Based on the KO phenotypes, Nr5a1 was speculated to have the potential to differentiate pluripotent cells into gonadal and adrenocortical cells. This potentiality was first discovered by Crawford et al. [29] in 1997 by successfully establishing steroidogenic cells from murine embryonic stem cells through stable Nr5a1 expression alone. Similar results were obtained by Gondo et al. [30] and Yazawa et al. [31] using bone marrow cells and mesenchymal stem cells, respectively. Although these studies indicate the potential for Nr5a1 to contribute to the differentiation of pluripotent stem cells into steroidogenic cells, Nr5a1 alone appeared to be insufficient to specify steroidogenic cells.

Studies of developing gonads have accumulated evidence that sets of transcription factors are required for the differentiation of Sertoli and Leydig cells [32]. Based on these observations, direct reprograming of Sertoli cells from mouse fibroblasts was achieved by Buganim et al. [33]. In this study, the complete reprograming process was divided into three steps. The first step, cell proliferation and the mesenchymal-epithelial transition, was promoted by Nr5a1, Wilms tumor 1 (Wt1), and doublesex and mab-3 related transcription factor 1 (Dmrt 1); the second step, cell aggregation, was promoted by Nr5a1, Wt-1, and SRY-box transcription factor 9 (Sox9); and the last step, conversion to Sertoli cells, was promoted by Nr5a1, Wt-1, Dmrt1, GATA binding protein 4 (Gata4), and Sox9. Notably, Nr5a1 is required for all three steps. By applying a similar experimental design, Liang et al. [34] revealed that two genes, NR5A1 and GATA4, are sufficient for the direct reprograming of human fibroblasts to Sertoli cells. Moreover, under the same concept, Yang et al. [35] successfully reprogramed mouse fibroblasts directly to Leydig cells using Nr5a1, Gata4, and Dmrt1. These studies have shown that Nr5a1/Nr5A1 has the potential to promote whole reprograming steps and, in combination with other transcription factors, to specify the cell types to differentiate.

The aforementioned studies have indicated the indispensable role of Nr5a1 during cell differentiation. NR5A2 (liver receptor homolog-1 [LRH1]), another member of the NR5A subfamily, recognizes the same nucleotide sequences as NR5A1 [2,36]. Therefore, both these NR5A family members were believed to potentially regulate the same sets of target genes. Although their cellular expression is different, the patterns of expression partially overlap. In fact, both NR5A1 and NR5A2 are expressed in the ovary to regulate Cyp11a1 genes [37].

Gu et al. [38] found that Nr5a2 is expressed in the inner cell mass and epiblast of early-stage mouse embryos. Using gene-disrupted mice, they revealed that Nr5a2 in those cells tightly regulates the expression of octamer-binding transcription factor 4 (Oct4), one of the genes essential for reprograming of differentiated cells to pluripotent stem cells. Consistent with this finding, Heng et al. [39] demonstrated that Oct4 can be replaced by Nr5a2 in the reprograming of murine somatic cells to pluripotent cells. Interestingly, both Nr5a1 and Nr5a2 have the potential to reprogram and establish pluripotent cells [38-40]. Given that these two factors share common binding sequences, their common activities were expected to be realized by regulating the same sets of target genes. Two chromatin immunoprecipitation (ChIP)-sequence studies were conducted to identify the target genes of the two factors: one with Nr5a2-expressing mouse fibroblasts [39] and the other with the NR5A1-expressing human embryonic stem cell line (Table 1) [40]. The results obtained from the mouse cells revealed that NR5A2 together with SOX2 and Kruppel like factor 4 (KLF4) regulates genes pivotal for the maintenance of the identity of embryonic stem cells, such as POU domain, class 5, transcription factor 5 (Pou5f1), homeobox transcription factor Nanog (Nanog), T-box transcription factor 3 (Tbx3), and Klf2, while those from the human cells revealed that NR5A1 accumulates at the developmental pluripotency associated 2 (DPPA2) and DPPA4 genes that are known to be key reprograming factors in mice.

Based on the findings of the studies described in this section, NR5A1/Nr5a1 can be inferred to be fundamental for multiple steps of cell differentiation, such as the establishment and reprograming of pluripotent cells and the differentiation of cells to gonadal somatic cells, including steroidogenic cells. Given that future studies would focus on unveiling the mechanism by which NR5A1 regulates these steps of cellular differentiation, identifying NR5A1 target genes is essential to delineate the mechanism.

As describe above, KO phenotypes strongly suggest a role for Nr5a1 in the regulation of cell proliferation. A possible correlation between the amount of Nr5a1 and cell proliferation was suggested for the first time by Bland et al. [41]. They reported that the adrenal glands of Nr5a1 heterozygous mice were smaller than those of wild-type mice [41]. Consistent with this finding, Beuschlein et al. [42] reported that compensatory adrenal growth after unilateral adrenalectomy was strongly affected in heterozygous mice. Conversely, forced Nr5a1 expression resulted in enlargement of the fetal adrenal cortex [43], possibly caused in part by enhanced proliferation [44]. This cell proliferation promotion activity was reproduced with cultured cells [44] and inhibited by inverse agonists of NR5A1 [45].

Along with these studies, the following studies have advanced our understanding of the mechanism by which NR5A1 regulates cell proliferation. Doghman et al. [44] raised the possibility that Nr5a1 regulates cell proliferation by activating FATE1 (a cancer testis antigen) expression. Ishimaru et al. [46] demonstrated that forced expression of Nr5a1 induced cyclin D1 expression, whereby promoting cell proliferation within the chick embryonic gonad. Similarly, Syu et al. [47] demonstrated that proliferation of Y-1 cells was promoted via Nr5a1-activated cyclin E1 expression. In a study investigating cell cycle regulation, Lewis et al. [48] demonstrated that the transcriptional activity of NR5A1 is regulated through phosphorylation by cyclin dependent kinase 7 (CDK7). CDK7, a cyclin-dependent kinase, forms a trimeric complex with cyclin H and Mat1 to act as a CDK-activating kinase complex; this complex is a component of transcription factor IIH (TFIIH) and thus associated with the regulation of basal transcription [49]. Taken together, CDK7 might be assumed to serve as a direct link between cell cycle progression through activation of other CDKs and transcription through NR5A1 phosphorylation. Moreover, CDK7 could couple NR5A1-driven transcriptional activation with activation of the basal transcriptional machinery.

Regarding regulation of cell proliferation by Nr5a1, interesting observations were reported by Lai et al. [50] and Wang et al. [51]. A series of studies reported that NR5A1 was localized at the centrosome. Dissociation of NR5A1 from the centrosome promoted DNA-dependent protein kinase (DNA-PK) recruitment and thereby activated CDK2, whose activity is required for the duplication of DNA and the centrosome. When Nr5a1 is knocked down, CDK2 is aberrantly activated by DNA-PK, and thus, the centrosome is over-duplicated. Consequently, cell proliferation is disordered. Although the studies described in this section are still fragmented rather than tightly connected, evidence supporting the role of Nr5a1 in cell cycle regulation has accumulated gradually from a wide variety of sources.

Early studies of Nr5a1 predominantly focused on steroidogenic genes. Thereafter, studies were conducted to identify target genes in nonsteroidogenic cells, such as Sertoli cells and pituitary gonadotrophs. Consequently, in addition to steroidogenic genes, several nonsteroidogenic gene targets were identified. However, it seemed unlikely that NR5A1, merely thought to regulate genes already identified, was responsible for a wide range of biological processes, such as differentiation and reprograming to particular cell types, cell proliferation, or survival. To comprehensively understand the actions of Nr5a1, genome-wide studies with DNA arrays and deep sequencing were conducted (Table 1) [40,52-58]. In fact, these genome-wide studies identified many novel target genes of NR5A1, and eventually unveiled novel roles of NR5A1 in biological, physiological, and pathological processes.

Human cohort studies have shown that strong NR5A1 expression in adrenocortical carcinoma correlates with a poor clinical outcome [59]. However, it remained to be clarified which genes, under the control of the highly-expressed NR5A1, were responsible for the poor prognosis. Through ChIP-seq, Ruggiero et al. [56] identified guanine nucleotide exchange factor 2 (VAV2) as one of the NR5A1 target genes in H295R cells (a human adrenocortical carcinoma cell line). VAV2, a member of the VAV family, has been characterized as a guanine exchange factor that activates the Rho/Rac family of GTPases, and thus promotes cellular remodeling and invasion [60]. Ruggiero et al. [56] clearly demonstrated that NR5A1-induced VAV2 activates the small GTPases cell division cycle 42 (CDC42) and Rac family small GTPase 1 (RAC1), and consequently promotes cytoskeleton remodeling and cell invasion. This result could provide the rationale behind highly-expressed NR5A1 causing poor clinical outcomes in adrenocortical carcinoma.

By using Y-1 adrenocortical and testicular Leydig cells, the role of Nr5a1 in the regulation of energy metabolism-related genes was identified through ChIP-seq. Baba et al. [55] demonstrated that nearly all glycolytic genes are regulated by NR5A1 (Fig. 3). Indeed, Nr5a1 knockdown resulted in a decrease in glucose consumption as well as in glycolytic gene expression. Along with the tricarboxylic acid cycle and oxidative phosphorylation, glycolysis is the main pathway for supplying the energized molecule, adenosine triphosphate (ATP). As expected, intracellular ATP concentration decreased with Nr5a1 knockdown. Another energized molecule, nicotinamide-adenine dinucleotide phosphate (NADPH, reduce form), is required for the synthesis of various biomolecules. Steroidogenic reactions mediated by cytochrome P450s consume NADPH [61]. This energized molecule is synthesized by multiple pathways and enzymes, such as the pentose phosphate pathway, malic enzymes, and methylenetetrahydrofolate dehydrogenases. Among these, genes encoding malic enzyme 1 and methylenetetrahydrofolate dehydrogenase 2 were identified as Nr5a1 targets [57]. NADPH concentration also decreased with Nr5a1 knockdown.

Moreover, many cholesterogenic genes were shown to be NR5A1 target genes by Baba et al. [58]. Concordantly, cholesterogenic activity was decreased by Nr5a1 knockdown. Cholesterol is utilized as a starting material for multiple steroidogenic reactions, wherein the first reaction is mediated by the cholesterol side-chain cleavage P450 localized in mitochondria. Therefore, cholesterol must be transported into mitochondria. Hypoxia up-regulated mitochondrial movement regulator (HUMMR)/mitochondria-localized glutamic acid-rich protein (MGARP) promotes cholesterol transfer to the mitochondrial outer membrane [62], and interestingly, the Hummr/Mgarp-encoding gene is an NR5A1 target together with cholesterogenic genes. Expectedly, the amount of cholesterol in mitochondria decreased with Nr5a1 knockdown [58].