPubMedCentralCrossRefGoogle Scholar
- Shima Y, Zubair M, Komatsu T, Oka S, Yokoyama C, Tachibana T, et al. Pituitary homeobox 2 regulates adrenal4 binding protein/steroidogenic factor-1 gene transcription in the pituitary gonadotrope through interaction with the intronic enhancer. Mol Endocrinol. 2008;22(7):1633–46.PubMedPubMedCentralCrossRefGoogle Scholar
- Zubair M, Ishihara S, Oka S, Okumura K, Morohashi K. Two-step regulation of Ad4BP/SF-1 gene transcription during fetal adrenal development: initiation by a Hox-Pbx1-Prep1 complex and maintenance via autoregulation by Ad4BP/SF-1. Mol Cell Biol. 2006;26(11):4111–21.PubMedPubMedCentralCrossRefGoogle Scholar