Introduction

Leydig cells are the primary source of androgens, producing approximately 95% of circulating testosterone to sustain spermatogenesis and male fertility [1, 2]. Leydig cells originate from stem Leydig cells (SLCs), which differentiate into steroidogenic adult Leydig cells (ALCs) under paracrine regulation [3, 4]. Steroidogenic factor 1 (Sf1), an orphan nuclear receptor transcription factor, serves as a master regulator of this process by activating steroidogenic enzymes (e.g., cytochrome P450, family 11, subfamily C, polypeptide 1 [Cyp11c1], 3β-hydroxysteroid dehydrogenase [3β-HSD]) to drive ALC maturation [5]. In mice, mutations of Sf1 exhibit arrested testicular development and abolished steroidogenesis [6–8], while human SF1 mutations are associated with 46,XY disorders of sexual development (DSD), such as Swyer syndrome [9]. Similarly, in Nile tilapia (Oreochromis niloticus), our previous work reveals that sf1 mutation disrupts testicular morphogenesis and downregulates the expression of key steroidogenic enzymes such as cyp11c1 [10]. Despite its established central role, the upstream signaling pathways that regulate Sf1 to commit the Leydig lineage remain poorly defined.

Hedgehog (Hh) signaling pathway is an evolutionarily conserved developmental regulator [11]. Canonical Hh signaling involves the binding of ligands to Patched (Ptch) receptors, which relieves the suppression on Smoothened (Smo) and ultimately activates Glioma-associated oncogene homolog (Gli) transcription factors to regulate target genes [12–14]. In vertebrates, there are three Hh ligands (i.e. Sonic Hh, Indian Hh and Desert Hh [Dhh]), two Ptch receptors (Ptch1 and Ptch2) and three Gli homologs (Gli1-3) [15–18]. A large number of studies have well demonstrated that Dhh signaling pathway is critical in gonadal development [14, 19]. Dhh, secreted by Sertoli cells, is critical for Leydig cell commitment in mammals [20, 21]. In mice, mutations of Dhh lead to Leydig cell dysfunction, testosterone insufficiency, and infertility [22], while human DHH mutations are associated with 46,XY DSD [23]. Strikingly, these phenotypes mirror those caused by Sf1 deficiencies, suggesting a potential regulatory nexus between Dhh and Sf1 signaling. However, the molecular circuitry connecting Dhh signaling to Sf1 activation remains elusive.

Studies demonstrate functional divergence between Ptch1 and Ptch2. For instance, Ptch1 knockout causes embryonic lethality due to constitutive Hh activation, whereas Ptch2 deficiency yields no overt defects [24, 25]. During testicular development, Ptch1 displays high expression at early stage critical for gonadal primordium formation, while Ptch2 shows testis-enriched dynamic regulation [26, 27]. However, the distinct roles of Ptch1 and Ptch2 in Leydig lineage commitment remain unknown to date. Among Gli transcription factors, Gli1 acts as a primary activator, while Gli2 and Gli3 exhibit dual regulatory functions [28, 29]. In murine testes, co-expression of Gli1 and Gli2 in Leydig cells illustrates functional redundancy—single knockouts show normal development, but combined pharmacological inhibition (GANT61) abolishes Leydig cells and spermatogenesis [30]. These findings underscore the complexity of Gli-mediated transcription in the testis and highlight the need to delineate the specific roles of individual Gli in SLC differentiation.

Nile tilapia (Oreochromis niloticus) is not only an important farm fish in global aquaculture but also a distinguished animal model for investigating gene function and endocrine regulation. This model organism possesses multiple advantageous attributes, including an XX/XY sex-determination system, early sexual maturity (approximately 3 months after hatching for males and 6 months after hatching for females), a short spawning cycle of approximately 14 days, year-round reproductive capacity under laboratory conditions, well-characterized sex-linked genetic markers, and a body size amenable to physiological sampling [31, 32]. Furthermore, a stem Leydig cell line (TSL) has been established from the testis of a 3-month-old Nile tilapia [33]. TSL expresses platelet-derived growth factor receptor α (pdgfrα), nestin, and chicken ovalbumin upstream promoter transcription factor II (coup-flla), which are usually considered as SLC-related markers in several other species [2, 34, 35]. Notably, this cell line exhibits the capacity to differentiate into 11-ketotestosterone (11-KT)-producing Leydig cells both in vitro and in vivo [33]. When cultured in a defined induction medium, TSL cells differentiate into a steroidogenic phenotype, expressing key steroidogenic genes including star1, star2, and cyp11c1, and producing 11-KT; upon transplantation into recipient testes, TSL cells successfully colonize the interstitial compartment, activate the expression of steroidogenic genes, and restore 11-KT production. In this study, by combining CRISPR/Cas9-mediated gene knockout with TSL transplantation technology, we unveil a Dhh-Ptch2-Gli1-Sf1 axis that is essential for Leydig cell lineage differentiation. Our findings provide fundamental insights into the endocrine regulation of Leydig cell development.

Results

Mutation of dhh disrupts testicular organization and androgen synthesis

The Nile tilapia Dhh open reading frame (GenBank ID: 100707022, 1377 bp) was cloned and validated. The encoded 458-amino acid protein shares over 56% identity with vertebrate homologs (Fig. S1A) and clusters phylogenetically with mammalian DHH orthologs (Fig. S1B). Transcriptomic analyses conducted on various tissues and testes across different developmental stages revealed that dhh expression peaked in the testes, particularly at 90 days after hatching (dah) (Fig. S1C, D).

CRISPR/Cas9 targeting of dhh exon 1 from Nile tilapia generated homozygous mutants (dhh^-/-) harboring a 13-bp frameshift deletion (Fig. S2A–D), resulting in premature termination and loss of the conserved Hedge domain (Fig. S2E). dhh mRNA levels in dhh^-/- fish were significantly reduced compared to WT (Fig. S2F). At 90 dah, dhh^-/- testes exhibited marked atrophy (Fig. 1A, B) and disorganized histology, with sparse spermatogonia, rare spermatids, and collapsed interstitial architecture (Fig. 1C–D’, K). Immunofluorescence confirmed significantly reduced germ cell (Vasa⁺), meiotic cell (Sycp3⁺), and Leydig cell (Cyp11c1⁺) populations (Fig. 1E-J’, L-N). Serum 11-KT levels in dhh^-/- fish were significantly lower than WT (Fig. 1O), indicating impaired androgen synthesis.

Fig. 1.

The differentiation of SLCs cannot be rescued by 11-KT, but by SAG

Given the established role of 11-KT in teleost testicular development and spermatogenesis [36], we first asked whether exogenous 11-KT could ameliorate the testicular defects in dhh^-/- mutants. Compared to untreated mutants, dhh^-/- individuals exposed to 11-KT exhibited a more organized testicular architecture, with significantly increased populations of Vasa⁺ and Sycp3⁺, as evidenced by both H&E staining and immunofluorescence (Fig. 2A–C’, E–G’, I–K’, R, S). This improvement was further supported by a recovery in GSI (Fig. 2Q). However, the population of Cyp11c1-positive Leydig cells remained profoundly depleted in dhh^-/- +11-KT testes (Fig. 2M–O’, T). Taken together, these data indicate that 11-KT treatment partially restored the development of germ cells but not Leydig cell lineage, implying that the Leydig cell defect is androgen-independent.

Fig. 2.

We next investigated whether direct activation of the Dhh pathway could rescue the dhh^-/- phenotype. Treatment of dhh^-/- mutants with SAG, a Hedgehog agonist, significantly rescued testicular development. This was marked by a significant recovery of germ cells (Vasa⁺), spermatocytes (Sycp3⁺), and critically, Leydig cells (Cyp11c1⁺), alongside a restoration of testicular histology (Fig. 2D–D’, H–H’, L–L’, P–P’, Q–T). These results demonstrate that dhh mutation does not disrupt survival of endogenous Leydig cell lineage, but directly impairs their differentiation, leading to secondary impairment in androgen production and germ cell development.

Mutation of dhh blocks SLCs differentiation in vivo

To dissect Leydig cell lineage impairment in dhh^-/- testes, we transplanted the TSL labeled with PKH26 (a fluorescent red hydrophobic membrane dye that enables tracking of transplanted cells) into WT and dhh^-/- testes (Fig. 3A). In WT recipients, TSL cells colonized the interstitium and differentiated into Cyp11c1 positive steroidogenic cells (Fig. 3B1–B3). In dhh^-/- testes, TSL engrafted equivalently but failed to express Cyp11c1 (Fig. 3C1–C3), indicating a differentiation-specific block. Rescue experiments showed that the TSL overexpressing tilapia Dhh (TSL-OnDhh) or treated with SAG (TSL+SAG) restored Cyp11c1 expression in dhh^-/- testes (Fig. 3D1–E3, F) and a significant increase in endogenous 11-KT levels (Fig. 3G). These results suggest that SLC differentiation is inhibited in dhh^-/- XY tilapia testes. In addition, the survival and engraftment of PKH26-labeled TSL cells were unaffected in the dhh^-/- environment (Fig. 3C3). Collectively, these results demonstrate that Dhh signaling is specifically required for the differentiation of SLCs into steroidogenic cells, but is dispensable for their survival or initial recruitment into the testicular niche.

Fig. 3.

Ptch2 is the functional receptor for Dhh signaling in SLCs

RT-PCR results showed that both ptch1 and ptch2 were expressed in TSL (Fig. S3). Fluorescence in situ hybridization (FISH) and IF revealed co-expression of ptch1 and ptch2 in Cyp11c1 positive Leydig cells (Fig. 4A). To identify the Dhh-driving receptor selectivity in SLCs, we established ptch1^-/- and ptch2^-/- TSL cell lines (Fig. S4A, B) and evaluated pathway activity using an 8×GLI luciferase reporter system (Fig. 4B). In WT TSL cells, the 8×GLI reporter demonstrated substantial luciferase activity that was markedly enhanced by OnDhh overexpression (Fig. 4C), confirming Dhh responsiveness in this system. Intriguingly, ptch1^-/- cells maintained Dhh-induced luciferase activation comparable to WT controls, whereas ptch2^-/- cells completely lost this responsiveness (Fig. 4D). This differential receptor requirement implies that Ptch2 likely acts as the functional receptor for transducing Dhh signals in TSL cells.

Fig. 4.

To investigate the function of Ptch2 in vivo, we generated ptch2^-/- and dhh^-/-;ptch2^-/- mutant Nile tilapias (Fig. S5A-E). Compared to WT fish at 90 dah, ptch2^-/-testes exhibited no significant differences in testicular histology (Fig. 4E–F’), GSI (Fig. 4Q), germ cell (Vasa⁺, Sycp3⁺) and Leydig cell (Cyp11c1⁺) populations (Fig. 4H-I’, K-L’, N-O’, R-T), and serum 11-KT levels (Fig. 4U). Strikingly, the loss of ptch2 in the dhh^-/- background (dhh^-/-;ptch2^-/-) fully rescued the testicular defects, resulting in normal histology (Fig. 4G–G’), Leydig cell differentiation (Fig. 4P–P’), and serum 11-KT levels (Fig. 4U). This genetic rescue is consistent with Ptch2 functioning as the inhibitory receptor for Dhh; its removal alleviates the suppression on the pathway, thereby bypassing the requirement for the Dhh ligand.

Gli1 is the principal transcriptional effector of Dhh signaling in SLCs

In the Hh pathway, Gli transcription factors (Gli1/2/3) mediate diverse developmental processes (26). RT-PCR results showed that gli1, gli2, and gli3 were all expressed in TSL (Fig. S3). FISH and IF revealed co-expression of gli1, gli2, and gli3 in Cyp11c1 positive Leydig cells (Fig. 5A). To identify Dhh-specific effectors, we generated gli1^-/-, gli2^-/-, and gli3^-/- TSL lines (Fig. S4C–E) and assessed pathway activity using the 8×GLI luciferase reporter. While TSL-gli2^-/- and TSL-gli3^-/-cell lines maintained Dhh-induced luciferase activation comparable to WT controls, the TSL-gli1^-/- cell line completely lost this responsiveness (Fig. 5B), implying that Gli1 is the principal transcriptional effector for Dhh signaling transduction in SLCs.

Fig. 5.

Sf1 is the critical downstream effector of Gli1 in SLCs differentiation

To identify key targets of the Dhh-Gli1 axis, we performed transcriptomic profiling of TSL cells under conditions of pathway activation: Dhh overexpression (TSL-OnDhh), Gli1 overexpression (TSL-OnGli1), and SAG treatment (TSL+SAG). Comparative RNA-seq analysis identified a core set of 33 genes consistently upregulated across all three conditions (Fig. 5C, S6A). Among these, sf1 emerged as a prominently upregulated candidate. Functional annotation of its promoter region identified two conserved Gli1-binding motifs (Fig. 5D). A dual-luciferase reporter assay confirmed that Gli1 potently transactivates the sf1 promoter (Fig. 5E). The specificity of this interaction was further demonstrated by a cold probe competition assay: a wild-type oligonucleotide containing the Gli-binding motif (GACCACCCA) competitively inhibited Gli1-driven sf1 transactivation in a dose-dependent manner, whereas a mutated version (TTAATTAAA) had no effect (Fig. 5E).

The functional necessity of Sf1 was confirmed by transplantation assays using sf1^-/- and Sf1-overexpressing (TSL-OnSf1) TSL cells. While TSL-sf1^-/- cells failed to differentiate into Cyp11c1-positive cells even in WT testes, TSL-OnSf1 cells successfully differentiated and expressed Cyp11c1 even in the dhh^-/- testicular environment (Fig. 5F–H). These results definitively position Sf1 as the critical downstream effector executing the Dhh-Ptch2-Gli1 signaling cascade during SLCs differentiation.

Discussion

This study delineates a critical Dhh-Ptch2-Gli1-Sf1 signaling axis essential for SLCs differentiation in Nile tilapia. By combining dhh^-/- XY models with SLCs transplantation, we definitively demonstrate that Dhh signaling is dispensable for Leydig cell lineage survival but indispensable for its differentiation. Furthermore, our integrated evidence from targeted gene knockout and functional assays suggests that Ptch2 acts as the primary receptor for Dhh signaling in SLCs, and points to Gli1 as the key transcriptional effector mediating the transactivation of sf1. Collectively, these findings implicate Sf1 as the critical downstream target bridging Dhh signaling to steroidogenic maturation, thereby resolving key ambiguities in Leydig cell lineage development and providing evolutionary insights into vertebrate reproductive biology.

Consistent with its established role in mammals, our data in Nile tilapia confirm that Dhh function in male testis development is evolutionarily conserved across vertebrates. While mutations in Dhh lead to testosterone insufficiency and infertility in mice [22] and are associated with 46,XY DSD in humans [23], the precise cellular mechanism remained unresolved. Our study demonstrates that dhh^-/- tilapia phenocopy these defects, exhibiting testicular dysgenesis, a profound deficiency of Leydig cells, and drastically reduced 11-KT levels (Fig. 1). To define the onset of Leydig cell differentiation, we performed a developmental time-course analysis. This revealed that Cyp11c1-positive steroidogenic cells first appear in wild-type testes at 30 dah, while being conspicuously absent in dhh^-/- mutants at this same stage (Fig. S7). This clear temporal pattern establishes ∼30 dah as the developmental window when SLCs initiate their differentiation program in the Nile tilapia. The functional indispensability of Dhh at this specific stage was further confirmed by rescue experiments, in which both SAG treatment and transplantation of Dhh-activated TSL cells restored Cyp11c1-positive cell populations in dhh^-/- testes (Fig. 3). Collectively, our findings position Dhh as a crucial niche signal that acts at a defined developmental checkpoint to drive the differentiation of SLCs into steroidogenic Leydig cells.

The functional divergence of Ptch1 and Ptch2 in Hh signaling represents an intriguing aspect of pathway regulation. Structurally, both receptors share conserved 12-transmembrane domains vital for cholesterol transport but diverge in their cytoplasmic C-terminal regions, which are implicated in differential Smoothened inhibition efficiency [37, 38]. In mammals, Ptch1 serves as the primary, broad-spectrum regulator of Hh signaling during embryogenesis, whereas Ptch2 often assumes context-dependent roles, such as in primordial germ cell migration via its co-receptor Gas1 [27]. However, their potential redundancy or specificity within the Leydig cell lineage has remained an open question. Our study provides significant insight into this issue. We confirmed that both ptch1 and ptch2 are expressed in Leydig cells (Fig. 4A). Functional interrogation using TSL cell lines revealed a marked difference: genetic ablation of ptch2 profoundly attenuated Dhh-induced signaling, as measured by a Gli-responsive luciferase reporter, whereas loss of ptch1 had no detectable effect under the same conditions (Fig. 4B-D). This in vitro data, pointing to a primary role for Ptch2, was further supported by in vivo genetic interaction studies. The observation that the dhh^-/- phenotype was fully rescued in dhh^-/-;ptch2^-/- double mutants (Fig. 4E-U) is consistent with Ptch2 acting as the inhibitory receptor whose loss alleviates pathway suppression. The apparent specificity for Ptch2 in this context might stem from unique co-receptor interactions or expression patterns within the testicular niche. Nonetheless, the early embryonic lethality of global ptch1 mutation in tilapia [39] precluded a direct assessment of its potential role in postnatal testis development. Therefore, while our findings strongly nominate Ptch2 as the principal receptor for Dhh in SLCs, a definitive exclusion of a role for Ptch1 will require future studies employing Leydig cell-specific conditional knockout models.

Gli is a nuclear transcription factor, characterized by the presence of a zinc finger domain located in the middle of the protein [40]. Downstream of the zinc finger domain is a phosphorylation cluster containing phosphorylation sites for protein kinase A and glycogen synthase kinase 3, which negatively regulate the transcriptional activity of Gli through phosphorylation [41]. In Drosophila, Ci is the sole Gli homolog, possessing both transcriptional activation and repression functions [42]. In vertebrates, there are three Gli homologs: Gli1, Gli2, and Gli3. Gli1 functions solely as a transcriptional activator because it lacks proteolytic processing domains. Gli2 has both activation and repression functions, while Gli3 primarily acts as a transcriptional repressor [43, 44]. However, there has been no research on the regulatory mechanism of Gli1/2/3 in SLCs differentiation. In this study, we found that although gli1, gli2, and gli3 are all expressed in Leydig cell lineage (Figure 5A-C), the results from TSL cell lines with mutations in gli1, gli2, and gli3 indicated that TSL-gli1^-/-, rather than TSL-gli2^-/-or TSL-gli3^-/-, failed to respond to Dhh signaling (Figure 5D). This suggests that Gli1 is the key transcription factor downstream of the Dhh signaling cascade in SLCs.

Sf1 and Dhh play important roles in Leydig cell commitment and testicular development [5, 11]. Studies suggest a potential interaction between these two factors. For example, in Dhh^-/- XY mice, testicular development is impaired, with a significant reduction in the number of Leydig cells and a marked downregulation of Sf1 expression [26]. In XX mouse ovaries where Smo is constitutively activated, a large number of ectopic Sf1-positive Leydig cells are generated [45]. Similarly, in sf1^-/- XY tilapia, testicular development is impaired, with a significant reduction in Leydig cell numbers and down regulation of dhh expression [10]. These findings imply that Dhh and Sf1 may interact through feedback loops or co-regulatory mechanisms. However, direct evidence for the regulation of Sf1 by the Hh signaling pathway remains lacking. This study elucidates the molecular mechanism of Sf1 as a critical downstream effect of the Dhh-Ptch2-Gli1 axis. Transcriptomic profiling revealed an increased expression of sf1 by Dhh regulation signaling in TSL cells (Fig. S6), identifying it as a key downstream target of Dhh signaling. Promoter analysis identified two conserved Gli-binding motifs (Fig. 5F) within the sf1 promoter, and luciferase assays confirmed that Gli1 transactivates sf1 transcription (Fig. 5G). Functional validation demonstrated that TSL-sf1^-/- cells failed to differentiate even in WT testes, whereas TSL-OnSf1 cells restored Leydig cell differentiation and steroidogenesis in dhh^-/- testes (Fig. 5H-J). These results establish Sf1 as both necessary and sufficient for Leydig cell lineage differentiation.

In conclusion, our study systematically deciphers the signaling cascade governing SLCs differentiation, from the niche signal to the core steroidogenic regulator. We definitively show that the Dhh signal from the niche is indispensable for SLCs differentiation, but not their survival. Within this pathway, Ptch2 serves as the critical receptor, relaying the signal through the transcription factor Gli1. The pivotal outcome of this cascade is the activation of sf1 by Gli1. Thus, we establish a complete Dhh-Ptch2-Gli1-Sf1 axis, resolving the long-standing question of how a key morphogen controls Leydig cell lineage development in vertebrates.

Materials and methods

Animals

Nile tilapias (Oreochromis niloticus) were kept in recirculating freshwater systems at 26°C under a 12-h light/dark photoperiod. All animal experiments were conducted in accordance with the regulations of the Guide for Care and Use of Laboratory Animals prescribed by the Committee of Laboratory Animal Experimentation at Southwest University, China (IACUC-20181015-12).

Generation of mutant lines by CRISPR/Cas9 in Nile tilapia

The sequences of tilapia dhh (GenBank ID: 100707022) and ptch2 (GenBank ID: 100692939) were retrieved from the NCBI. CRISPR/Cas9-mediated mutagenesis was performed as described previously [46]. Guide RNAs (gRNAs) targeting the first exon of dhh (GCGGGCCCGGTCCGCATCCC) and ptch2 (GTCCCAGGGGCCGGCGTATT) were designed using ZiFit (http://zifit.partners.org/ZiFiT/). gRNA and Cas9 mRNA were synthesized following established protocols [46]. The primers listed in Table S1. A mixture of synthetic gRNA (500 ng/μl) and Cas9 mRNA (1000 ng/μl) was microinjected into one-cell-stage fertilized embryos. Mutants were validated by polyacrylamide gel electrophoresis (PAGE) and Sanger sequencing. F0 chimeric XY males were crossed with WT XX females to generate heterozygous F1 offspring. F1 siblings carrying a 13-bp deletion in dhh or a 25-bp deletion in ptch2 were intercrossed to generate homozygous F2 mutants. For double mutants, dhh^+/- XY males were mated with ptch2^+/- XX females to obtain dhh^+/-;ptch2^+/- offspring. These double heterozygotes were then intercrossed to generate dhh^-/-;ptch2^-/- mutants.

Generation of mutant lines by pCas9-NtU6sgRNA system in TSL Cells

The tilapia stem Leydig cell line TSL was maintained in ESM4 medium as described [33]. Gene sequences for ptch1 (GenBank ID: 100700194), ptch2 (GenBank ID: 100692939), gli1 (GenBank ID: 100710687), gli2 (GenBank ID: 100704788), gli3 (GenBank ID: 100707331), and sf1 (GenBank ID: 100534561) were retrieved from NCBI. Knockout plasmids targeting these genes were designed following established protocols [47]. gRNAs were generated using ZiFit, with the following exon-specific targets: ptch1 (GGAGGCGCTCCTGCAGCACC, exon 3), ptch2 (GTCCCAGGGGCCGGCGTATT, exon 1), gli1 (TGACTCATGGATCAGGACCA, exon 2), gli2 (GATTCGCCTGTCACCCCACG, exon 3), gli3 (TGAGGAGCCCTCTACGTCTA, exon 2), and sf1 (GGCTGTGTACTGGTACTGGG, exon 4; designed as previously described) [47]. Primers are listed in Table S1. Plasmids were transfected into TSL cells using the MDMP E-1 microchannel transfection instrument (MENDGENE) with transwell cell compartments (BIOFIT), per manufacturer guidelines. Post-transfection (48 hours), cells were selected in ESM4 medium containing 500 μg/mL G418 for 7 days. Single-cell clones were isolated via dilution and serial passage, followed by genomic DNA (gDNA) extraction and Sanger sequencing to confirm mutations.

Histological analyses of testes from WT, dhh^-/-, ptch2^-/- and dhh^-/-;ptch2^-/- XY fish

GSI was calculated for WT, dhh^-/-, ptch2^-/- and dhh^-/-; ptch2^-/- XY fish at 90 dah as: (gonad weight / body weight) × 100%. Testes were fixed in Bouin’s solution for 24 hours at room temperature with agitation, dehydrated through an ethanol series, and embedded in paraffin. Sections (5 μm thickness) were stained with hematoxylin and eosin (H&E) and imaged using an Olympus BX53 microscope. Germ cell classification (spermatogonia, spermatocytes) followed established criteria [48].

Immunofluorescence

Testes from WT, dhh^-/-, WT+11-KT, dhh^-/-+11-KT, ptch2^-/- and dhh^-/-; ptch2^-/- XY fish at 90 dah were analyzed for germ cell marker Vasa, Leydig cell marker Cyp11c1, and meiosis marker Sycp3. Tissue fixation and embedding. Sections (5 μm) were deparaffinized, rehydrated, and subjected to antigen retrieval. After blocking with 5% bovine serum albumin (BSA; 30 minutes, room temperature), sections were incubated overnight at 4°C with rabbit polyclonal primary antibodies (Vasa, Cyp11c1, Sycp3; 1:1,000 dilution), validated previously [49]. Alexa Fluor 488-conjugated goat anti-rabbit secondary antibody (1:1,000; Invitrogen) was applied for 1 hour at 37°C, followed by DAPI nuclear counterstaining (1:1,000; Sigma-Aldrich). Fluorescence signals were imaged using an Olympus Fv3000 confocal microscope. For each biological replicate (n=5-6 fish per genotype), three non-serial, non-adjacent testis sections were analyzed. From each section, three representative fields of view were captured to ensure non-overlapping sampling. All positive cells number of Vasa, Sycp3 and Cyp11c1 was quantified by Image J Pro 1.51 software using default parameters.

Fluorescence in situ hybridization (FISH) and immunofluorescence

Testes from WT XY tilapia at 90 dah were fixed in 4% paraformaldehyde (Sigma) overnight at 4°C with agitation, paraffin-embedded, and sectioned (5 μm). FISH was performed as described [48]. Partial cDNA fragments of ptch1, ptch2, gli1, gli2, and gli3 were cloned into the pGEM-T Easy vector. Sense (control) and antisense RNA probes were synthesized using a DIG RNA Labeling Kit (Roche) and T7 mMESSAGE mMACHINE Kit (Ambion). Signals were amplified with the TSA^TM Plus TMR system (PerkinElmer) per manufacturer protocols (primers in Table S1). Concurrently, Leydig cell marker Cyp11c1 expression was assessed by immunofluorescence.

11-KT and SAG rescue experiments

To rescue dhh^-/- XY fish, offspring from dhh^+/- XY male × dhh^+/- XX female crosses were divided into two groups. The experimental group received 11-KT via immersion from 30 to 90 dah, with treatments refreshed every 2 days. 11-KT concentrations increased incrementally from 57 ng/L at 30 dah to 1254 ng/L by 90 dah (39.9 ng/L increase per treatment cycle; see Fig. 2, 5, S7). For SAG rescue, fish received intraperitoneal injections of 10 mg/kg SAG from 30 to 90 dah, with booster doses administered weekly. Testes were collected at 90 dah for morphological and histological evaluation.

TSL cell labeling, transplantation and IF

TSL, previously isolated from 3-month-old XY tilapia testes [33], was used to assess stem Leydig cell differentiation in WT and dhh^-/- XY fish. TSL cells were treated as follows: (1) untreated TSL, (2) TSL + 0.5 μM SAG (Hh pathway agonist), (3) TSL-OnDhh (overexpressing tilapia Dhh), (4) TSL-OnSf1 (overexpressing tilapia Sf1), and (5) TSL-sf1^-/- (sf1 mutation). For the SAG treatment experiment, TSL cells were incubated with 0.5 μM SAG for 48 hours before transplantation. Cells were labeled with PKH26 fluorescent dye (Sigma) and transplanted (5,000 cells per testis) into WT or dhh^-/- XY tilapia via the urogenital tract [33]. For TSL-OnDhh, the Dhh ORF (1377 bp) was amplified from gDNA, cloned into pcDNA3.1 (Invitrogen), and transfected into TSL cells (primers in Table S1). At 10 days post-transplantation, testes were analyzed by IF to quantify Cyp11c1 positive cells among PKH26 positive populations.

GLI report assay

A firefly luciferase reporter plasmid (8×GLI) was generated by cloning eight tandem copies of the GLI-binding motif (GACCACCCA) into the pGL4.23 vector (minimal promoter). To validate 8×GLI functionality, TSL cells were transiently transfected with: (1) pGL4.23 (negative control), (2) 8×GLI, or (3) 8×GLI plus Dhh-overexpression plasmid (OnDhh), for 48 hours. The 8×GLI construct was further transfected into TSL (WT), TSL-ptch1^-/-, TSL-ptch2^-/-, TSL-gli1^-/-, TSL-gli2^-/- or TSL-gli3^-/-, with or without OnDhh. Transfections included pRL-TK as an internal control. Luciferase activity was measured 48 hours post-transfection using the Dual Luciferase Assay System (Promega) and a Luminoskan™ Ascent luminometer (Thermo Fisher Scientific). Relative activity was calculated as firefly/Renilla luciferase ratios.

Western blot

The ORFs of Dhh (1377 bp) and Gli1 (4404 bp) were amplified from gDNA by PCR. Two tandem FLAG tag sequences (GATTATAAAGATGATGATAAA) were inserted flanking each ORF (after the start codon and before the stop codon) and cloned into the pcDNA3.1 vector (Invitrogen), generating pcDNA3.1-OnDhh and pcDNA3.1-OnGli1 (primers in Table S1). TSL cells were transiently transfected with pcDNA3.1-OnDhh or pcDNA3.1-OnGli1 for 48 hours. Protein expression was analyzed by Western blot following established methods [49]. Briefly, total protein lysates were separated by 15% SDS-PAGE, transferred to nitrocellulose membranes, and blocked with 5% BSA in TBST (10 mM Tris pH 7.9, 150 mM NaCl, 0.1% Tween-20) for 1 hour at 37°C. Membranes were incubated overnight at 4°C with rabbit anti-FLAG antibody (1:1000; Cell Signaling Technology), followed by HRP-conjugated goat anti-rabbit secondary antibody (1:1000; Invitrogen) for 1 hour at 37°C. Signals were detected using the BeyoECL Plus Kit (Beyotime) and imaged on a Fusion FX7 system (Vilber Lourmat).

Measurement of 11-KT levels by ELISA

Blood was collected from the caudal vein of 90 dah XY fish (WT, dhh^-/-, ptch2^-/-, dhh;ptch2^-/-, WT+TSL, dhh^-/-+TSL, dhh^-/-+TSL+SAG, dhh^-/-+TSL-OnDhh and dhh^-/-+TSL-sf1^-/-), incubated at 4°C overnight, and centrifuged to isolate serum. Tissue fluid samples from WT and dhh^-/- XY fish at 5, 10, 20, and 30 dah were similarly prepared. All samples were stored at −80°C until analysis. Serum and tissue fluid 11-KT levels were quantified using a 11-keto Testosterone ELISA Kit (Cayman) validated for teleost samples, per manufacturer instructions. Individual samples (serum or tissue fluid; n = 6 per genotype) were analyzed in triplicate on a single 96-well plate, with concentrations calculated from standard curves.

Transcriptome analyses

Tissue-and stage-specific expression of dhh was assessed using transcriptomic data from tilapia tissues (brain, heart, head kidney, kidney, liver, muscle, ovary, testis) and gonads at developmental stages spanning 5 to 360 dah [50, 51]. For global gene expression profiling, RNA was extracted from TSL-WT, TSL-OnDhh, TSL-OnGli1, and TSL + 0.5 μM SAG cells using RNAiso Plus (Takara). For the SAG treatment experiment, TSL cells were incubated with 0.5 μM SAG for 48 hours before collection. For each genotype, cells from three independent culture wells were pooled. Libraries were sequenced on an Illumina platform (GENEBOOK Biotechnology), and clean reads were aligned to the reference genome with HISAT2. Gene expression levels (FPKM: fragments per kilobase per million mapped reads) were quantified using featureCounts. Differentially expressed genes (DEGs) were identified for each condition (TSL-OnDhh, TSL-OnGli1, TSL+SAG) compared to TSL-WT controls using edgeR (threshold: FDR < 0.05, |log2(foldchange)| ≥ 1.5).

Real-time PCR

Testes from WT and dhh^-/- XY fish at 90 dah (n = 3 per genotype), along with TSL-WT, TSL-OnDhh, TSL-OnGli1, and TSL + 0.5 μM SAG cells, were analyzed for gene expression. Total RNA was extracted using RNAiso Plus (Takara), reverse-transcribed into cDNA (PrimeScript II 1st Strand cDNA Synthesis Kit; Takara), and amplified by real-time PCR on an ABI-7500 system (Applied Biosystems). Relative mRNA levels were normalized to β-actin and calculated via the formula R = 2^−ΔΔCt (primers in Table S1).

Dual luciferase report assay

The 5015-bp promoter region upstream of the sf1 start codon was amplified and cloned into the pGL3-basic vector (Promega), generating pGL3-sf1 (primers in Table S1). To assess Gli1 binding to the sf1 promoter, HEK293 cells were co-transfected with pGL3-sf1, pcDNA3.1-OnGli1, cold probe (GACCACCCA), and mutant cold probe (TTAATTAAA) using the TransIT-X2 transfection reagent (Mirus Bio). Control transfections included empty pcDNA3.1 and pGL3 vectors, with pRL-TK for normalization. Luciferase activity was measured 48 hours post-transfection.

Statistical analyses

Data are presented as mean ± SD from at least three independent replicates. A two-tailed independent Student’s t-test was used to determine the differences between the two groups. One-way ANOVA, followed by Tukey multiple comparison, was used to determine the significance of differences in more than two groups. P < 0.05 was used as a threshold for statistically significant differences.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request

Acknowledgements

This research was funded by the National Key Research and Development Program of China (Grant No. 2022YFD1201600), the National Natural Science Foundation of China (Grant No. 32172969, 31972776, 32102780, 32473159).

Additional information

Author contributions

Changle Zhao: Writing – original draft, Software, Resources, Methodology, Investigation, Formal analysis, Data curation. Yongxun Chen: Writing – original draft, Methodology, Data curation. Lei Liu: Data curation, Conceptualization. Xiang Liu: Software, Investigation. Hesheng Xiao: Software, Data curation. Feilong Wang: Investigation, Formal analysis. Qin Huang: Resources, Software. Xiangyan Dai: Resources, Investigation. Wenjing Tao: Resources, Funding acquisition. Deshou Wang: Supervision, Investigation, Funding acquisition. Jing Wei: Writing – review & editing, Validation, Supervision, Project administration, Funding acquisition, Data curation, Conceptualization.

Funding

MOST | National Key Research and Development Program of China (NKPs) (2022YFD1201600)

• Deshou Wang

MOST | National Natural Science Foundation of China (NSFC) (32172969)

• Jing Wei

MOST | National Natural Science Foundation of China (NSFC) (31972776)

• Jing Wei

MOST | National Natural Science Foundation of China (NSFC) (32102780)

• Wenjing Tao

MOST | National Natural Science Foundation of China (NSFC) (32473159)

• Xiangyan Dai

Additional files

Table S1

Graphical abstract