NEDD4 Promotes Sertoli Cell Proliferation and Adult Leydig Cell Differentiation in the Murine Testis

Abstract

Successful testis development relies on the coordinated differentiation and assembly of various cell types to establish both endocrine and reproductive functions. The ubiquitin ligase NEDD4 has emerged as a key player in murine testis development, with this enzyme being implicated in gonadal sex determination and spermatogonial stem cell differentiation. Here, we report hitherto uncharacterized roles of NEDD4 in postnatal testis development. Utilizing Nr5a1- and Amh-Cre drivers to conditionally ablate Nedd4 in testicular somatic cells, we show that NEDD4 promotes Sertoli cell proliferation through the modulation of the PI3K-AKT signaling pathway. This ubiquitin ligase also ensures proper differentiation of adult Leydig cells and may contribute to murine steroidogenesis. Furthermore, NEDD4 is essential for adrenal gland differentiation, as its loss results in adrenal dysgenesis. These findings highlight NEDD4 as a crucial factor in testis development, emphasizing the importance of ubiquitination and post-translational modifications in reproductive biology.

Successful testis development requires the coordinated differentiation and assembly of multiple cell types to enable the endocrine and reproductive capabilities of the mature testis. In mammals, testis fate is determined by the expression of the sex-determining region on the Y chromosome (Sry) in somatic cell progenitors. This initiates a rapid cascade in which SRY upregulates the expression of a related family member, SRY-box containing gene 9 (Sox9), driving the differentiation of the supporting cell lineage into Sertoli cells. The critical role of SRY and SOX9 in sex determination is evidenced by complete XY male to female gonadal sex reversal upon loss of either factor, while their overexpression in XX gonads induces testis formation (reviewed in (1, 2)).

Once testicular fate is determined, Sertoli cells play a crucial role in orchestrating the differentiation of other testicular cell types, including peritubular myoid cells and steroidogenic fetal Leydig cells (1, 2). They also drive early testicular morphogenesis including testis cord formation and testis vascularization (1, 2). In mice, fetal Leydig cells rapidly increase in number between 12.5 and 15.5 days post coitum (dpc) as a result of paracrine recruitment and differentiation of progenitor Leydig cells (3-5). Fetal Leydig cells ultimately regress postnatally and are replaced by adult Leydig cells. While the origins of fetal and adult Leydig cells are incompletely understood, both Leydig cell types are derived from progenitors within the testis interstitium (6-10). Adult Leydig cells arise from Leydig stem cells which express Nestin, Pdgfrb, and Nr2f2, which then differentiate into adult Leydig cell progenitors, generally characterized by the upregulation of Lhr and expression of Akr1c14 and Srd5a1. Within a week, these progenitors proliferate and differentiate into immature Leydig cells, expressing Hsd17b3, which increases upon further differentiation into adult Leydig cells (11-14).

Beyond their essential role in fetal development, Sertoli cells remain of crucial importance in the adult testis, where they support and nourish germ cells throughout spermatogenesis (15). Since each Sertoli cells can sustain only a limited number of germ cells, testis size and sperm production are directly correlated with Sertoli cell number (16). This Sertoli cell population is established during the late fetal and early perinatal period when they undergo a phase of rapid proliferation, before becoming mitotically inactive at approximately postnatal day (P) 15 (17-20).

The Neural precursor cell Expressed Developmentally Down-regulated protein 4 (NEDD4) is the founding member of the Nedd4 family of ubiquitin protein ligases (21). Ubiquitin ligases covalently attach ubiquitin or ubiquitin chains to specific substrates. Ubiquitin modifications play essential roles in virtually all aspects of cell signaling through proteolytic and nonproteolytic mechanisms, including proteasomal degradation and protein trafficking. NEDD4 regulates a complex array of pathways during development (21) and ablation of NEDD4 in mice results in embryonic lethality (22).

A role for NEDD4 in murine testis development is becoming increasingly evident, with this ubiquitin ligase being a key regulator of the cell fate decisions of various testicular cell types. In the postnatal testis, NEDD4 promotes spermatogonial stem cell differentiation through the targeted degradation of, the inhibitory factor, NANOS2 (23). NEDD4 is also of critical importance to the earliest stages of testis development, with Nedd4-deficient mice exhibiting complete, testicular to ovarian, gonadal sex reversal, owing to a reduction in progenitor cell proliferation, delayed onset of the testis-determining program and ectopic expression of ovarian-promoting pathways (24). To further elucidate the cell autonomous roles of NEDD4 in murine testis development, the Cre/Lox system was employed, allowing for conditional ablation of this E3 ubiquitin ligase specifically in somatic cells of the developing testis. This strategy circumvented the embryonic lethality and growth restriction observed in conventional Nedd4 constitutive knockout mice (22) enabling spatially restricted ablation of NEDD4 within in the somatic cells of the developing testis.

Here we report that conditional ablation of Nedd4 in testicular somatic cells did not replicate the sex reversal observed upon constitutive ablation of Nedd4 (24). However, it uncovered a later role for NEDD4 in testis development, characterized by a significant reduction in testis size due to impaired Sertoli cell proliferation and disrupted adult Leydig cell differentiation. These results, together with previous studies (23-25), highlight distinct and stage-specific roles of the ubiquitin protein ligase NEDD4 in murine testis development, emphasizing its importance beyond early sex determination.

Materials and Methods

Mouse Lines

Conditional Nedd4 knockout mice, wherein Nedd4 is ablated in the somatic cells of the developing testis, were generated using Nedd4^flox/flox mice (Nedd4^tm3.1Bros), in which exon 9 of Nedd4 is flanked by 2 loxP sites (26), crossed with Nr5a1-Cre positive mice (Tg(Nr5a1-cre)2Klp), which drives Cre expression under the Nr5a1 promoter (27), a gene expressed in the earliest stages of genital ridge formation (28). Nr5a1-Cre;Nedd4^Flox/+ mice were time-mated with Nedd4^Flox/Flox mice in order to generate Nr5a1-Cre;Nedd4^Flox/Flox mice, with noon of the day on which a vaginal plug was observed deemed as 0.5 dpc. For more accurate staging of fetuses up to 12.5 dpc, the tail somite stage (ts) was determined by counting the number of somites posterior to the hind limb, with 11.5 dpc corresponding with 18 ts, 12.0 dpc corresponding to 24 ts, and 12.5 dpc corresponding to 30 ts (29). For tissue collected postnatally, noon of the day of birth was determined to be P0.

Mice were genotyped using PCR on genomic DNA derived from tail tissue. Primers Nedd4Floxed_F: 5′-GTACATTTTAGTTCATGGTTCTCACAGG-3′ and Nedd4Floxed_R: 5′-CAGAGCTCACATGGCTGTGGG-3′ resulted in PCR products of 168 base pairs (bp) and 202 bp for the wildtype (WT) and floxed alleles respectively, while primers Cre408_F: 5′-GCATTACCGGTCGATGCAACGAGTGATGAG-3′ and Cre408_R: 5′-GAGTGAACGAACCTGGTCGAAATCAGTGCG-3′ generated a 408 bp product in Cre-positive mice. Genetic sex was determined as described previously (30).

For Sertoli cell specific ablation of Nedd4, Nedd4^flox/flox mice (Nedd4^tm3.1Bros) were mated with mice containing the Amh-Cre transgene (Tg(AMH-cre)1Flor) (31) to generate Amh-Cre;Nedd4^Flox/Flox mice, which were genotyped using the Nedd4Floxed primers described above and Cre26: 5′-CCTGGAAAATGCTTCTGTCCG-3′ and Cre36: 5′-CAGGGTGTTATAAGCAATCCC-3′, which amplified a 400 bp Cre fragment.

All animal experiments were performed with the approval of the Animal Ethics Committee of the University of Melbourne (approval number 1513725).

Immunofluorescence

Section immunofluorescence on paraformaldehyde-fixed, paraffin embedded mouse fetuses was performed as described previously (25). Primary antibodies used were mouse anti-NEDD4 (BD Transduction Laboratories, Catalog # 611481, RRID:AB_398941; 1:100), goat anti-DDX4 (R&D Systems, Catalog # RDSAF2030, RRID:AB_2277369; 1:300), rabbit anti-SOX9 (32) (RRID: AB_3696860; 1:200), rabbit anti-SRY (32) (RRID:AB_2687957; 1:100), rabbit anti-FOXL2 (33) (RRID:AB_2687958; 1:300), goat anti-GATA4 (Santa Cruz, Catalog # sc1237, RRID:AB_2108747; 1:300), rabbit anti-Laminin (Sigma Aldrich, Catalog # L9393, RRID:AB_477163; 1:300), rabbit anti-CYP11A1 (34) (RRID:AB_3676670; 1:300), mouse anti-POU5F1 (Santa Cruz, Catalog # sc5279, RRID:AB_628051; 1:100), mouse anti-SYCP3 (Abcam, Catalog # ab97672, RRID:AB_10678841; 1:100), goat anti-AMH (Santa Cruz, Catalog # sc6886, RRID:AB_649207; 1:300), sheep anti-SOX9 (35) (RRID:AB_2755027; 1:200), rat anti-NR5A1 (Transgenic Inc, Catalog # KO610, RRID:AB_3696859; 1:500), rabbit anti-tyrosine hydroxylase (Sigma-Aldrich, Catalog # AB152, RRID:AB_390204; 1:100), mouse anti-Ki67 (BD Transduction Lab, Catalog # 550609, RRID:AB_393778; 1:100) and mouse anti-NR2F2 (Perseus Proteomics, Catalog # PP-H7147-00, RRID:AB_2314222; 1:200). All secondary antibodies were purchased from Invitrogen and were used at a dilution of 1:300.

Quantitative Real-time RT-PCR

Mouse fetuses were collected at 14.5 dpc, gonad pairs were dissected and the underlying mesonephros removed before being snap-frozen in liquid nitrogen. Similarly, whole testes were dissected from P60 mice and snap-frozen in liquid nitrogen. Total RNA was isolated using Trizol (Ambion) and single stranded cDNA synthesized using the Protoscript II First Strand cDNA Synthesis Kit (New England Biolabs). qPCR was performed with the SensiFAST SYBR No-ROX Kit (Bioline) using the Rotor-Gene 3000 system (Qiagen). Relative mRNA levels were quantified using the 2^−ΔΔCt method with Sdha (fetal) or Tbp (postnatal) serving as the internal housekeeping control gene. Values are displayed as fold change relative to controls. Each reaction was performed in technical triplicate and each experiment was performed with at least 5 biological replicates. Numbers are specified in each corresponding figure legend. Values are plotted as the mean, with error bars representing the standard error of the mean (SEM) between biological replicates. Statistical significance between groups was determined using a 2-tailed, unpaired Student's t-test. Primer sequences can be found in Table S1 (36).

Testis and Seminal Vesicle Weight Measurement

Testes of P7, P14, P28, and P60 (Nr5a1-Cre;Nedd4^Flox/Flox) and P15 and P60 (Amh-Cre;Nedd4^Flox/Flox), and seminal vesicles of P60 Nr5a1-Cre;Nedd4^Flox/Flox and Amh-Cre;Nedd4^Flox/Flox control and mutant animals were dissected from each mouse and weighed. Weights were normalized to body weight and are displayed as the mean ± SEM with the average of controls set to 100%. Statistical significance between groups was determined using a 2-tailed, unpaired Student's t-test.

Histology

Testes were fixed overnight in Bouin's fixative and embedded in paraffin. To evaluate the histology of these organs, 5 μm transverse sections were stained with hematoxylin and eosin.

Sertoli Cell Proliferation Assay

Paraformaldehyde-fixed, paraffin embedded 15.5 dpc Nr5a1-Cre;Nedd4^Flox/Flox (n = 5) and control littermate fetuses (n = 4) were sectioned and co-immunolabelled with rabbit anti-SOX9 and mouse anti-Ki67. Between 3 and 5 whole testis cross-sections, at least 50 μm distant from one another were selected and imaged on a Zeiss, LSM800 confocal microscope. Sections were manually counted by blinded observers. Proliferating Sertoli cells (KI67⁺/SOX9⁺) are shown as a percentage of total Sertoli cells (SOX9⁺). Statistical significance between groups was determined using a 2-tailed, unpaired Student's t-test.

Immunoblotting

P60 mouse testes were homogenized, lysed in RIPA buffer (50 mM Tris-HCL [pH 7.5], 150 mM NaCl, 1% NP-40, 1 mM EDTA, 1 mM EGTA, 0.1% SDS and 0.5% Sodium deoxycholate) supplemented with cOmplete Protease Inhibitor Cocktail (Roche) and total protein measured using Bradford Assay (Thermo Scientific) before being further diluted in RIPA buffer as required; 4× SDS loading buffer (200 mM Tris-HCl [pH 6.8], 0.4M DTT, 8% SDS, 0.2% bromophenol blue, 40% glycerol) was added to each sample and boiled for 10 minutes before separating on 7.5%, 10%, or 12% polyacrylamide gels for 2 hours at 100 V. Separated protein samples were transferred onto polyvinylidene difluoride membranes, washed 3 times with Tris buffered saline supplemented with 0.1% Tween 20 detergent (TBST) and blocked with 5% bovine serum albumin in TBST for 1 hour. Membranes were incubated with primary antibody diluted 1:1000 in blocking solution overnight at 4 °C, washed 3 times in TBST and incubated with a HRP coupled secondary antibody for 2 hours. Membranes were washed 3 more times with TBST, prepared for imaging through the addition of a 1:1 mixture of H₂O₂ and Luminol (SuperSignal West Pico PLUS Kit, Thermo Fisher Scientific) and imaged using a ChemiDoc MP (BioRad). Primary antibodies used in this study are mouse anti-NEDD4 (BD Transduction Laboratories, 611481), rabbit anti-PTEN (Cell Signaling Technology, Catalog # 9559, RRID:AB_390810), rabbit anti-phospho-AKT (Ser473) (Cell Signaling Technology, Catalog # 9271, RRID:AB_329825), rabbit anti-β-actin (Cell Signaling Technology, Catalog # 4967, RRID:AB_330288), rabbit anti-Beclin (Cell Signaling Technology, Catalog # 3495S, RRID:AB_1903911), rabbit anti-p62 (Cell Signaling Technology, Catalog # 39749, RRID:AB_2799160), rabbit anti-LC3B (Cell Signaling Technology, Catalog # 3868, RRID:AB_2137707), rabbit anti-p53 (Cell Signaling Technology, Catalog # 32532, RRID:AB_2757821), mouse anti-PDCD6IP (aka. ALIX) (Cell Signaling Technology, Catalog # 92880, RRID:AB_2800192), and rabbit anti-GAPDH (Cell Signaling Technology, Catalog # 2118, RRID:AB_561053).

Sperm Count and Sperm Motility

For Amh-Cre;Nedd4^Flox/Flox mice cauda epididymides and vas deferens of P60 males were harvested in prewarmed M2 media, and fat and connective tissues were removed. Sperm cells were released in 1 mL of M2 media by manual dissociation of the tissue followed by an incubation of 10 minutes at 37 °C. A 1:40 dilution of the sperm suspension was loaded in a prewarmed 100-µm-deep counting chamber (Leja Products, Nieuw-Vennep, The Netherlands) and submitted to computer-assisted sperm analysis (CASA) (CEROS II, Hamilton Thorn Research Inc, Beverly MA). The settings employed for analysis were as follows: acquisition rate, 60 Hz; number of frames, 100; minimum illumination, 70; maximum illumination, 90; minimum cell area, 8; maximum cell area, 100; minimum elongation gate, 5; maximum elongation gate, 90; minimum cell brightness, 129; minimum tail brightness, 102; magnification factor, 0.74. The motility parameters measured were curvilinear velocity, average path velocity, and amplitude of lateral head displacement. Static, motile, progressive, and hyperactivated sperm were characterized by the default parameters of the software. At least 300 sperm cells were analyzed for each assay. The number of sperm present in the initial 1-mL suspension was calculated by the CASA system (10⁶ sperm cells/mL).

Results

Fetal Testis Development Proceeds Normally in Nr5a1-Cre;Nedd4^Flox/Flox Mice

Previous studies identified a role for NEDD4 in murine gonadal sex determination, where constitutive Nedd4 ablation led to complete male to female sex reversal in XY mice (24). To further investigate the cell autonomous roles of NEDD4 within the somatic gonad, while circumventing the fetal lethality associated with constitutive ablation, we employed the Cre-Lox system to conditionally delete Nedd4 in gonadal somatic cells, utilizing Cre recombinase under the control of the Nr5a1 promoter (Nr5a1-Cre;Nedd4^Flox/Flox mice).

In Nedd4-null gonads, sex reversal is characterized by insufficient progenitor cell proliferation, delayed onset of SRY expression, negligible SOX9 levels and ectopic activation of the ovarian program (24). In contrast, immunofluorescence analysis of Nr5a1-Cre;Nedd4^Flox/Flox and control gonads at E11.5 revealed normal SRY and SOX9 expression in Sertoli cells, along with DDX4-positive germ cells (Fig. 1A). Although a few FOXL2-positive cells were observed in Nr5a1-Cre;Nedd4^Flox/Flox gonads, they were insufficient to promote ovarian fate. By 12.5 dpc, SRY expression was down-regulated and SOX9-positive Sertoli cells had upregulated anti-Müllerian hormone (AMH) and formed testis cords with germ cells (Fig. 1B). Despite an apparently normal testis-determining program, Nr5a1-Cre;Nedd4^Flox/Flox testes were slightly smaller than control by 12.5 dpc (Fig. 1B, bottom panel). Taken together, these findings indicated that XY Nr5a1-Cre;Nedd4^Flox/Flox mice did not replicate the sex reversal observed upon constitutive ablation of Nedd4.

Figure 1.

The testis-determining program is unaffected in XY Nr5a1-Cre;Nedd4^Flox/Flox mice. Section immunofluorescence on 11.5 dpc (A) and 12.5 dpc (B) XY Nr5a1-Cre;Nedd4^Flox/Flox gonads (bottom) alongside Cre-negative XY littermate controls (top) stained for Sertoli cell markers SRY (green, left panel), SOX9 (green, middle panel), and AMH (magenta right panel in B), granulosa cell marker FOXL2 (green, right panel) and germ cell marker DDX4 (magenta in A and left and middle panels of B). The anterior pole of each gonad is positioned at the top of each panel. Scale bars = 100 μm.

To investigate the discrepancy in gonadal phenotypes between Nr5a1-Cre;Nedd4^Flox/Flox and Nedd4-null mice, and to confirm Nedd4 ablation, we performed NEDD4 immunofluorescence alongside the germ cell marker DDX4 at 11.5 dpc, approximately 32 hours after NR5A1 is first expressed (37). While NEDD4 was ubiquitously expressed throughout the gonad and surrounding tissues of control littermates, Nr5a1-Cre;Nedd4^Flox/Flox mice exhibited a reduction in NEDD4 expression in most DDX4-negative somatic gonadal cells, although expression was still observed in some somatic cells within the gonad (Fig. S1A (36)). Some of these cells may be endothelial cells, as these are not derived from NR5A1-positive progenitors. Unexpectedly, NEDD4 expression remained detectable in coelomic epithelial cells of XY Nr5a1-Cre;Nedd4^Flox/Flox mice, indicating that Nedd4 ablation was incomplete by 11.5 dpc. At 14.5 dpc, Nedd4 transcripts were significantly reduced (Fig. S1B (36)), yet NEDD4 protein was still present within XY Nr5a1-Cre;Nedd4^Flox/Flox gonads, albeit at a much lower level that in controls. While NEDD4 was widely expressed in control gonads (Fig. S1C (36)), consistent with our previous studies (24, 25), it was largely absent in SOX9-positive Sertoli cells and interstitial Leydig cells in XY Nr5a1-Cre;Nedd4^Flox/Flox mice (Fig. S1C (36)). While NEDD4 expression in germ cells within testis cords of Nr5a1-Cre;Nedd4^Flox/Flox mice was expected (Fig. S1C (36), white box), due to their Nr5a1-independent origins, the persistence of NEDD4 within the coelomic epithelium was unexpected (Fig. S1C (36), arrowheads).

Given that sex reversal in XY Nedd4^−/− gonads is attributed to insufficient proliferation of the coelomic epithelium, leading to a reduced pre-Sertoli cell pool, it is plausible that residual NEDD4 expression within the coelomic epithelium in XY Nr5a1-Cre;Nedd4^Flox/Flox gonads is sufficient to overcome this proliferation deficit, allowing the testis-determining program to proceed unperturbed.

Although XY Nr5a1-Cre;Nedd4^Flox/Flox mice did not replicate the gonadal sex reversal seen in XY Nedd4^−/− mice, they provided a valuable model to investigate the role of NEDD4 in the somatic cells of the developing testis, a task previously hindered by the complete gonadal sex reversal of XY Nedd4^−/− mice.

To explore this, we performed immunofluorescence on 14.5 dpc XY Nr5a1-Cre;Nedd4^Flox/Flox gonads, assessing key testicular markers (SOX9, GATA4, CYP11A1, AMH) and ovarian marker (FOXL2) to determine somatic cell identity. Additionally, we examined germ cell markers (DDX4, POU5F1, SYCP3) to evaluate potential indirect effects of somatic cells in instructing the development of fetal germ cells. Our analysis revealed that the testes of XY Nr5a1-Cre;Nedd4^Flox/Flox mice were morphologically and molecularly similar to Cre-negative littermate controls. Sertoli cells expressing SOX9 (Fig. 2A and 2F), GATA4 (Fig. 2B and 2G), and AMH (Fig. 2D and 2I), successfully formed testis cords which enclosed DDX4-positive germ cells (Fig. 2A, 2F, 2E, and 2J) and were outlined by the extracellular matrix protein laminin (Fig. 2B and 2G). Steroidogenic fetal Leydig cells, marked by CYP11A1 (Fig. 2C and 2H), properly differentiated within in the testis interstitium. No FOXL2-positive cells, indicative of gonadal sex reversal, were observed (Fig. 2E and 2J). Finally, while gonadal sex reversal extended to the developing germ cells in Nedd4^−/− mice (24 ), germ cells differentiated in line with that expected of gonocytes within the developing testis, retaining expression of the pluripotency marker POU5F1 (Fig. 2C and 2H) and staining negative for SYCP3 (Fig. 2D and 2I), a meiosis marker typical of germ cells within the developing ovary at this age.

Figure 2.

Unperturbed fetal testis differentiation in XY Nr5a1-Cre;Nedd4^Flox/Flox mice. Section immunofluorescence on 14.5 dpc XY Nr5a1-Cre;Nedd4^flox/flox testes alongside XY littermate controls stained for Sertoli cell markers SOX9 (green in A, E), GATA4 (green in B, G), and AMH (magenta in D, I), extracellular matrix protein laminin (magenta in B, G), Leydig cell marker CYP11A1 (green in C, H), germ cell marker DDX4 (magenta in A, E, F, J), pluripotency marker POU5F1 (magenta in C, H), meiosis marker SYCP3 (green in D, I), and granulosa cell marker FOXL2 (green in E, J). The anterior pole of each gonad is positioned at the top of each panel. Scale bars = 100 μm. K) RT-qPCR analyses of Sox9, Amh, Fgf9, Fgfr2, Inhba, Cyb26b1, Pou5f1, Nanog, Nodal, Nanos2, Dazl, Stra8, and Sycp3 expression at 14.5 dpc on XY Nr5a1-Cre;Nedd4^flox/flox gonads (blue, n = 5) and XY littermate controls (gray, n = 5). Values are normalized to Sdha and are expressed as fold change relative to controls. Mean ± SEM; t-test; no significant differences.

RT-qPCR analysis of micro-dissected tissue further confirmed that testis development remained unperturbed in Nr5a1-Cre;Nedd4^Flox/Flox testes (Fig. 2K). Mutant testes showed no significant differences in transcripts associated with Sertoli cells (Sox9, Amh, Fgf9, Fgfr2, Inhba, and Cyp26b1), male embryonic germ cell development (Pou5f1, Nanog, Nodal, and Nanos2), or female germ cell markers indicative of gonadal sex reversal (Dazl, Stra8, and Sycp3) (Fig. 2K). These findings confirm that by 14.5 dpc, Nr5a1-Cre;Nedd4^Flox/Flox develop normally, without deviation from control.

Taken together, these data suggest that, beyond its crucial role in establishing gonadal precursors, NEDD4 is not required for the maintenance of Sertoli or fetal Leydig cell identity.

Adrenal Cortex Dysgenesis in Nr5a1^Cre/+; Nedd4^Flox/Flox Mice

Given the shared developmental origins of the gonads and adrenal glands (28, 38), and the expression of Nr5a1-Cre in both gonadal somatic cells and the adrenal cortex (27), we also examined adrenal gland development in Nr5a1-Cre;Nedd4^Flox/Flox mice. In Cre-negative animals, NEDD4 was strongly expressed in the cytoplasm of NR5A1-positive cells of the adrenal cortex but was largely absent from the neural crest–derived chromaffin cells of the adrenal medulla at 14.5 dpc. In contrast, while NEDD4 expression was still observed in surrounding tissues of Nr5a1-Cre;Nedd4^Flox/Flox mice, expression was absent from the NR5A1-positive cell population, confirming successful ablation of Nedd4 in the adrenal cortex (Fig. S2A (36)). Notably, in both XX and XY Nedd4-ablated adrenal glands, the cortical cell population was significantly reduced compared to Cre-negative controls (Fig. S2 (36)), suggesting an essential role for NEDD4 in adrenal cortex development.

To visualize this phenotype further, immunofluorescence staining was performed on Nr5a1-Cre;Nedd4^Flox/Flox and control embryos at 14.5, 16.5, and 18.5 dpc using antibodies for NR5A1 and the chromaffin cell marker tyrosine hydroxylase (39). This analysis confirmed that Nedd4-deficient adrenal glands exhibited a severely diminished NR5A1-positive cortical cell population, resulting in a hypoplastic adrenal cortex and a failure to fully enclose the medulla (Fig. S2B (36)). From 14.5 dpc the developing adrenal gland becomes encapsulated by fibrous mesenchyme–derived tissue, establishing distinct cortical and medullary compartments, a process that is nearly completed by birth (40, 41). In Cre-negative controls, a well-defined adrenal cortex progressively expanded around the medulla, with these cell populations appearing near mutually exclusive by 18.5 dpc. In contrast, while NR5A1-positive adreno-cortical cells in Nr5a1-Cre;Nedd4^Flox/Flox embryos increased over time, their numbers remained insufficient to fully encapsulate the chromaffin cell population, resulting in a disorganized cortex–medulla boundary (Fig. S2B (36)).

Collectively, these data show that NEDD4 is expressed within the NR5A1+ cell population of the adrenal cortex and is required for adrenal cortex development, with the absence of Nedd4 resulting in adrenal cortex dysgenesis.

Aberrant Sertoli Cell Proliferation Upon Loss of Nedd4

To further investigate the role of NEDD4 in testicular somatic cell development, we analyzed adult Nr5a1-Cre;Nedd4^Flox/Flox testes at P60. Mutant testes exhibited a significant reduction in size (Fig. 3A, right testis) with testis weight decreased by 47.8% compared with littermate controls (Fig. 3B). Despite this size reduction, histological analysis using hematoxylin and eosin staining revealed little difference in the seminiferous epithelium of Nr5a1-Cre;Nedd4^Flox/Flox testes when compared with controls, with both containing seminiferous tubules capable of supporting spermatogenesis, as indicated by the presence of spermatozoa within the tubule lumen (Fig. 3C and 3D). Immunofluorescence analysis for SOX9 and FOXL2 at P60 further confirmed that Sertoli cells maintained their identity, with SOX9-positive Sertoli cells lining the basement membrane of seminiferous tubules, while the ovarian program was repressed, as evidenced by the absence of FOXL2 expression (Fig. 3E and 3F). To determine when testis size differences first emerged, we assessed testis weight at P7, P14 and P28. At all ages Nr5a1-Cre;Nedd4^Flox/Flox testes were significantly smaller than controls, exhibiting reductions of 39.5% at P7, 48.6% at P14, and 38.7% at P28 (Fig. 3G).

Figure 3.

Impaired Sertoli cell proliferation in Nr5a1-Cre;Nedd4^flox/flox mice. Testes (A) and Testis Weight (B, n = 14) of postnatal day (P) 60 Cre-negative littermate control (left, gray) and Nr5a1-Cre;Nedd4^flox/flox (right, blue) mice. Hematoxylin and Eosin staining of Cre-negative (C) and Nr5a1-Cre;Nedd4^flox/flox (D) P60 testes. Section immunofluorescence on P60 Cre-negative control (E) and Nr5a1-Cre;Nedd4^flox/flox (F) testes stained for Sertoli cell marker SOX9 (green), granulosa cell marker FOXL2 (magenta). Scale bars = 100 μm. (G) Testis weights of control (gray) and Nr5a1-Cre;Nedd4^flox/flox mice (blue) at P7 (n = 5), P14 (n = 6 control, 5 mutant), and P28 (n = 11 control, 9 mutant). Testis weights are normalized to body weight and are shown as a percentage of controls. (H) Section immunofluorescence on 15.5 dpc control and Nr5a1-Cre;Nedd4^flox/flox testes stained for proliferation marker KI67 (green) and Sertoli cell marker SOX9 (magenta), counterstained with DAPI (gray). Scale bars = 100 μm. (I) Quantification of Sertoli cell proliferation of control (gray, n = 4) and Nr5a1-Cre;Nedd4^flox/flox (blue, n = 5) fetal testes, shown as the proportion of proliferating Sertoli cells (KI67+ and SOX9+) within the total Sertoli cell pool (SOX9+). (J) Immunoblot on control (left) and Nr5a1-Cre;Nedd4^flox/flox (Nedd4^Δ/Δ, right) P60 testis lysates probed with NEDD4, PTEN, phosphorylated AKT (pAKT) and beta-actin (ACTB). Size markers (kilo Daltons) are shown on the lefthand side. (K) Quantification of pAKT immunoblot of control (gray, n = 3) and Nr5a1-Cre;Nedd4^flox/flox (blue, n = 3) P60 testis lysates. Values are normalized to beta-actin and are shown relative to littermate testes. All graphs display mean ± SEM; t-test; *P < .05, ***P < .001, ****P < .0001.

Since testis size is directly correlated with the total number of adult Sertoli cells (16), and reduced Sertoli cell proliferation leads to decreased testis weight (17), we next assessed the proliferative capacity of fetal Sertoli cells, to determine whether diminished Sertoli cell proliferation could contribute to the postnatal weight reduction observed in Nr5a1-Cre;Nedd4^Flox/Flox testes. Indeed, at 15.5 dpc, fetal Nr5a1-Cre;Nedd4^Flox/Flox testes exhibited 42.1% fewer proliferative Sertoli cells (KI67+ and SOX9+) than littermate controls (Fig. 3H and 3I). NEDD4 is recognized for its oncogenic potential through the negative regulation of the tumor suppressor phosphatase and tensin homolog (PTEN), a well-characterized negative regulator of the PI3K-Akt signaling pathway (42). Since NEDD4 and PTEN levels are often inversely correlated (43-45) and PTEN has been shown to negatively regulate Sertoli cell proliferation (46), we assessed whether increased PTEN expression, and diminished PI3K-AKT signaling might explain the reduced Sertoli cell proliferation in Nedd4-mutant testes. As expected, NEDD4 protein levels were significantly reduced in P60 Nr5a1-Cre;Nedd4^Flox/Flox testes but mutant testes showed no significant differences in total PTEN levels (Fig. 3J). Despite this, signaling along the PI3K-AKT axis was still diminished, with a 53.7% reduction of phosphorylated AKT (Fig. 3J and 3K). Together, these findings suggest that NEDD4 promotes Sertoli cells proliferation by regulating PI3K-AKT signaling, and its loss leads to impaired Sertoli cell proliferation and reduced testis size.

To further explore the role of NEDD4 in Sertoli cells, we generated Amh-Cre;Nedd4^Flox/Flox mice, in which Nedd4 was selectively ablated in Sertoli cells from approximately 15 dpc (31). While Nr5a1-Cre;Nedd4^Flox/Flox testes were significantly smaller than those of their littermates as early as P7, this was not the case upon Sertoli cell specific ablation of Nedd4, with Amh-Cre;Nedd4^Flox/Flox mutant testes being comparable in weight and size to controls at P15. By P60, however, Amh-Cre;Nedd4^Flox/Flox testes showed a 29.7% reduction in weight compared to littermates (Fig. 4A and 4B). At P60, mutant testes frequently displayed abnormal morphology, appearing shorter in length and often associated with a larger blood vessel (Fig. 4C, red arrows) in contrast to littermates (Fig. 4C, white arrows). Similar to Nr5a1-Cre;Nedd4^Flox/Flox mice, histological analysis of mutant testes revealed no major alteration of the seminiferous epithelium (Fig. 4D). Amh-Cre;Nedd4^Flox/Flox mice were viable, reached adulthood and exhibited normal sexual behavior, internal, and external genitalia, and unaltered seminal vesicles weights (Fig. 4E).

Figure 4.

Reduced testis size and abnormal testis morphology upon Sertoli cell–specific ablation of Nedd4. Testes (A) from control (left) and Sertoli cell–specific Nedd4 mutants (Amh-Cre;Nedd4^flox/flox, right) at postnatal (P) day 15 (top) and P60 (bottom). (B) Testis weight of P15 and P60 Cre-negative littermate controls (left, gray, n = 9 [P15], n = 7 [P60]) and Amh-Cre;Nedd4^flox/flox mice (right, red, n = 10 [P15], n = 11 [P60]) normalized to body weight and shown as a percentage of controls. (C) Testes of P60 controls (upper panel) and Amh-Cre;Nedd4^flox/flox mice (lower panel) showing abnormal large blood vessels (red arrowhead) in mutants compared with their control littermates (white arrowhead). (D) Hematoxylin and eosin staining of P60 control and Amh-Cre;Nedd4^flox/flox testes. Seminal vesicle weight (n = 5 control, 8 mutant) (E), sperm count (n = 6 control, 10 mutant) (F), and sperm motility measures (n = 6 control, 10 mutant) (G) of P60 control (left, gray) and Amh-Cre;Nedd4^flox/flox mice (right, red). All graphs display mean ± SEM; t-test; ns = not significant, ***P < .001.

The reduction in testis weight, upon Sertoli cell specific loss of Nedd4, was not associated with impaired sperm production. CASA revealed no significant differences in sperm counts between mutants and controls (Fig. 4F). Additionally, sperm produced by Amh-Cre;Nedd4^Flox/Flox mice exhibited normal levels of motility as those produced by their littermates (Fig. 4G). These data suggest that the absence of NEDD4 in Sertoli cells only partially contributes to the phenotype observed in Nr5a1-Cre;Nedd4^Flox/Flox mice, where Nedd4 is deleted in almost all testicular somatic cells.

Aberrant Leydig Cell Differentiation Contributes to Reduced Testis Size in Nr5a1-Cre;Nedd4^Flox/Flox Testes

Since Amh-Cre;Nedd4^Flox/Flox mice did not fully replicate the phenotypes observed in Nr5a1-Cre;Nedd4^Flox/Flox mice, we next assessed whether defective Leydig cell development or function, could contribute to the testis size reduction. While fetal Leydig cell differentiation seemingly occurred as expected (Fig. 2), we hypothesized that impaired adult Leydig cell differentiation might exacerbate the observed reduction in testis weight. Immunofluorescence at P60 for CYP11A1, a steroidogenic marker that increases with Leydig cell differentiation (11, 47), and NR2F2, a marker for both fetal (48) and adult (8) Leydig progenitor cells, revealed that while Nr5a1-Cre;Nedd4^Flox/Flox testes contained both progenitor and differentiated Leydig cells, their interstitial spaces appeared less cellularly dense than controls (Fig. 5A and 5B). This is also apparent in hematoxylin and eosin–stained sections (Fig. 3C and 3D). To confirm whether perturbed adult Leydig cell differentiation contributes to this diminished cellularity, we next utilized RT-qPCR to assess the relative expression of transcripts of adult Leydig cells at various states of differentiation.

Figure 5.

Perturbed adult Leydig cell differentiation in Nr5a1-Cre;Nedd4^flox/flox testes. Section immunofluorescence on P60 Cre-negative control (A) and Nr5a1-Cre;Nedd4^flox/flox (B) testes stained for Leydig cell markers CYP11A1 (green) and NR2F2 (magenta). Scale bars = 100 μm. (C) RT-qPCR analyses for fetal Leydig cell transcripts Ren1 and Crhr1 (red section), Leydig stem cell transcripts Nestin, Pdgfrb and Nr2f2 (yellow section), Leydig progenitor cell transcripts Akr1c14 and Srd5a1 (green section), immature adult Leydig cell transcript Hsd17b3 (light blue section), and adult Leydig cell transcripts Bhmt and Sult1e1 (purple section) on P60 control (gray, n = 10) and Nr5a1-Cre;Nedd4^flox/flox testes (blue, n = 9). Anogenital distance (D, E, n = 7) and seminal vesicle weight (F, n = 11 control, 10 mutant) of P60 Cre-negative littermate control (Left, gray) and Nr5a1-Cre;Nedd4^flox/flox mice (Right, blue). Seminal vesicle weight and anogenital distance are normalized to body weight and are shown as a percentage of controls. (G) RT-qPCR analyses for expression of genes encoding Ar, Lhr, Star, Cyp11a1, Cyp17a1, and Insl3 at P60 in Nr5a1-Cre;Nedd4^flox/flox testes (blue, n = 5) and Cre-negative littermate controls (gray, n = 6). RT-qPCR values are normalized to Tbp and are expressed as fold change relative to controls. Mean ± SEM; t-test; ns = not significant, *P < .05, **P < .01, ****P < .0001.

This analysis revealed that, while transcripts indicative of fetal Leydig cells (Ren1 and Crhr1), Leydig stem cells (Nestin, Pdgfrb, and Nr2f2) and progenitor adult Leydig cells (Akr1c14 and Srd5a1) were comparable between Nr5a1-Cre;Nedd4^Flox/Flox testes and controls at P60, transcripts representing immature and adult Leydig cells (Hsd17b3, Bhmt, and Sult1e1) were significantly reduced in the Nedd4-mutant testes (Fig. 5C), in line with a seeming reduction in the adult Leydig cell population. Consistent with defective steroidogenesis, Nr5a1-Cre;Nedd4^Flox/Flox mice displayed reduced anogenital distance (Fig. 5D and 5E, 11.2% reduction) and decreased seminal vesicle weight (Fig. 5F, 15.3% reduction), two androgen-dependent measures that rely heavily on steroid hormones produced by the testis (49, 50).

Finally, given the reduction in androgen-dependent measures in Nr5a1-Cre;Nedd4^Flox/Flox testes, and the critical role of adult Leydig cells in steroid hormone synthesis, we next assessed the expression of genes whose products contribute to the regulation of steroidogenesis. These included hormones (Insl3), hormone receptors (Ar and Lhr) and gene products involved in hormone biosynthesis (Star, Cyp11a1, and Cyp17a1). At P60, Nr5a1-Cre;Nedd4^Flox/Flox testes had a significant reduction in expression of Lhr and Cyp17a1, while expression of Ar, Star and Cyp11a1 remained unchanged. Surprisingly, however, mutant testes expressed Insl3 transcripts at levels nearly twice that of littermate controls (Fig. 5G). This provided compelling evidence that, beyond its role in Sertoli cell proliferation, NEDD4 contributes to the differentiation and function of the adult Leydig cell population within the postnatal testis.

Discussion

Here we report previously uncharacterized roles of the ubiquitin ligase NEDD4 in the developing testis. While Nr5a1-Cre-driven ablation of Nedd4 did not replicate the gonadal sex reversal observed upon constitutive loss of this enzyme, its deletion in testicular somatic cells resulted in a significant reduction in postnatal testis weight. This phenotype stemmed from impaired Sertoli cell proliferation and defective adult Leydig cell differentiation.

Inefficient Ablation of Nedd4 in Nr5a1-Cre;Nedd4^Flox/Flox Testes

Despite targeting Nedd4 deletion in Nr5a1-positive cells, which are evident in the developing genital ridge as early as 10.2 dpc (37), Nr5a1-Cre;Nedd4^Flox/Flox mice did not phenocopy the gonadal sex reversal observed in Nedd4^−/− mice (24). Investigation into NEDD4 ablation efficiency revealed that, while its expression was largely absent in most somatic cells within the developing gonad, it persisted in the coelomic epithelium until at least 14.5 dpc (Fig. S1C (36)). This is consistent with previous reports showing incomplete deletion of genes such as Fgfr2 (51), Gata4, and Fog2 (52), and Gata4 and Gata6 (53) when utilizing the same Nr5a1-Cre mouse line (27). Despite continued expression of the deleted gene product within the coelomic epithelium, sex reversal phenotypes have been reported for Fgfr2, Gata4, and Sox9 (51, 52, 54). These sex reversals, however, are often incomplete and are characterized by an inability to maintain the Sertoli cell fate. In contrast, earlier ablation of these genes with alternative Cre-drivers, causes complete sex reversal, often as a result of proliferation defects within the genital ridge, presumably resulting in an insufficient number of Sertoli cells to maintain the testicular fate (51, 52).

The complete sex reversal in XY Nedd4^−/− gonads, but the absence of such a phenotype in Nr5a1-Cre;Nedd4^Flox/Flox mice, supports the hypothesis that NEDD4 is essential for the proliferation of the coelomic epithelium and the establishment of the Sertoli cell stock, rather than for maintaining the male fate. This aligns with its oncogenic potential and its function in promoting Sertoli cell proliferation (Fig. 3).

Adrenal Cortex Dysgenesis in Nr5a1-Cre; Nedd4^Flox/Flox Mice

The adrenal dysgenesis observed in Nr5a1-Cre;Nedd4^Flox/Flox mice is not unexpected, given the shared developmental origins of the adrenal cortex and the genital ridges, the male to female sex reversal observed in XY Nedd4^−/− mice, and the similar adrenal and gonadal phenotypes reported in Nr5a1 (55), Wt1 (56), Cbx2 (57), Cited2 (58), Pbx1 (59), Odd1 (60), Igf1r/Insr (61), and Six1/4 (62, 63) mutant mice. The same Cre driver has successfully generated adrenal gland phenotypes with other floxed alleles (41, 64-66).

Although Nedd4 deletion in the adrenal cortex of Nr5a1-Cre;Nedd4^Flox/Flox mice may be incomplete, as seen in the genital ridges, complete ablation of Nedd4 may not be required to induce adrenal cortex dysgenesis. The adrenal cortex is highly sensitive to gene dosage, particularly for Nr5a1. Complete loss of this gene results in adrenal agenesis (55 , 67-69 ), while heterozygous mice develop adrenal glands 12-fold smaller than WT littermates (70). Indeed, given the near 50% reduction in Nr5a1 transcripts at 11.5 dpc in XY Nedd4^−/− gonads (24), it is likely that Nr5a1 levels were sufficiently reduced in Nr5a1-Cre;Nedd4^Flox/Flox adrenal cortexes to cause adrenal dysgenesis, such is the case in Cbx2- and Cited2-deficient mice (57, 58). Interestingly, adrenocortical cell proliferation and steroidogenic activity increase later in development upon Nr5a1 haploinsufficiency, perhaps as a means of compensating for earlier developmental defects (70-72 ), ultimately allowing fully functional adrenal gland in adulthood (73). This could also be the case in Nr5a1-Cre;Nedd4^Flox/Flox mice, given the progressive increase in the adrenocortical cell population between 14.5 and 18.5 dpc (Fig. S2B (36)). This may explain why these mice survive into adulthood, whereas more severe adrenocortical deficits in other models of adrenal dysgenesis results in early postnatal lethality (41, 65).

NEDD4 Promotes Sertoli Cell Proliferation

The absence of sex reversal in Nr5a1-Cre;Nedd4^Flox/Flox mice provided a unique opportunity to explore the role of NEDD4 in the somatic cells of the postnatal testis, a function that remains largely uncharacterized due to the perinatal lethality (22) and gonadal sex reversal observed in XY Nedd4^−/− mice.

The establishment of an adequate number of Sertoli cells in the mammalian testis is largely dependent on the length of the proliferative phase and the rate of division of these cells during development. In Nr5a1-Cre;Nedd4^Flox/Flox testes, Sertoli cells exhibited reduced proliferative potential as early as 15.5 dpc (Fig. 3), likely contributing to the reduction in postnatal testis-weight. Testis weight in Nr5a1-Cre;Nedd4^Flox/Flox were reduced by 48.6% and 47.8% by P14 and P60, respectively (Fig. 3B and 3G). However, in Amh-Cre;Nedd4^Flox/Flox, testis weight was unchanged at P15 and only 29.7% reduced by P60 (Fig. 4B). This likely stems from the timing of Nedd4 ablation. The Nr5a1-Cre line deletes genes as early as 11.5 dpc, whereas Amh-Cre ablation occurs around 15 dpc (27, 31). Consequently, Nr5a1-Cre;Nedd4^Flox/Flox Sertoli cells experience earlier disruptions in proliferation than those of Amh-Cre;Nedd4^Flox/Flox Sertoli cells, reducing the total rounds of division, and leading to a more pronounced reduction of Sertoli cell number and testis weight. Interestingly, Nr5a1-Cre;Nedd4^Flox/Flox testes were already smaller by 12.5 dpc (Fig. 1), suggesting that this reduction in proliferative potential arose early in testis differentiation, though not early enough to cause gonadal sex reversal. Furthermore, the additional defects in adult Leydig cell differentiation (Fig. 5) may further exacerbate the reduced testis size in Nr5a1-Cre;Nedd4^Flox/Flox mice compared with Amh-Cre;Nedd4^Flox/Flox mice.

The role of NEDD4 in Sertoli cell proliferation aligns with its previously reported contributions to genital ridge proliferation (24), and its purported role as an oncogene (43-45). Our findings indicate that NEDD4 modulates the PI3K-AKT signaling pathway, which was diminished in Nedd4-mutant testes (Fig. 3J and 3K). Given that numerous signaling cascades converge on AKT activation, many of which are critical for Sertoli cell proliferation and homeostasis (74), NEDD4 likely facilitates testis growth via this pathway. Diminished AKT activation in Nr5a1-Cre;Nedd4^Flox/Flox testes, however, occurs in the absence of ectopic PTEN, a NEDD4 substrate known to drive similar phenotypes upon loss of Nedd4. These observations add to the complexity of the NEDD4 and PTEN relationship. While some in vivo studies supports negative regulation of PTEN by NEDD4, with PTEN levels increasing upon Nedd4 deletion (75-77) and decreasing with ectopic NEDD4 expression (77), PTEN stability and localization remain largely unaffected in various Nedd4-deficient tissues, including fibroblasts (22, 78), T cells (79, 80), neurons (81), and spermatogonial stem cells (23 ). Our study (Fig. 3J), similarly found no change in PTEN expression, suggesting that the regulation of PTEN by NEDD4 may be context dependent or that redundant mechanisms maintain PTEN stability and localization in the absence of NEDD4. Indeed, this may be true of many in vitro identified NEDD4 substrates as known substrates Beclin1, p62, LC3, and PDCD6IP (21, 82) remain unchanged in Nedd4-mutant testes (Fig. S3 (36)).

Beyond its role in PTEN regulation, NEDD4 is required for animal growth and this has been linked to the control of IGF-1 and insulin signaling via the PI3K-AKT pathway. Ubiquitination of GRB10 and IRS2 by NEDD4 have been implicated in such regulation (22, 83). In keeping with this, insulin and IGF-1 signaling is diminished upon loss of Nedd4. Phenotypes resembling those of Nedd4-mutants arise upon dual ablation of the insulin receptor (Insr) and insulin-like growth factor 1 receptor (Igf1r). In this way XY Insr^−/−;Igf1r^−/− mice exhibit diminished proliferation of somatic progenitors within the genital ridges, XY gonadal sex reversal and perturbed adrenal development (61). Insr;Igf1r-mutants also exhibit significant reductions in testis weight, owing to reduced Sertoli cell proliferation (20), though this is rescued upon ablation of Pten (46). Given the known regulation of insulin and IGF-1 signaling by NEDD4, and the importance of this signaling axis across multiple levels of testis differentiation, perturbed insulin and IGF-1 signaling may underlie the phenotypes observed in Nedd4 mutants, particularly given the marked reduction in pAKT observed in mutant testes (Fig. 3). While the precise mechanism of NEDD4's control over the PI3K-AKT axis in murine Sertoli cells remains to be elucidated, NEDD4 is known to interact with multiple components along this signaling cascade (21, 82). Investigating how NEDD4 modulates these interactions should prove illuminating to further our understanding of Sertoli cell dynamics.

NEDD4 Regulates Adult Leydig Cell Differentiation and Steroidogenesis

Beyond its roles in adrenal cortex development and Sertoli cell proliferation, NEDD4 is essential for proper adult Leydig cell differentiation. In Nr5a1-Cre;Nedd4^Flox/Flox testes, the transition from progenitor to immature adult Leydig cells was notably disrupted (Fig. 5C). While testosterone levels were not measured in these mice, a defect could be observed with respect to reduced androgen-sensitive parameters, such as anogenital distance and seminal vesicle weight (Fig. 5D-5F), alongside reduced expression of genes encoding steroidogenic enzymes (Fig. 5G).

Interestingly, while other models of adult Leydig cell dysfunction show a near absence of Cyp11a1, Cyp17a1, Star, Insl3, and Lhr (84), of these, only Cyp17a1 and Lhr were significantly decreased in Nr5a1-Cre;Nedd4^Flox/Flox testes, suggesting a more nuanced role for NEDD4 in steroidogenesis. Notably, Star and Cyp11a1 act early in steroidogenesis. STAR regulates cholesterol transport into mitochondria while CYP11A1 converts cholesterol to pregnenolone (85). CYP17A1 acts downstream of these proteins, ultimately aiding in the conversion of pregnenolone and progesterone to their major products: estradiol, testosterone, DHEA and cortisol (86). Given the observed expression data, it is likely that NEDD4, if it were intrinsically involved in this process, would exert its influence downstream of STAR and CYP11A1 and upstream of CYP17A1. Hsd17b3 transcripts are also diminished in Nedd4-mutant testes which, itself, acts downstream of CYP17A1, further supporting this hypothesis.

Disrupted LHR Signaling Contributes to Impaired Adult Leydig Cell Differentiation

The downregulation of Lhr in Nr5a1-Cre;Nedd4^Flox/Flox testes provides further insight into the Leydig cell phenotype observed upon Nedd4 deletion. LHR signaling is essential for adult Leydig cell differentiation, yet dispensable for the development of fetal Leydig cells. While Lhr is expressed in fetal Leydig cells (87), Lhr-deficient mice exhibit normal Leydig cell development at fetal and neonatal stages (88). In contrast, adult Lhr^−/− mice exhibit significant reductions in Leydig cell numbers, anogenital distance, testis weight and seminal vesicle weight (88, 89). Surviving Leydig cells in Lhr-deficient testes are also functionally immature, as evidenced by a decrease in cytoplasmic volume (90). These observations align with the unaffected differentiation of fetal Leydig cells in Nr5a1-Cre;Nedd4^Flox/Flox testes (Fig. 2) and the postnatal Leydig cell dysfunction observed later (Fig. 5). This suggests that impaired LHR signaling contributes to the Leydig cell defects observed upon ablation of Nedd4. However, since Lhr levels are only reduced by approximately 28% in Nr5a1-Cre;Nedd4^Flox/Flox testes, rather than absent in Lhr-deficient mice, the Leydig cell phenotype in Nr5a1-Cre;Nedd4^Flox/Flox is comparatively milder. While a thorough characterization of Lhr heterozygous mice is lacking, in which Lhr transcript levels are significantly reduced, the serum hormone profile of these mice is altered and Leydig cells do appear morphologically different to WT littermates (88, 89), perhaps suggesting that a mere reduction in Lhr expression is sufficient to induce Leydig cell associated phenotypes, as is observed in this study. Cyp17a1 expression has also been shown to be luteinizing hormone (LH) dependent, yet independent of testosterone, unlike other genes encoding steroidogenic components (91), suggesting that the reduction in Cyp17a1 may merely be a consequence of reduced Lhr expression. This may further explain why Cyp17a1 is reduced in Nr5a1-Cre;Nedd4^Flox/Flox testes while Cyp11a1 and Star remain unchanged (Fig. 5G).

Curiously, despite the observed downregulation of Cyp17a1 and Lhr, Nr5a1-Cre;Nedd4^Flox/Flox testes showed a significant increase in Insl3 expression. This finding is intriguing, as INSL3 is widely recognized as a biomarker of Leydig cell differentiation and function (92, 93). It is perhaps paradoxical that a mouse model exhibiting perturbed Leydig cell function exhibit increased levels of Insl3 transcripts. One possible explanation is that the reduction in Lhr expression triggers a compensatory increase in circulating LH. Indeed, gonadotropins (LH and human chorionic gonadotropin) are known to upregulate Insl3 expression (94) and Lhr deficient mice have increased circulating LH (88, 89, 95). In this way, it is feasible that the observed reduction in Lhr in Nr5a1-Cre;Nedd4^Flox/Flox testes results in a compensatory increase in circulating LH which subsequently upregulates Insl3.

How increased levels of Insl3 affect Leydig cell function, however, remains to be seen. While transgenic Insl2/Insl3 expression was capable of rescuing the cryptorchidism of Insl3^−/− mice, the transient increase in Insl3 expression in transgenic mice had no impact on male fertility (96), and no testicular abnormalities were reported.

In conclusion, this study highlights the distinct and multifaceted role of the ubiquitin ligase NEDD4 in murine testis development. We demonstrate that NEDD4 plays a crucial role in the early establishment of the adrenogonadal primordium, with its retention in the gonadal coelomic epithelium being sufficient to rescue the gonadal sex reversal observed in constitutive knockouts. However, this is not the case in the adrenal cortex, where Nedd4 ablation results in adrenal dysgenesis, underscoring the differential sensitivity of these tissues to Nedd4 ablation. Consistent with its reported role as an oncogene, we have shown that NEDD4 promotes Sertoli cell proliferation, through the modulation of the PI3K-AKT signaling pathway. NEDD4 also ensures proper differentiation of adult Leydig cells and may contribute to murine steroidogenesis. Overall, this study establishes that NEDD4 functions at multiple levels to orchestrate testis development, reinforcing the importance of ubiquitination, and post-translational modifications in the differentiation of the mammalian testis.

Abbreviations

AMH

anti-Müllerian hormone

CASA

computer-assisted sperm analysis

dpc

days post coitum; Igf1r, insulin-like growth factor 1 receptor; Insr, insulin receptor

LH

luteinizing hormone; NEDD4, The Neural precursor cell Expressed Developmentally Down-regulated protein 4

P

postnatal day; PTEN, phosphatase and tensin homolog

RT-qPCR

quantitative real-time RT-PCR; SOX9 = SRY-box containing gene 9; SRY, sex-determining region on the Y chromosome

TBST

Tris buffered saline supplemented with 0.1% Tween 20 detergent

ts

tail somite stage

WT

wildtype

Funding

This work was supported by an Australian Research Council Discovery Grant (150101448) awarded to D.W. and S.K. and partly supported by a Monash University Biomedicine Discovery Institute Early Career Award awarded to S.W. S.K. was supported by an Australian National Health and Medical Research Council Investigator Grant (GNT 2007739). S.N. and Y.N. received funding from the Swiss National Science Foundation under grants 31003A_173070 and 310030_200316.

Disclosures

The authors have nothing to disclose.

Data Availability

The data that support the findings of this study are available in the Materials and Methods and/or Results section of this article. Supplementary Materials are available at Figshare, under the collection “Supplementary Data for Windley et al (2025) ‘NEDD4 Promotes Sertoli Cell Proliferation and Adult Leydig Cell Differentiation in the Murine Testis’’ (36).