Introduction

Sexual development in mammals is conditioned by the gonadal sex established during fetal life. In XY embryos, the initially undifferentiated gonads develop into testes that can produce testosterone which stimulates the growth of male internal and external genitalia (including epididymis, vas deferens, seminal vesicles, and penis) leading to the masculinization of the fetus. In contrast, fetal ovaries do not produce testosterone and female internal and external genitalia (oviducts, uterus, vagina, and vulva) develop in XX individuals. Abnormalities in the masculinization process are mainly associated with defects in the androgen synthesis or signaling (1). Although this molecular cascade is well defined, the differentiation of androgen-producing cells in the embryo is only partially understood.

Androgen production in the developing testis relies mainly on Fetal Leydig Cells (FLC), which express all enzymes required for the biosynthesis of androstenedione from cholesterol (2–5). The final step of conversion of androstenedione into testosterone, catalyzed by HSD17B3, takes place in a different cell type, the fetal Sertoli cells (4, 6). In addition to their role in androgen synthesis, FLC also produce INSL3, a hormone required for testis descent (7, 8). Defects in testis descent, regulated also by androgens, result in cryptorchidism, a condition which impacts fertility and constitutes a risk factor for testicular cancer (9). While a fraction of FLC persist after birth, others de-differentiate or involute and adult testosterone production is ensured by a distinct population of steroidogenic cells, the Adult Leydig cells (ALC), which differentiate at puberty (10, 11).

FLC differentiate from embryonic day 12.5 (E12.5) in mice and increase in number during fetal life through the recruitment and differentiation of proliferative progenitors located in the interstitial space of the testes (10–12). Interstitial progenitors also give rise to the contractile peritubular myoid cells (PTM) that will surround the future seminiferous tubules (5). Despite being the most abundant cell population in the fetal testis (5, 13), little is known about the genetic control of the proliferation, specification, and differentiation of the interstitial steroidogenic progenitor cells. Lineage tracing and single cell transcriptomic analyses have revealed that interstitial progenitors have a dual origin. The majority are derived from the coelomic epithelium of the undifferentiated gonad which harbors early bipotential progenitors able to differentiate along the supporting (the future Sertoli cells) or the steroidogenic lineage (5, 14–16). In addition, Nestin- positive cells migrate from the adjacent mesonephros into the gonad and differentiate along the steroidogenic lineage from E13.5 to give rise to up to a third of the FLC population by the end of gestation (5, 17, 18).

Positive and negative paracrine signals drive FLC differentiation by up-regulating the transcription factors NR5A1, GATA4 and GATA6 which in turn regulate the expression of genes related to cholesterol metabolism and steroidogenesis (12, 19–26). Desert Hedgehog (DHH) is secreted by Sertoli cells and acts on the interstitial progenitors expressing the Hedgehog receptor Patched1 (PTCH1) and the Hedgehog effectors GLI1, 2 and 3 to trigger FLC and PTM differentiation (27–33). In addition FLC development requires the activation of signaling pathways downstream of PDGFRA in the steroidogenic progenitors (34–36). On the other hand, ligands present in vascular and peri-vascular cells activate the NOTCH2 receptor and the expression of the effectors HES1 and HEYL in interstitial progenitors to maintain their undifferentiated state and inhibit FLC differentiation (18, 37–39). In addition to the paracrine signals from adjacent cell populations, FLC differentiation is also regulated by the cell-autonomous action of the transcription factors ARX, TCF21, PBX1, MAF and MAFB expressed in the interstitial steroidogenic progenitors, although their precise roles remain elusive (5, 40–43).

Nuclear Receptor subfamily 2 group F member 2 (NR2F2, also known as COUP-TFII) is abundantly expressed in interstitial progenitors of the fetal and adult testis in rodents and humans (3, 5, 44–49). NR2F2 activates or represses transcription depending on the cellular context by directly binding to DNA responsive elements or by interacting with other transcription factors. NR2F2 regulates cell differentiation during organogenesis, adult tissue homeostasis and tumorigenesis (50). NR2F2 function is essential for cardiac and vascular development so that Nr2f2 mutation in mouse leads to embryonic lethality at mid-gestation (51). The study of mouse conditional mutants has shown that NR2F2 is essential for ALC differentiation in the post-natal testis before puberty (48). However, the function of NR2F2 in the developing testis during fetal life has not been addressed and its role in the interstitial progenitors that give rise to the FLC lineage is currently unknown. It was initially proposed that NR2F2 could act as a negative regulator of steroidogenesis at fetal stages based on the inverse correlation between NR2F2 expression and steroidogenesis genes and testosterone levels in mouse and rat fetal testes treated with endocrine disruptors (46). More recently, rare variants in NR2F2 have been associated with cryptorchidism, hypospadias and defective penile growth in human patients (52–54). These phenotypes can be attributed to defective testosterone and INSL3 production during gestation (9, 55), suggesting a positive role for NR2F2 in promoting FLC differentiation and/or function in the fetal testis.

In this study we show that NR2F2 is expressed in interstitial progenitors of coelomic and mesonephric origin of the mouse fetal testes and is downregulated upon FLC differentiation. By using two Cre lines that drive Nr2f2 deletion in mouse embryonic gonads, we show that NR2F2 is required for fetal mouse testis morphogenesis and for FLC differentiation. Absence of Nr2f2 does not impair paracrine signals known to regulate FLC differentiation nor the proliferation or survival of the steroidogenic progenitor population. Our findings reveal that NR2F2 promotes the initiation of FLC differentiation as well as FLC maturation. Taken together, these results establish NR2F2 as an essential factor that positively regulates the development of steroidogenic cells in the mouse fetal testis.

Results

NR2F2 is expressed in steroidogenic progenitors of the developing testis

We analyzed the spatio-temporal distribution of NR2F2 in the developing testis by immunofluorescence. At E11.5 (18-21 tail somites, ts) NR2F2 was detected in both the coelomic epithelium and the mesonephric mesenchyme adjacent to the gonads, two tissues that contribute to the population of interstitial steroidogenic progenitors (5, 14, 18) (Fig. 1 A and C). The majority of mesenchymal cells in the gonad expressed the transcription factor RUNX1 (RUNX1+) (56), indicating that most of the gonadal somatic cells at this stage belong to the supporting lineage (Fig. 1 A and D). Nevertheless, NR2F2+ RUNX1- cells were observed, revealing that interstitial progenitors were already present at this stage (Fig. 1A). NR2F2+ cells were either GATA4+ or GATA4-, suggesting that interstitial progenitors of coelomic and mesonephric origins respectively, were both present (Fig. 1 A-C).

Figure 1.

In E12.5 testes, NR2F2+ cells were detected in the coelomic epithelium and in the interstitial space outside the developing testis cords (Fig. 1 E-H, Fig. S1 A-D). NR2F2+ cells co- expressed the steroidogenic progenitor marker ARX (5, 40), and were actively proliferating (Fig.1 E-H, Fig. S1 E-H). At E14.5, NR2F2 expression was maintained in interstitial cells co-expressing PDGFRA (34), as well as in the peritubular myoid cells lining the testis cords and in the cells beneath the surface of the testis that will contribute to the future tunica albuginea (Fig. 1 I and J, Fig. S1 M-P). NR2F2 expression was also found in NESTIN expressing peri-vascular cells that correspond to the mesonephros derived steroidogenic progenitors (18) (Fig. 1 K and L). In contrast, NR2F2 protein was either absent or detected at very low levels in FLC marked by the expression of the steroidogenic enzyme HSD3B (Fig. 1 M-P, Fig. S1 I-L). Together our results show that NR2F2 is expressed in the coelomic epithelium and the mesonephros as well as in the interstitial progenitors derived from both sources and is downregulated upon FLC differentiation in the developing testis.

NR2F2 is required for fetal testicular morphogenesis and FLC development

In order to investigate the function of NR2F2 in the developing mouse testis, we used a Nr2f2^flox conditional allele (9), and a knock-in WT1^CreERT2 line, in which tamoxifen inducible CreERT2 is produced by WT1 expressing cells (10). Wt1 is expressed from E9.5 in the coelomic epithelium of the gonadal ridge and in the adjacent mesonephros (59) and WT1^CreERT2 mediated recombination can be induced in all somatic gonadal cells upon tamoxifen administration at E9.5 and E10.5 (60) (Fig. S2A).

NR2F2 is co-expressed with WT1 in the gonadal coelomic epithelium, in the mesonephros and in interstitial cells (61)(Fig. S2 C-F). Tamoxifen treatment at E9.5 and E10.5 triggered an efficient NR2F2 deletion in gonadal and mesonephric tissues of WT1^CreERT2; Nr2f2^flox/flox embryos analyzed at E12.5 and E14.5 (Fig. 2 A-E, Fig. S2B). NR2F2 expression was completely absent in gonadal interstitial cells including NESTIN+ cells, demonstrating that all interstitial steroidogenic progenitors were targeted in WT1^CreERT2; Nr2f2^flox/flox embryos (Fig. 2 D and E). Morphological examination of the urogenital system of E16.5 WT1^CreERT2; Nr2f2^flox/flox embryos revealed hypoplastic undescended testes (Fig. S2 G and H), indicating that NR2F2 function is required for testicular development. Sertoli cells expressing SOX9 (Fig. 2 B and C), and interstitial progenitors marked by ARX (Fig. 2 F and G) were present in WT1^CreERT2; Nr2f2^flox/flox embryos at E12.5. However, differentiated steroidogenic FLC (marked by HSD3B expression) were almost completely absent in the mutant gonads (Fig. 2 J, K and N). This phenotype was not due to a delay in the initiation of FLC differentiation, as the number of HSD3B positive cells remained strongly reduced in WT1^CreERT2; Nr2f2^flox/flox mutant gonads at E14.5 (60% reduction, Fig. 2 L-N). In addition, the transcripts of Cyp11a1 and Cyp17a1, two genes encoding steroidogenic enzymes expressed in FLC, and of Insl3 were strongly reduced (Fig. 2O), providing a possible explanation for the observed undescended testes phenotype (Fig. S2 G and H). The mutant gonads showed enlarged and irregular shaped testis cords at E14.5 (Fig. 2E, Fig. S2J). Moreover, the expression of ACTA2 was strongly reduced both in the periphery of the gonad and in cells lining the testis cords, indicating that tunica cell and peritubular myoid cell development were impaired in WT1^CreERT2; Nr2f2^flox/flox mutants (Fig. 2 H and I). Together these results demonstrate that NR2F2 function is not essential for the initial specification of the interstitial and supporting cells of the testis but is required for testicular morphogenesis and for FLC development.

Figure 2.

Sertoli cell development is impaired in WT1^CreERT2; Nr2f2^flox/flox testes

FLC differentiation relies on signals produced by Sertoli cells such as DHH and PDGFA acting on interstitial cells (12, 19). The number of Sertoli cells expressing SOX9 and the levels of Sox9 transcripts were not altered in WT1^CreERT2; Nr2f2^flox/flox mutants at E14.5 (Fig. 2N and O). However, the expression of Dhh, Pdgfa and Amh, another marker of differentiated Sertoli cells, were reduced in WT1^CreERT2; Nr2f2^flox/flox mutants mutant compared to Nr2f2^flox/+ control gonads (Fig. S2K) indicating that Sertoli cell differentiation is abnormal in WT1^CreERT2; Nr2f2^flox/flox gonads. WT1 is required for Sertoli cell development (62) and previous work suggested that Wt1 heterozygosity in WT1^CreERT2/+ gonads results in delayed testicular development (63). To discriminate between the effects of Nr2f2 loss of function and Wt1 heterozygosity in the phenotype of WT1^CreERT2; Nr2f2^flox/flox mutants, we analyzed gene expression WT1^CreERT2/+ testes compared to wild-type littermates. We found that the expression of the FLC markers Cyp11a1 and Cyp17a1 were not significantly different in WT1^CreERT2/+ testes compared to wild-type littermates (Fig. S2L). In contrast, transcript levels of Dhh, Pdgfa and Amh were significantly reduced (Fig. S2L). While these results indicate that impaired FLC development in WT1^CreERT2; Nr2f2^flox/flox mutants is associated with the loss of Nr2f2 function, the potential contribution of Sertoli cell defects caused by Wt1 heterozygosity to the FLC phenotype cannot be ruled out in this model.

NR2F2 is required in the steroidogenic lineage for FLC development

In order to elucidate the specific function of NR2F2 in the steroidogenic lineage for FLC development we used the transgenic Nr5a1-Cre line which drives robust recombination in somatic gonadal cells from E11.5, after the supporting and steroidogenic lineages have been specified (Fig. S3 A and B) (64, 65).

At E11.5, NR5A1 is co-expressed with NR2F2 in the gonadal coelomic epithelium and in interstitial cells, but is absent from the mesonephros and mesonephros derived cells (15, 18)(Fig. S3 D-G). Nr5a1-Cre; Nr2f2^flox/flox mutants show efficient Nr2f2 deletion in gonadal interstitial cells at E12.5 and E14.5 (Fig. 3 A-D, S3C). Consistent with previous reports on the activity of Nr5a1-Cre (65), NR2F2 was still detected in the coelomic epithelium layer and in interstitial cells just beneath it, particularly in the anterior part of the gonad of Nr5a1-Cre; Nr2f2^flox/flox mutants (Fig. 3 A-D). NESTIN+ NR2F2+ cells were still present, confirming that steroidogenic progenitors of mesonephric origin were not targeted by Nr5a1-Cre (Fig. 3 C and D, Fig. S3 H and I) (18). Together these results demonstrate that in Nr5a1-Cre; Nr2f2^flox/flox mutants, NR2F2 is deleted after E11.5 in interstitial cells derived from the coelomic epithelium except in the outermost layer of the testis.

Figure 3.

Nr2f2 deletion by Nr5a1-Cre did not affect the initial formation of the supporting and steroidogenic lineages as evidenced by SOX9 (Fig. 3 A, B and G, Fig. S3N and O) and ARX expression (Fig. 3 E and F), nor the differentiation of Sertoli cells as shown by normal expression levels of Dhh, Pdgfa and Amh (Fig. 3G, Fig. S3N). In addition, ACTA2 was detected in peritubular myoid cells and was only slightly reduced in the tunica albuginea of the posterior region in the mutant testes (Fig. S3 J and K). In contrast, the FLC population marked by HSD3B or CYP11A1 was decreased (40% reduction, Fig. 3 H-L, Fig. S3 L and M) and the expression of the FLC markers Cyp11a1, Cyp17a1 and Insl3 was strongly down-regulated in Nr5a1-Cre; Nr2f2^flox/flox mutant testes (Fig. 3M). In agreement with reduced FLC function during embryogenesis, the testes of Nr5a1-Cre; Nr2f2^flox/flox embryos were undescended and exhibited an abnormal abdominal position at postnatal day (P) 3 (Fig. 3 N and O).

We conclude that Nr2f2 deletion after E11.5 by Nr5a1-Cre leads to FLC reduction without Sertoli cell defects suggesting that NR2F2 is required cell-autonomously in the interstitial cells for FLC development.

NR2F2 is required for the initiation of FLC differentiation

FLC differentiate from proliferating interstitial progenitors that progressively lose their mitotic ability, downregulate the transcription factors ARX and NR2F2, up-regulate the master regulator of steroidogenesis NR5A1 and activate the expression of steroidogenesis related genes (3, 5, 40). We wanted to determine which of these steps of FLC differentiation are NR2F2 dependent.

The transcription factor ARX is required in the pool of proliferating interstitial progenitors for FLC development (40). We first examined whether NR2F2 regulates the survival, proliferation, or identity of the ARX+ cell population. We found that the expression of Arx mRNA was not modified in Nr5a1-Cre; Nr2f2^flox/flox mutants (Fig. S4 A). The total number of gonadal cells, the percentage of ARX+ cells among the total number of gonadal cells and the fraction of proliferating cells among the ARX+ population were similar to controls at E12.5 and E14.5 (Fig. 4 A, B and E, Fig. S4 B-E). In addition, we did not find evidence of increased cell death in Nr5a1-Cre; Nr2f2^flox/flox mutants (Fig. S4 F-H). Together these results indicate that ARX+ steroidogenic progenitor cells are present and proliferate at normal rates in Nr5a1-Cre; Nr2f2^flox/flox mutants.

Figure 4.

As the ARX+ steroidogenic progenitors adopt a FLC identity, NR2F2 is progressively lost and the nuclear receptor NR5A1 is strongly up-regulated (3, 5, 46). NR5A1 directs FLC differentiation by regulating the expression of genes associated with cholesterol metabolism and steroidogenesis (20, 23, 66, 67). In Nr5a1-Cre; Nr2f2^flox/flox mutants, the cells expressing high levels of NR5A1+ in the interstitial compartment were reduced compared to controls (Fig. 4 C-E, Fig. S4 I and J). This observation indicates that NR2F2 function is required for NR5A1 up-regulation at the initial step of steroidogenic cell differentiation.

DHH, PDGFRA and NOTCH pathways are not impaired in Nr5a1-Cre; Nr2f2^flox/flox mutants

We next investigated the impact of Nr5a1-Cre; Nr2f2 ^flox/flox mutation on the activation of the signaling pathways involved in FLC differentiation. Desert Hedgehog (DHH) secreted by Sertoli cells, binds to its receptor PTCH1 expressed in interstitial progenitors and activates its target genes, including Gli1, to promote FLC differentiation (27, 28). The expression levels of Dhh and of its target Gli1, a readout of Hedgehog pathway activation, were not modified in Nr5a1-Cre; Nr2f2^flox/flox mutants (Fig. 3G, Fig. 4F, Fig. S3N). Gli1 transcripts exhibited a similar distribution in interstitial cells of control and Nr5a1-Cre; Nr2f2^flox/flox mutants at E14.5 (Fig. 4G and H). PDGFA produced by Sertoli cells binds to the PDGFRA receptor expressed in interstitial progenitors to activate downstream signaling required for FLC development (34, 36). The expression levels of Pdgfa and Pdgfb, coding for two PDGFRA ligands expressed in developing testes (34) were not altered in Nr5a1-Cre; Nr2f2^flox/flox mutants (Fig. 3G, Fig. 4F). Transcript levels of Pdgfra, were unchanged in Nr5a1-Cre; Nr2f2^flox/flox mutants (Fig. 4F) and PDGFRA protein was detected at the plasma membrane of interstitial cells in both control and Nr5a1-Cre; Nr2f2^flox/flox mutant testes at E14.5 (Fig. 4I and J). In addition, the expression of Sgpl1, a PDGFRA signaling target involved in steroidogenic differentiation (36), was not reduced in Nr5a1-Cre; Nr2f2^flox/flox mutants (Fig. 4F). These observations indicate that the activities of Hedgehog and PDGFRA signaling, two pathways that positively regulate FLC differentiation, are not impaired in Nr5a1-Cre; Nr2f2^flox/flox mutants.

In addition to the positive signals, FLC differentiation is also negatively modulated by NOTCH signaling triggered by ligands expressed in vascular and peri-vascular cells. An increase in testicular endothelial cells results in a reduction in FLC numbers (18, 37–39). The distribution and abundance of endothelial cells marked by PECAM-1 was not altered in Nr5a1-Cre; Nr2f2^flox/flox mutants (Fig. 4K and L). In addition, Cdh5 transcript levels (a readout of the abundance of endothelial cells) were similar in controls and Nr5a1-Cre; Nr2f2^flox/floxmutants (Fig. 4M). These observations indicate that the reduction in FLCs in Nr5a1-Cre; Nr2f2^flox/floxmutants was not associated with an increased population of endothelial cells. The receptor NOTCH2 expressed in interstitial cells is involved in restricting FLC differentiation (18, 37–39). Notch2 mRNA levels were not altered in Nr5a1-Cre; Nr2f2^flox/flox mutants (Fig. 4M). NOTCH signaling activates the expression of its target genes (that also act as effectors of the pathway) including Heyl, specifically expressed in interstitial cells and strongly up-regulated upon vascular depletion (18) and Hes1, expressed in interstitial cells and involved in restricting FLC differentiation (37, 38). Transcripts of the NOTCH pathway targets Hes1 and Heyl were detected at similar levels in controls and Nr5a1-Cre; Nr2f2^flox/floxmutants (Fig. 4M), indicating that the observed decrease in FLC numbers cannot be attributed to an increase in NOTCH signaling.

Together these results indicate that NR2F2 deficiency does not impact the activity of the signaling pathways known to induce or repress FLC formation, suggesting that NR2F2 acts downstream or in parallel of these pathways to regulate FLC differentiation from interstitial progenitors.

NR2F2 is required for FLC maturation

A small fraction of HSD3B positive cells formed in absence of NR2F2 function, even in the case of WT1^CreERT2; Nr2f2^flox/flox mutant gonads that exhibit widespread NR2F2 deletion in steroidogenic progenitors of coelomic and mesonephric origin. One possibility is that these cells are adult Leydig cells that have prematurely differentiated in Nr2f2 mutant fetal testes. Bhmt (Betaine-homocysteine methyltransferase) is specifically expressed in ALC and absent in FLC (68). Bhmt transcripts were not detected in control or Nr2f2 mutant testes by RT-qPCR at E14.5 while they were present in adult testes (data not shown) indicating that the steroidogenic cells of the mutants do not have adult characteristics.

To evaluate the steroidogenic capacities of the remaining FLC in Nr2f2 mutant gonads, we analyzed the transcript levels of Cyp11a1 and Cyp17a1 normalized to the FLC number as quantified by HSD3B immunofluorescence (Fig. 2N, Fig. 3L) (31). Normalized data showed reduced steroidogenic gene expression in Nr5a1-Cre; Nr2f2^flox/flox and in WT1^CreERT2; Nr2f2^flox/flox mutant testes (Fig. 5A and B) suggesting that FLC formed in the mutant testes have reduced steroidogenic function.

Figure 5.

FLC differentiation is accompanied by a change in cell shape from spindle shaped progenitors to round shaped FLC, and an increase in cytoplasmic volume (46, 69). HSD3B positive FLC present in Nr5a1-Cre; Nr2f2^flox/flox and in WT1^CreERT2; Nr2f2^flox/floxmutants were smaller and more elongated than those in control littermates (Fig. 5 C-H, Fig. S5). These cellular characteristics have been associated with immature FLC at the initial stages of FLC differentiation before the formation of large and round testosterone producing FLC (46, 69).

Taken together, our results show that NR2F2 is required in interstitial steroidogenic progenitor cells for initiation of FLC differentiation but also for the subsequent step of FLC maturation leading to robust steroid production (Fig. 5 I).

Discussion

NR2F2 protein is expressed in interstitial cells of coelomic epithelium and mesonephric origin and is absent or detected at very low levels in FLC as soon as they are formed at E12.5. Our observations are consistent with NR2F2 positive cells being progenitors for FLC (3, 5) and indicate that NR2F2 is quickly down-regulated upon differentiation of the fetal steroidogenic lineage, when the cells begin to synthetize steroid hormones.

We used two Cre lines to address the role of NR2F2 in the mouse fetal testis. In both cases Nr2f2 mutants exhibit reduced FLC numbers, decreased steroidogenic gene expression and undescended testes. These findings identify NR2F2 as a positive regulator of FLC development in the mouse fetal testis. Nr2f2 mutation does not impair the survival or proliferation of the interstitial steroidogenic progenitor population. We found that the expression of the master regulator of steroidogenic differentiation NR5A1 is not up-regulated in Nr2f2 mutants, which is sufficient to account for the decrease in FLC numbers (20–23). NR2F2 regulates gene expression by directly interacting with NR5A1 in MA-10 cells, an in vitro model for immature ALC (44, 70, 71). Whether such an interaction could contribute to the strong up-regulation of Nr5a1 itself in differentiating FLCs and whether NR2F2 cooperates with other factors for the regulation of Nr5a1 will be the subject of future investigations.

The activity of the main paracrine pathways regulating FLC differentiation was not significantly altered in Nr2f2 mutants, as shown by the expression of target genes of the DHH, PDGFRA and NOTCH signaling. These observations indicate that NR2F2 is a permissive factor in steroidogenic progenitors acting downstream or in cooperation with the signaling pathways regulating FLC differentiation. NR2F2 expression has been found to be modulated by Hedgehog and/or NOTCH signaling in other cell types, and whether a similar regulation of Nr2f2 expression by DHH and NOTCH signaling occurs in the steroidogenic progenitors is not known (72, 73).

In contrast to the situation in Nr5a1 mutants where FLC are completely absent (23), a fraction of FLC differentiates in Nr2f2 mutants even when progenitors of both coelomic epithelium and mesonephric origins are targeted. This observation indicates that additional factors cooperate with NR2F2 to regulate the transition from the steroidogenic progenitor state to the differentiating FLC. The homeodomain protein ARX is expressed in Nr2f2 mutant steroidogenic progenitors and is a good candidate to be such a factor. Arx mutants exhibit decreased FLC numbers without defects in paracrine signals driving their differentiation, similar to Nr2f2 mutants (40).

FLC and ALC are morphologically, transcriptionally and functionally distinct (11, 68), yet NR2F2 function is required for the differentiation of both lineages. In contrast to the situation in the fetal testis, NR2F2 is maintained in cells that have started to express steroidogenesis genes in the post-natal testis (3, 5, 44). NR2F2 regulates the transition from the adult progenitor Leydig cell (characterized by their elongated shape, their ability to proliferate and a low level of testosterone synthesis) to the immature ALC (characterized by their round shape, low mitotic activity and increased testosterone production) (48, 69). In agreement with its in vivo role promoting maturation along the ALC lineage, NR2F2 co-operates with NR5A1 and GATA4 to activate the transcription of Insl3, Star (encoding the cholesterol transporter) and Amhr2 (44, 70, 71, 74, 75) in MA-10 cells. Here we found that the steroidogenic cells that differentiate in Nr2f2 mutant fetal testes exhibit a small size, an elongated shape and reduced steroidogenic gene expression, features of the earliest stages of FLC differentiation (19, 46, 69). A similar phenotype is obtained when Nr5a1 is deleted after the onset of FLC differentiation (67). This suggests that in addition to controlling the initial engagement of steroidogenic progenitors into the FLC lineage and similar to the situation in the post-natal testis, NR2F2 promotes FLC maturation possibly by directly regulating the expression of genes involved in steroidogenesis.

NR2F2 is required for additional aspects of fetal testis morphogenesis and differentiation. Testis cords are enlarged and abnormally shaped in Nr2f2 mutants. Testis cord development involves the formation of Sertoli-germ cell masses after E11.5 and their subsequent partition by growing wedges of interstitial cells and associated vascular branches at E12.5 (76). How NR2F2 dependent regulation of interstitial cell adhesion or migration contributes to this process will be the aim of future research.

Pathogenic variants in NR2F2 have been associated with congenital malformations including congenital heart disease, congenital diaphragmatic hernia and syndromic 46,XX testicular or ovo-testicular difference/disorder in sex development (DSD) (50, 77). More recently, defects in the external genitalia (micropenis, hypospadias), and cryptorchidism have been associated with rare heterozygous variants in NR2F2 in 46, XY patients (52–54). These phenotypes can be attributed to defects in testosterone-dependent masculinization and INSL3-dependent testis descent during gestation and could be explained by a failure of FLC differentiation in the fetal testis. NR2F2 is abundantly expressed in interstitial cells of fetal human testes, a population that likely contains the progenitors for FLC (45–47, 49). The present work demonstrating that NR2F2 is required in the steroidogenic progenitors of the murine fetal testis for the initiation and progression of FLC differentiation provides an entry point to understand the etiology of 46, XY DSD associated with pathogenic NR2F2 variants.

Materials and Methods

Mouse Strains and Genotyping

The experiments described herein were carried out in compliance with the relevant institutional and French animal welfare laws, guidelines, and policies. Mouse lines were kept on a mixed background B6CBAF1/JRj. The Nr2f2^tm1Vc line where Nr2f2 exon1 sequences (encoding the DNA binding domain) are deleted upon CRE mediated recombination (referred to as Nr2f2^flox), the knock-in Wt1^{tm2(cre/ERT2)Wtp}line where tamoxifen inducible Cre^ERT2 is produced by WT1 expressing cells (referred to as WT1^CreERT2), and the transgenic Tg(Nr5a1-cre)2Klp line where Cre expression is driven by Nr5a1 regulatory sequences (referred to as Nr5a1-Cre ), were genotyped as previously described (57, 58, 64). WT1^CreERT2; Nr2f2^flox/+or Nr5a1-Cre^tg/0; Nr2f2^flox/+ males were crossed with Nr2f2^flox/+ females to obtain mutant embryos at different stages. Embryos were named controls (Nr2f2^+/+ or Nr2f2^flox/+) or mutants (WT1^CreERT2; Nr2f2^flox/flox or Nr5a1-Cre^tg/0; Nr2f2^flox/flox). Genotypes of mice and embryos were determined using PCR assays on lysates from ear biopsies or tail tips. Genotyping primers are listed in Table S1. To activate the Cre^ERT2 recombinase in embryos, tamoxifen (TAM, T5648, Sigma-Aldrich) was directly diluted in corn oil to a concentration of 40 mg/mL, and two TAM treatments (200 mg/kg body weight) were administered to pregnant females by oral gavage at E9.5 and E10.5. For proliferation assays, 5-Bromo-2’-deoxy-Uridine (BrdU) (B5002, Sigma-Aldrich) diluted to a concentration of 10 mg/mL in sterile H2O, was administered to the pregnant females (50 mg/kg body weight) by intraperitoneal injection, and pregnant females and their embryos were humanely killed after 3 hours and 30 min. The day when a vaginal plug was found was designated as embryonic day E0.5. E11.5–E12.5 embryos were staged by counting the number of tail somites (ts) with 18 ts corresponding to E11.5.

Immunofluorescence staining and in situ hybridization

Embryos were fixed in 4% (w/v) paraformaldehyde (PFA, 15710-S, EMS) overnight, processed for paraffin embedding, and sectioned into 5 µm thick sections. Immunofluorescence and DAPI staining were performed as described in (78). Proliferation analysis was performed by using a BrdU detection kit (11 296 736 001, Roche). Gli1 mRNA was detected with the RNAscope technology (probe 311001) according to Advanced Cell’s instructions using the RNAscopeMultiplex Fluorescent Reagent Kit v2 Assay. Images were obtained on a motorized Axio Imager Z1 microscope (Zeiss) coupled with an AxioCam MRm camera (Zeiss) and processed with Fiji (Bethesda, MD, USA). The DAPI staining marking the nuclei was adjusted to visualize the tissues and may vary between samples. However, for the immunofluorescence analysis, the exposure time of the acquisition of the fluorescent signal was identical in the same experiment to allow comparison between controls and mutants. Images were assembled using the open-source software platform OMERO (https://www.openmicroscopy.org/omero/). Antibodies are listed in Table S2. At least three embryos of each genotype were analyzed for each marker.

Cell Quantifications

The gonadal area for each section was measured by creating a gonadal Region of Interest (ROI) drawn manually in Fiji. The number of HSD3B positive cells (2 to 493 cells per section depending on the stages and genotypes), SOX9 positive cells (231 to 882 cells per section depending on the stages and genotypes), NR5A1 positive cells (3 to 256 cells per section depending on the stages and genotypes), ARX positive cells (666 to 2487 cells per section depending on the stages and genotypes), ARX positive cells that had incorporated BrdU (232 to 656 cells per section depending on the stages and genotypes), Activated Caspase 3 positive cells (0 to 6 cells per section depending on the stages and genotypes) or the number of nuclei labelled by DAPI (1582 to 4279 cells per section depending on the stages and genotypes), were counted manually in the entire gonadal section using the cell counter Plugin from Fiji. For each genotype, gonads of 3 or 4 biological replicates were analyzed. Two to three coronal sections spaced by at least 30 µm in the medial regions of the gonads were analyzed for each individual. Statistical significance was assessed by Mann-Whitney U two-tailed test. * indicates P value ≤ 0.05; ns indicates P value > 0.05.

Quantification of area and circularity of HSD3B positive cells

Gonadal (ROI) were drawn manually, and HSD3B positive cells were segmented using Stardist Deep Learning plugin of Fiji with a minimum area of 20 µm² to remove small particles. The area and circularity of each segmented cell were measured with Fiji. Circularity = 4π*area/perimeter^2. A value of 1.0 indicates a perfect circle. As the value approaches 0.0, it indicates an increasingly elongated shape. For each genotype, 2 to 3 biological replicates were analyzed. The data are shown as violin plots (with median and quartiles) for control and mutant samples.

RNA extraction and Quantitative PCR analysis

Individual gonads were dissected from the mesonephros in PBS, snap-frozen in liquid nitrogen and kept at −80 °C. RNA was extracted by RNeasy Micro Kit (74004, Qiagen) and reverse transcribed by M-MLV reverse transcriptase (M170A, Promega). The cDNA was used as a template for quantitative PCR analysis using the SYBR Green I Master (04887352001, Roche) and a LightCycler 480 System (Roche). Primer sequences are listed in Table S3.

All biological replicates of different genotypes (N = 3–9) were run in the same plate and run as duplicate technical replicates. Relative gene expression of each gonad was normalized to the expression of the housekeeping genes Shda and Tbp (79) by the 2-ΔΔCt calculation method. GeNorm , BestKeeper algorithms and the comparative delta-Ct method provided through the online tool RefFinder (https://www.ciidirsinaloa.com.mx/RefFinder-master/?type=reference#) were used to confirm reference gene stability in the experimental datasets. Fold change in gene expression was obtained by dividing the normalized gene expression in gonads of a given genotype by the mean of the normalized gene expression in control gonads. Data are shown as means ± SEM. Statistical significance was assessed by Mann-Whitney U two-tailed test (GraphPad Prism 10.2.1). * indicates P value ≤ 0.05; ns indicates P value > 0.05.

Acknowledgements

We acknowledge the help from members of the Experimental Histpathology Platform, the PRISM Imaging Platform and the Animal house at iBV (Institut de Biologie Valrose, Université Côte d’Azur, CNRS, Inserm, iBV, France). We are grateful to members of the A. Schedl, M.C. Chaboissier and S. Nef groups for helpful discussions and to D. Wilhelm for critical reading of the manuscript. We are indebted to Dr. Chaponnier, Dr. Inoue, Pr. Morohashi, Dr. M. Vasseur-Cognet and Dr. Wilhelm for sharing mouse lines and reagents. This research was funded by Agence Nationale de la Recherche ANR-23-CE14-0012, Heterosex, and by a scholarship from the China Scholarship Council (to F.T.).