Strain‐Dependent Functional Variation of a Sex Specific Sox9 Enhancer

ABSTRACT

In mammalian sex differentiation, Sox9 expression in undifferentiated gonads determines male development. Sox9 is regulated by multiple enhancers, of which Enh13/mXYSRa is the most critical. Disruption of the GATA4 binding site in Enh13/mXYSRa causes XY sex reversal in the C57BL/6J (B6J) strain, but not in the mixed genetic background of B6J × DBA/2. To clarify the cause of these differences, mice with the same mutation were generated in DBA/2. Homozygous mutants developed as XY males. We hypothesized that the phenotypic differences between strains were due to functional differences of another enhancer. Mice lacking TESCO and carrying a mutation in the GATA4 binding site were generated and analyzed. The mutants developed according to their chromosomal sex despite reduced Sox9 expression. These findings imply that the DBA/2 genetic background contains factors preventing Sox9 expression from falling below the threshold required for male development. Bulk RNA‐seq analysis of fetal gonads during sex determination identified strain‐dependent differences in intermediate mesoderm development and protease inhibition pathways; the study also revealed that the expression of Sox8 has a redundant role in Sertoli cell differentiation to that of Sox9 in male embryonic gonads. These differences may contribute to the phenotypic variation observed between strains.

The same mutations on GATA4‐binding site in Sox9 distal enhancer (mXYSRa/Enh13) caused XY female in C57BL/6 and XY male in DBA/2. It is possible that another enhancer (TESCO) contributed to XY male in DBA/2; however, double mutants of the GATA4‐binding site and TESCO were also XY males. Transcriptome analysis implies that Sox8 and SERPINA family genes would be involved as potential causes.

1. Introduction

Although SRY expression in undifferentiated embryonic gonads determines the destination of sex to males in most mammals, the only established role of SRY as a transcription factor is the upregulation of Sox9 expression. SOX9 functions as a master regulator that controls the expression of a wide range of downstream genes necessary for male sex differentiation, and transgenic mice ectopically expressing Sox9 in XX gonads exhibit male phenotypes (Vidal et al. 2001). In addition, SOX9 regulates its coding genes in a positive feedback manner in the male embryonic testis (Sekido and Lovell‐Badge 2008). SOX9 belongs to the SOX family of transcription factors and is characterized by a conserved high‐mobility group (HMG) box DNA‐binding domain. Based on HMG box similarity, SOX proteins are classified into several groups; SOX9 is a member of Group E, which includes SOX8 and SOX10 (Bowles et al. 2000). Although Sox8 knockout mice do not show obvious defects in sex determination (O'Bryan et al. 2008), the combined loss of Sox8 and Sox9 results in more severe ovarian development than that observed in Sox9 conditional knockout mice, indicating partially redundant roles of these factors in testis development. Consistently, overexpression of Sox10 induced testis development in XX gonads, further supporting the functional overlap among Group E SOX proteins. Regulation of Sox9 expression during sexual differentiation is mediated by multiple enhancers, among which Enh13/mXYSRa plays a critical role. Deletion of this enhancer in mice results in male‐to‐female sex reversal (Gonen et al. 2018; Ogawa et al. 2018). Its nucleotide sequence is evolutionarily conserved among vertebrates. In humans, the corresponding sequence is located in a deletion region commonly found in several 46,XY disorder cases, in the upstream sequence of SOX9 (Croft et al. 2018; Kim et al. 2015).

Enh13/mXYSRa is a 557/711‐bp sequence containing binding sites (BSs) for SRY, NR5A1, GATA4, and SOX9. In NR5A1‐BS deletion mice with XY sex chromosome composition, Sox9 expression levels in embryonic testis during sex differentiation were reduced to 80% of wild‐type levels, and the phenotype was male. XY mice lacking SRY‐BS or SOX9‐BS were male, whereas mice with simultaneous deletions in both BSs showed complete female phenotypes (Ridnik et al. 2024). Recently, we identified another functional sequence, GATA4‐BS, located between SRY‐BS and SOX9‐BS on Enh13/mXYSRa, using the CRISPR/Cas9 system to induce a series of deletions in the enhancer in mice (Ogawa et al. 2025). Almost all XY mice carrying a two‐base‐pair substitution in GATA4‐BS exhibit a complete female phenotype in a C57BL/6J (B6J) genetic background. However, this was not the case for the mixed B6J × DBA/2 background, where considerable inter‐individual variabilities were observed. This observation likely reflects the widely accepted notion that B6 is prone to sex reversal (Bullejos and Koopman 2005; Correa et al. 2012; Eicher et al. 1996; Washburn et al. 2001). However, the underlying mechanism remains unclear.

In this study, we took advantage of the reduced susceptibility to sex reversal in the DBA/2 genetic background to investigate the function of GATA4‐BS in Enh13/mXYSRa and explored factors that may contribute to strain‐dependent differences in phenotype.

2. Results

2.1. Deficiency of Enh13/mXYSRa GATA4‐BS in the DBA/2 Genetic Background Resulted in XY Male; However, the Expression Level of Sox9 Was Reduced

Our previous study showed that a two‐base‐pair substitution of GATA4‐BS (GATA4‐BS^2sub) in Enh13/mXYSRa led to different phenotypes: XY complete sex reversal in the pure B6J strain and XY males in the mixed B6J‐DBA/2 genetic background (Ogawa et al. 2025). To investigate this difference, we generated a mouse strain carrying the same substitution in the GATA4‐BS of Enh13/mXYSRa on a DBA/2 genetic background (Figure 1A). By microinjection of the CRISPR/Cas9 tool into oocytes obtained from DBA/2, finally 16 individuals were obtained, of which three contained the GATA4‐BS^2sub allele. Homozygous, heterozygous, and wild‐type mice obtained by mating heterozygous mutants showed that all XX and XY individuals exhibited phenotypes of the reproductive system being female‐ and male‐type, respectively, and were fertile irrespective of genotype. Embryos were generated by mating heterozygous males with either heterozygous or homozygous females. At embryonic day 13.5 (E13.5), the early stage of sex differentiation, the morphologies of the testes in all 24 GATA4‐BS^2sub/2sub XY embryos were indistinguishable from those of wild‐type DBA/2 (Figure 1B, Table S1). They formed characteristic structures of the testis, such as the testis cord and angiogenesis, as with the wild type. At the adult stage, the phenotypes of the reproductive system were indistinguishable from those of wild‐type DBA/2 in all five GATA4‐BS^2sub/2sub mice examined (Figure 1C). Anogenital distance (AGD) divided by their body weights (BW), a quantitative indicator of the degree of male development, was not significantly different between the genotypes. Hematoxylin–eosin (HE) staining revealed that homozygous testes displayed normal histological architecture, including spermatogenesis (Figure 1Cc). As no differences were observed in the histology between the mutants and the wild type, we examined the expression levels of Sox9 in the embryonic gonads. The Sox9 expression level in whole embryonic gonads in GATA4‐BS^2sub/2sub tended to be reduced to approximately 60% of that in wild‐type DBA/2 (Figure 1Bb). Immunofluorescence analysis revealed that the testis cord structures were similar to those in wild‐type males, and SOX9 was clearly detected in the testis cords (Figure 1D). FOXL2, a marker of ovarian differentiation, was not detected in GATA4‐BS^2sub/2sub, XY gonads.

FIGURE 1.

Phenotypes of GATA4‐BS^2sub/2sub. (A) a An schematic diagram of the GATA4‐BS^2sub allele. b A result of genome editing of the GATA4‐BS^2sub/2sub. Skyblue highlighting indicates substituted bases. Enh13* indicates that it is Enh13/mXYSRa, in which two base substitutions have been introduced into GATA4‐BS. (B) Embryonic phenotypes at E13.5. a Pictures of embryonic gonads. Scale bar: 1 mm. b Bar plots of expression analysis of Sox9 in embryonic gonads. Statistical significance was assessed using an unpaired two‐tailed Student's t‐test. p < 0.05. (C) Adult phenotypes at 8 weeks old. a Bar plots of anogenital distance/body weight (AGD/BW). Statistical significance was assessed using an unpaired two‐tailed Student's t‐test. p < 0.05. b Pictures of adult internal reproductive organs. Scale bar: 5 mm. c HE staining of testis. Scale bar: 50 μm. (D) Immunofluorescence analysis of embryonic gonads at E13.5. Scale bar: 100 μm.

2.2. Whole Enh13/mXYSRa Deletion Caused Sex Reversal in DBA/2

To examine the effect of Enh13/mXYSRa on testis development in the DBA/2 background, mice with a deletion of the entire Enh13/mXYSRa locus were generated. The microinjection to fertilized oocytes obtained from DBA/2 was performed using a mixture of Cas9 protein, two gRNAs (sgRNA8 and mXYSRaDBA2T_gRNA), and a plasmid designed for another ongoing project. Finally, seven pups were obtained, one of which contained a 0.7 kb deletion of the enhancer (Enh13/mXYSRa^del) as a byproduct (Figure 2A). Wild‐type, heterozygous, and homozygous mice were generated by mating. While XX and XY heterozygous mice developed normally, with female‐ and male‐type internal and external genitalia, respectively, and were fertile, all five homozygous XY mutants analyzed exhibited completely female‐type internal and external genitalia during the adult stage (Figure 2B). Immunofluorescence analysis of embryonic XY gonads showed that FOXL2 was expressed in a pattern comparable to that of wild‐type females, whereas SOX9 was not expressed in all five XY individuals (Figure 2C).

FIGURE 2.

Phenotypes of Enh13/mXYSRa^del/del. (A) a A schematic diagram of the Enh13/mXYSRa^del allele. The dotted line indicates the area deleted. b A result of genome editing of the Enh13/mXYSRa^del/del. Skyblue highlighting indicates the junction of edited ends. (B) Images of adult internal reproductive organs showing adult phenotypes. a XX,Enh13/mXYSRa^del/wt; b XY,Enh13/mXYSRa^del/wt; c XY,Enh13/mXYSRa^del/del. Scale bar: 5 mm. (C) Immunofluorescence analysis of embryonic gonads at E13.5. Scale bar: 100 μm.

2.3. Double Mutant Mice of GATA4‐BS on Enh13/mXYSRa and TESCO Resulted in XY Male in DBA/2

As the mutation that caused sex reversal in B6J did not cause phenotypic changes in DBA/2, we investigated the possibility that the known enhancer TESCO is involved in Sox9 expression levels in DBA/2. To this end, two mutant alleles were generated in a DBA/2 background, one of which had GATA4‐BS^2sub and a deletion of TESCO in cis (GATA4‐BS^2sub; TESCO^del), and the other had only a TESCO deletion (TESCO^del) by CRISPR/Cas9 system using fertilized eggs of DBA/2 (Figure 3A). Transcription factor binding sites for SRY and NR5A1, which were reported to be required for TESCO activity (Sekido and Lovell‐Badge 2008), were encompassed by the deletions. At E13.5, the gross morphologies of the embryonic gonads of GATA4‐BS^2sub/2sub; TESCO^del/del, TESCO^del/del and wild‐type were indistinguishable in all individuals examined (Figure 3B, Table S1). Sox9 expression levels of embryonic gonads of TESCO^del/del and GATA4‐BS^2sub/2sub; TESCO^del/del were reduced to approximately 70% and 30% of the wild‐type levels, respectively (Figure 3C). However, at the adult stage, the sexual phenotypes of TESCO^del/del and GATA4‐BS^2sub/2sub; TESCO^del/del were consistent with their sex chromosome composition in all individuals examined. In both strains, the testis size and AGD/BW ratio were not significantly different from those in wild‐type mice, and no abnormality was identified by HE staining (Figure 3D,E). Immunofluorescence analysis of embryonic gonads showed differences between GATA4‐BS^2sub/2sub; TESCO^del/del and wild‐type mice (Figure 3F). While the signal intensities of SOX9 were comparable between GATA4‐BS^2sub/2sub; TESCO^del/del and wild‐type DBA/2, SOX9 signals were identified in the testis cords of wild‐type and were disorganized in GATA4‐BS^2sub/2sub; TESCO^del/del. In contrast, in TESCO^del/del, testis cords were well formed, and SOX9 signals were detected along the cords. FOXL2 expression was detected only in the GATA4‐BS^2sub/2sub and TESCO^del/del. FOXL2‐positive cells were predominantly localized to the cranial pole of the gonad, and adjacent to disorganized seminiferous cords; however, no cells expressed both SOX9 and FOXL2 simultaneously. To facilitate comparison of gonadal phenotypes among genotypes and genetic backgrounds, the phenotypic characteristics of all mutant mice analyzed in this study are summarized in Table S1.

FIGURE 3.

Phenotypes of TESCO^del/del and GATA4‐BS^2sub/2sub/TESCO^del/del. (A) a Schematic diagram of the TESCO^del and GATA4‐BS^2sub/TESCO^del. Enh13* indicates Enh13/mXYSRa with two base substitutions at GATA4‐BS. The dotted line indicates the area deleted. b Results of genome editing of the TESCO^del/del and GATA4‐BS^2sub/2sub/TESCO^del/del. Skyblue highlighting indicates substituted bases and the junction of edited ends. (B) Embryonic phenotypes at E13.5. Scale bar: 1 mm. (C) Bar plots of expression analysis of Sox9 in embryonic gonads. Statistical significance was assessed using an unpaired two‐tailed Student's t‐test. p < 0.05. (D) Adult phenotypes at 8 weeks old. a Pictures of internal reproductive organs. Scale bar: 5 mm. b HE staining of testis. Scale bar: 50 μm. (E) Bar plots of anogenital distance/body weight (AGD/BW). Statistical significance was assessed using an unpaired two‐tailed Student's t‐test. p < 0.05. (F) Immunofluorescence analysis of embryonic gonad at E13.5. Scale bar: 100 μm.

2.4. Hundreds of Genes Were Differentially Expressed in Undifferentiated Gonads Between B6J and DBA/2

To identify genes whose expression levels in undifferentiated gonads are different between B6J and DBA/2 at the stage of sex determination, bulk RNA‐sequencing (RNA‐seq) was performed using RNA obtained from gonads at E11.5 (20 tail somite stage). A total of 1402 and 1438 genes were identified as differentially expressed between the strains in the XX and XY gonads, respectively (Figure 4A). In addition, 283 and 423 genes showed sex‐biased expression in B6J and DBA/2, respectively. The expression levels of genes involved in sex differentiation were measured (Figure S1). Greater inter‐individual variation in the expression of many of these genes was observed in the XY B6J gonads than in the XY DBA/2 gonads. To identify genes that affect the maintenance of Sox9 expression and prevent male‐to‐female sex reversal in DBA/2, sexually differentially expressed genes (SDEGs) were compared between strains (Figure 4B). When XY‐biased genes were compared, 91 genes were identified as strain‐specific SDEGs in DBA/2, 131 in B6J, and 102 were shared between the two strains. Genes associated with male development, such as Sox9 and Amh, were among the SDEGs identified in both DBA/2 and B6J. The top 10 genes, including Sox8, ranked in ascending order by adjusted p‐values among the 91 DBA/2‐specific SDEGs are shown in Table 1. Among the XX‐biased genes, 203 SDEGs were identified in DBA/2, 21 in B6J, and 31 were shared between the two strains.

FIGURE 4.

Results of bulk RNA‐seq using E11.5 (20 tail somite stage) undifferentiated gonads. (A) Volcano plots. Blue and red dots indicate genes with an adjusted p‐value < 0.05 and an absolute log2 fold change greater than 1. (B) A Venn diagram showing the overlap of SDEG sets among each strain. a Inter‐strain comparison in XY. b Inter‐strain comparison in XX. (C) Results of GO analysis. The y‐axis indicates enriched GO terms, and the x‐axis represents the GeneRatio, defined as the proportion of input genes associated with each GO term. The size of each dot corresponds to the number of genes assigned to the term, and the color indicates the adjusted p‐value. GO terms with lower adjusted p‐values are shown in red, indicating stronger enrichment. a An inter‐strain comparison in XX mice. b An inter‐strain comparison in XY mice. c An inter‐sexual comparison in B6J. d An inter‐sexual comparison in DBA/2. (D) Results of GSEA. Genes were ranked according to their differential expression between the two groups. Each curve represents the running enrichment score for an individual gene set across the ranked gene list. Positive enrichment scores indicate gene sets enriched among genes upregulated in the first group, and negative enrichment scores indicate enrichment among genes upregulated in the second group. The peak of each curve corresponds to the maximum enrichment score, and the vertical bars indicate the positions of genes belonging to each gene set within the ranked list. a Comparison between strains, with XY,B6J as the first group and XY,DBA/2 as the second group. b Comparison between strains, with XX,B6J as the first group and XX,DBA/2 as the second group.

TABLE 1.

Top 10 DBA/2‐specific XY‐biased genes ranked by adjusted p‐value.

gene_id	gene_name	biotype	chr	log2FC	lfcSE	padj
ENSMUSG00000087259	2610035D17Rik	lncRNA	11	−1.619244266	0.225899	5.14E‐10
ENSMUSG00000021613	Hapln1	protein_coding	13	−1.14214426	0.162012	1.17E‐09
ENSMUSG00000030325	Klrb1c	protein_coding	6	−7.146319087	1.122574	8.20E‐08
ENSMUSG00000027444	Cstdc1	protein_coding	2	−3.553611585	0.573464	2.16E‐07
ENSMUSG00000021675	F2rl2	protein_coding	13	−1.85075618	0.317079	1.64E‐06
ENSMUSG00000036958	Cst11	protein_coding	2	−2.612095168	0.458662	3.38E‐06
ENSMUSG00000054423	Cadps	protein_coding	14	−1.571925681	0.276082	3.38E‐06
ENSMUSG00000032420	Nt5e	protein_coding	9	−2.685347101	0.476545	4.59E‐06
ENSMUSG00000036111	Lmo1	protein_coding	7	−1.090016619	0.193599	4.67E‐06
ENSMUSG00000024176	Sox8	protein_coding	17	−2.224347914	0.405303	9.37E‐06

Gene Ontology (GO) analysis of differentially expressed genes identified by strain comparisons within each sex revealed significant enrichment of pathways related to the regulation of protease activity and negative regulation of proteolysis. Examination of the genes contributing to these terms indicated that this enrichment was largely driven by elevated expression of several Serpina family genes. The GO term was also identified by gene set enrichment analysis (GSEA), together with other terms, including kidney development and renal system development.

3. Discussion

In mammalian sex determination, the induction of SOX9 expression is a key event that initiates testis differentiation, and its expression must be maintained robustly and stably. SOX9 is a dosage‐sensitive gene, and reduced SOX9 expression is associated with a spectrum of human disorders, such as campomelic dysplasia and disorders of sex development. In the context of sex determination, there is a quantitative threshold of Sox9 expression level for male development, and failure to reach this threshold during the critical sex‐determining window results in sex‐reversed individuals in the mouse (Gonen and Lovell‐Badge 2019). Here, we demonstrated that mice possess a robustness, conferred by multiple enhancers and redundant regulatory factors, which prevents Sox9 expression from dropping below the threshold required for normal sex determination.

In mice carrying a two‐base pair substitution in the GATA4‐BS of Enh13/mXYSRa, sex reversal occurred in the B6J genetic background but not in the DBA/2 background, indicating that the contribution of GATA4‐BS to Sox9 expression is strain dependent. The reduction of Sox9 expression in the embryonic testis of both genetic backgrounds indicates that GATA4‐BS is a functional sequence. However, Sox9 expression decreased to approximately 60% of wild‐type levels in DBA/2 mice, whereas 10% in B6J mice. Given that the threshold level of Sox9 required for testis differentiation is approximately 25%, as suggested by Gonen and Lovell‐Badge (2019), the Sox9 expression level in DBA/2 was above the threshold, thereby maintaining the male phenotype.

Gonen et al. (2017) reported that TESCO‐deleted mice with a B6J–CBA mixed genetic background were XY males, although Sox9 expression in the embryonic testis was reduced to approximately 60% of wild‐type levels. This observation is consistent with our results for TESCO^del/del; DBA/2 mice. The comparable reduction in Sox9 expression levels suggests that the function of TESCO is independent of genetic background.

In GATA4‐BS^2sub/2sub; TESCO^del/del on DBA/2, the Sox9 expression level in the embryonic testis was reduced to approximately 30% of wild‐type levels, accompanied by disorganization of testis cords and the presence of ectopic FOXL2‐positive cells. Nevertheless, these mice developed as males without undergoing sex reversal. While Sox9 expression was reduced to approximately 60% of the wild‐type levels in both GATA4‐BS^2sub/2sub and TESCO^del/del mice, the further reduction observed in GATA4‐BS^2sub/2sub; TESCO^del/del mice was likely attributable to the combined loss of function of both enhancers. However, whether these enhancers function in an additive or synergistic manner remains unclear. These enhancers might be physically interacted during Sox9 expression (Ridnik et al. 2021), potentially through the binding of GATA4 to Enh13/mXYSRa. Furthermore, the phenotype of double‐mutant mice suggests that, in addition to GATA4‐BS in Enh13/mXYSRa and TESCO, additional regulatory elements contribute to Sox9 expression during testis differentiation, such as Enh14, which exhibits testis‐specific activity (Gonen et al. 2018) and is involved in the sex determination of Tokudaia osimensis (Terao et al. 2022). These elements may contribute more strongly to Sox9 expression in the DBA/2 genetic background than in the B6J background.

XY mice lacking the entire Enh13/mXYSRa region developed as females, indicating that, even in the DBA/2 genetic background, this enhancer is required for Sox9 expression and testis development. Since the enhancer sequences were identical between B6J and DBA/2 mice, NR5A1‐BS, SRY‐BS, and SOX9‐BS were likely functional in the DBA/2 background, as reported in the B6‐mixed background (Ogawa et al. 2023; Ridnik et al. 2024). SRY and SOX9 are key regulators of sex determination. The phenotypic difference between GATA4‐BS^2sub/2sub in DBA/2 and Enh13/mXYSRa^del/del in DBA/2 suggests that GATA4 makes a smaller contribution to enhancer activity than other factors. However, because the loss of NR5A1‐BS, SRY‐BS, or SOX9‐BS alone was insufficient to reduce Sox9 expression to a level that induced sex reversal, GATA4 likely functions cooperatively with other factors. Taken together, our results indicate that although the contribution of GATA4‐BS is relatively modest in the DBA/2 background, the essential role of this enhancer itself is conserved across genetic backgrounds.

Phenotypic analyzes of mutant mice clearly demonstrated that the function of GATA4‐BS in Enh13/mXYSRa differed between the genetic backgrounds. To identify the potential modifying factors underlying this difference, we performed bulk RNA‐seq using B6J and DBA/2 fetal gonads collected shortly after the onset of sex determination, with the developmental stage precisely matched with the tail somite stage. GO analysis of the differentially expressed genes identified by strain comparisons in XY mice revealed enrichment of pathways related to the regulation of proteolysis. However, because similar GO terms often arise from overlapping sets of genes, this enrichment is likely driven by the elevated expression of Serpina family genes in DBA/2. Since multiple Serpina paralogs form a genomic cluster (Forsyth et al. 2003), strain‐dependent variations in shared regulatory elements may coordinately affect the expression of these genes. Indeed, similar GO terms were also identified in the strain comparison of XX‐biased genes, indicating that the expression of this gene family differed between DBA/2 and B6J independent of sex. GSEA comparing XY gonads between strains revealed that GO terms associated with kidney development and negative regulation of proteolysis were significantly enriched among the genes upregulated in DBA/2. Because the kidney and gonad originate from the intermediate mesoderm, renal‐related GO terms are frequently detected in the analyzes of gonadal transcriptomes and likely reflect shared developmental programs. GO terms related to renal development were also detected in the GSEA of XX mice, although strain differences were smaller than those observed in XY mice. Taken together, genes associated with the inhibition of proteolysis, including Serpina family genes, as well as genes involved in gonadal development, were expressed at higher levels in DBA/2 than in B6J. However, because these differences were observed in both sexes, they likely represent strain‐dependent variations independent of sex determination, making it difficult to identify candidate modifier genes using these approaches.

We hypothesized that the modifier genes contributing to the phenotypic differences are specifically expressed in either strain of XY mice. To test this, we extracted genes that showed sex‐biased expression within each strain and were significantly upregulated in XY samples, and compared them between DBA/2 and B6J. As a result, 91 genes were identified as being specifically upregulated in XY DBA/2 mice. When these genes were ranked by adjusted p‐value, Sox8 was listed as the tenth gene (Table 1). SOX8 is a transcription factor that functions redundantly with SOX9 (Barrionuevo et al. 2006). A comparison of Sox8 expression levels between the XX and XY samples within each strain further showed that, except for one sample, the sex difference in B6J tended to be smaller than that observed in DBA/2. These results raise the possibility that in the DBA/2 genetic background, SOX8, a member of the SOX family Group E, together with SOX9, may contribute to the maintenance of male developmental programs in Enh13/mXYSRa mutants. To determine whether SOX8 is responsible for the strain‐dependent phenotypic differences, future studies analyzing mice carrying simultaneous mutations in Enh13/mXYSRa and Sox8 in the DBA/2 genetic background are required.

In mammalian sex determination, the testis and ovary developmental pathways mutually antagonize each other (Garcia‐Moreno et al. 2019). In the B6J genetic background, Sry expression is prone to delay or decrease, which may compromise the activation of the testis pathway and lead to insufficient repression of the ovary pathway, thereby increasing the likelihood of sex reversal (Bullejos and Koopman 2005; Correa et al. 2012; Eicher et al. 1996; Washburn et al. 2001). In addition, our RNA‐seq analysis produced results consistent with this framework. Despite restricting sample collection to the 20 tail somite stage, gene expression levels in XY gonads from the B6J strain showed substantially greater inter‐individual variability than those from DBA/2 mice (Figure S1). In particular, in the XY B6J gonads, one individual exhibited markedly higher expression of numerous testis‐promoting genes, including Sry, Sox9, and Dmrt1, compared with the other two biological replicates. A similar pattern was observed for Wt1, Nr5a1, and Gata4, which are all upstream regulators of Sry expression (Bradford et al. 2009; Miyamoto et al. 2008; Sekido and Lovell‐Badge 2008; Tevosian et al. 2002). In contrast, expression differences between this individual and the remaining replicates were relatively small for ovary‐promoting genes, such as Foxl2, Rspo1, and Wnt4. These observations suggest that in the B6J genetic background, the transcriptional environment that supports Sry activation in the bipotential gonad is intrinsically variable among individuals, thereby increasing the likelihood that fluctuations in the balance between the testis‐ and ovary‐promoting pathways result in XY sex reversal. Furthermore, because genes involved in ovarian differentiation, whose expression does not directly depend on SRY, exhibited comparatively limited inter‐individual variation, this transcriptional instability does not appear to reflect a general increase in expression noise but rather a phenomenon largely restricted to the testis‐determining pathway. In contrast, in the DBA/2 genetic background, it is likely that there are mechanisms that maintain the expression of testis‐promoting genes above the threshold required for robust male sex determination. One potential contributor to this buffering capacity is SOX8, which was expressed at higher levels in DBA/2 gonads. In addition, gene sets related to the inhibition of proteolysis were significantly enriched in DBA/2; further studies are required to clarify the relationship between the function of SERPINE and this mechanism.

4. Experimental Procedures

4.1. Animal Experiments

All animal experiments were approved by the Animal Protection and Use Committee of the National Center for Child Health and Development (Tokyo, Japan). This study complied with all relevant ethical regulations for animal use. DBA/2Cr and C57BL/6J mice were obtained from Sankyo Laboratory Services (Tokyo, Japan), and Jcl:ICR mice were purchased from CLEA Japan (Tokyo, Japan). The animals were kept under a 12 h light/dark cycle, and food and water were provided ad libitum. Euthanasia was performed by cervical dislocation or isoflurane overinhalation. E0.5 was defined as noon of the day of embryonic transfer in the case of artificial fertilization, or as the day a vaginal plug was observed in the case of natural mating.

4.2. Preparation of sgRNA

Locations of sgRNAs were based on previous reports: Gonen et al. (2017) for deletion of TESCO and Ogawa et al. (2025) for substitution of GATA4‐BS. For generation of deletion of Enh13/mXYSRa, sgRNA8 (Ogawa et al. 2018) and mXYSRaDBA2T sgRNAs were used. sgRNAs were synthesized using a CUGA7 gRNA Synthesis Kit (Nippon Genetech, Tokyo, Japan) according to the manufacturer's instructions. For PCR amplification of the double‐stranded DNA templates, primers TESCO_KO_gRNA1temp and TESCO_KO_gRNA2temp were used for TESCO deletion, mXYSRa_GATA_gRNA for GATA4‐BS substitution in Enh13/mXYSRa, and gRNA8temp and mXYSRaDBA2T_gRNAtemp for deletion of the Enh13/mXYSRa region (Table 2).

TABLE 2.

Primers used in this study.

Primer name	Sequence (5′ to 3′)
TESCO_KO_gRNA1temp	CTAATACGACTCACTATAGGTTGGAGTTCCGATTTAGACGTTTTAGAGCTAGAAATAGCA
TESCO_KO_gRNA2temp	CTAATACGACTCACTATAGTCTGTTAACGTATGACACGAGTTTTAGAGCTAGAAATAGCA
mXYSRa_GATA_gRNA	CTAATACGACTCACTATAGTGAGACAATCATATCTTCCAGTTTTAGAGCTAGAAATAGCA
gRNA8temp	CTAATACGACTCACTATAGAGACTTGAAATTATTAAGAGTTTTAGAGCTAGAAATAGCA
mXYSRaDBA2T_gRNAtemp	CTAATACGACTCACTATAGCATCTTTATGTTTCAAGATTTGGGTTTTAGAGCTAGAAATAGCA
DBA2_mXYSRa_geno_F1	CCCACATAAGAGATGAAAGC
DBA2_mXYSRa_geno_R1	TGCTAGAATAAAGCTAAGCC
TESCO_KO_geno_F1	TAAGATACTATCCCAAGAGC
TESCO_KO_geno_R2	GCTACCAAGAGAGCTTAAGG
Sox9_qPCR_MY_F2	GACCAGTACCCGCATCTGCAC
Sox9_qPCR_IT_R1	CACGAAGGGTCTCTTCTCGC
beta‐actinF	CGTGAAAAGATGACCCAG
beta‐actinR	TGGTACGACCAGAGGCATACAG

4.3. In Vitro Fertilization, Microinjection

For superovulation, 0.2 mL of CARD HyperOva (Kyudo, Saga, Japan) was administered intraperitoneally to four‐week‐old females or 0.2 mL of anti‐inhibin serum (AIS) (Central Research, Tokyo, Japan) to over eight‐week‐old females 72 h before in vitro fertilization (IVF), and 5 IU of human chorionic gonadotropin (hCG) (Sigma Aldrich CG10) 24 h before IVF. Gametes were collected from mice and fertilized in vitro in CARD mHTF medium (Kyudo, Saga, Japan) containing 0.4 mg/mL of l‐glutathione reduced (Sigma Aldrich G4251). Wild‐type female and male DBA/2 mice were used for IVF to generate Enh13/mXYSRa^del, GATA4‐BS^2sub and TESCO^del alleles. After washing the fertilized eggs with M2 medium (Sigma Aldrich M7167), 250 ng/μL of sgRNAs and 100 ng/μL of Cas9 (Nippon Genetech, Tokyo, Japan) were mixed and injected to cytoplasm of the fertilized eggs to generate the TESCO^del/del allele. To generate GATA4‐BS^2sub and delete Enh13/mXYSRa, single‐stranded oligodeoxynucleotides and a plasmid, which was for another ongoing project, respectively, were added to the gRNAs/Cas9 protein mixture and microinjected into the nucleus and cytoplasm of the fertilized eggs. The eggs were cultured in KSOM medium (Arc Resources, Kumamoto, Japan) at 37°C in 5% CO₂ overnight, and those that developed to the two‐cell stage were transferred into the ampulla of the uterine tube of pseudopregnant ICR female mice.

4.4. Sexing and Genotyping

For PCR‐based sexing and genotyping, genomic DNA was extracted from newborn fingertip samples (within 1 week postpartum), adult ear punches, and embryonic tails using a lysis buffer containing 0.1 g/L proteinase K, 10 mM Tris–HCl (pH 7.5), 1 mM EDTA (pH 8.0), 0.1 M NaCl, and SDS (1%), followed by phenol‐chloroform extraction.

The PCR was performed using BIOLINE Taq (BioLine, London). Sex was determined using the primer sets Zfy_F/Zfy_R and Xist_F/Xist_R (Obata et al. 1998). For genotyping, the GATA4‐BS^sub allele was amplified using mXYSRa_geno_F1 and mXYSRa_geno_R1, followed by digestion with NlaIII (New England Biolabs, R0125L). The TESCO deletion allele was analyzed using TESCO_KO_geno_F1 and TESCO_KO_geno_R2 with the same enzyme. Deletion of Enh13/mXYSRa was confirmed by PCR using mhXYSRa_geno_F1 and mhXYSRa_geno_R2. The primers used for genotyping are listed in Table 2.

For the genotyping of F₀ and F₁ mice, PCR amplicons were treated with ExoSAP‐IT PCR Product Cleanup Reagent (Thermo Fisher Scientific) and subjected to Sanger sequencing using the corresponding PCR primers and BigDye Terminator v3.1 (Applied Biosystems).

4.5. Phenotypic Analysis

Eight‐week‐old mice were used for phenotypic analysis of adulthood. The mice were euthanized by inhalation anesthesia using isoflurane. The AGD was determined as the mean of two measurements.

4.6. Histological Analysis

Gonads removed from E13.5 embryos or adult mice were fixed with 4% paraformaldehyde in phosphate‐buffered saline (PBS) at 4°C overnight and dehydrated with a series of concentrations of methanol diluted in PBS. Then, they were embedded in paraffin, sectioned at 7 μm thickness, and covered with slide glasses coated with silane. For HE staining, the slides were deparaffinized by immersing them in xylene and a series of ethanol solutions with concentrations ranging from 100% to 70%. Next, the slides were immersed in Mayer's hematoxylin solution (Wako, 131‐09665) diluted four times with water, water, 1% Eosin Y Solution (Wako, 051‐06515) diluted four times with EtOH, a series of ethanol concentrations from 70% to 100%, and xylene, in that order. Finally, the tissue sections were mounted on MountQuick (Daido Sangyo, Saitama, Japan) and observed using a microscope, BX53 (EVIDENT). For immunofluorescence, the slides were deparaffinized by immersing them in xylene and a series of ethanol concentrations ranging from 100% to 70%. Antigen retrieval was performed by autoclaving slides soaked in 10 mM citric acid monohydrate solution for 20 min at 121°C. The slides were incubated in 10% fetal bovine serum (FBS) in PBS for 1 h at room temperature for blocking. Primary antibodies, anti‐FOXL2 (1:200, ab5096; Abcam, Cambridge, UK) and anti‐SOX9 (1:1000, ab5535; Merck Millipore, Burlington, Massachusetts, USA), were diluted in 1% FBS in PBS, applied to the slides, and incubated overnight at 4°C. After washing with PBS containing 0.1% Triton‐X100, Donkey anti‐rabbit IgG(H + L) conjugated with AlexaFluor488 (1:500, A21206; Life Technologies, Carlsbad, California, USA) and donkey anti‐goat IgG (H + L) conjugated with Cy3 (1:2000, Ab6949; Abcam) diluted in PBS were applied to the slides and incubated for an hour at room temperature. 4,6‐diamidino‐2‐phenylindole (DAPI) solution (1:1000,19178‐91; Nacalai Tesque Inc., Kyoto, Japan) was used for counterstaining. Images were obtained using a confocal laser microscope, FV3000 (EVIDENT). Fluorescence images were processed using custom Python scripts. Image contrast was adjusted by linearly rescaling pixel intensities with a fixed display range (minimum = 0, maximum = 130). Images from the green and red channels were subsequently merged by combining the corresponding RGB images. Image processing was performed in Python 3 using the NumPy (v2.3.1) and tifffile (v2025.12.12) packages (Figure S2).

4.7. Expression Analysis

Total RNA was extracted from E13.5 gonads using ISOGEN (Nippon Genetech) and subsequently used for reverse transcription reaction with SuperScript II (Thermo Fisher Scientific) and an oligo(dT) primer. Real‐time quantitative PCR (RT‐qPCR) was performed using SYBR Green Master Mix (Thermo Fisher Scientific) and primer sets Sox9_qPCR_MY_F2 and Sox9_qPCR_IT_R1 for Sox9 and beta‐actinF/beta‐actinR for Actb as an internal control (Table 2). Gene expression levels were quantified by delta–delta ∆∆Ct method. All calculations were performed using R‐4.3.3 program.

4.8. Statistical Analysis

Comparisons between the wild‐type and mutant groups were performed using two‐sided Welch's t‐test. Data are presented as mean ± SEM from three independent biological replicates. Individual sample sizes (n) are shown in the figures. Statistical significance was defined as p‐value < 0.05.

4.9. Bulk RNA‐Seq

Total RNA was extracted from E11.5 gonads (20 tail somite stage) from XX and XY embryos of the B6J and DBA/2 strains (n = 3 per group; 12 samples in total). Sequencing libraries were constructed using the MGIEasy Fast RNA Library Prep Set and the DNBSEQ‐T7RS High‐Throughput Sequencing Reagent Kit V3.0 (MGI, Shenzhen, China). All the RNA extraction and sequencing procedures were performed at Genesis Healthcare (Tokyo, Japan). FASTQ files were processed using Trimmomatic‐0.39 with the following parameters: LEADING:28, TRAILING:28, SLIDINGWINDOW:4:20, and MINLEN:30. Trimmed reads were aligned to the mouse reference genome (GRCm39) using STAR‐2.7.11. Read counts mapped to each ENSEMBL gene ID were quantified using featureCounts in subread (v2.0.8) based on the GRCm39 GTF annotation file. Statistical analyzes to identify differentially expressed genes (DEGs) were performed using the DESeq2 R package (v1.42.0), and plots were generated using the ggplot2 R package (v3.5.2). Enrichment analyzes were performed using the org.Mm.eg.db R package (v3.18.0) and the clusterProfiler R package (v4.10.0).

Author Contributions

Suzune Furuhashi: investigation. Iku Tsuchiya: data curation, formal analysis, investigation, validation, visualization, writing – original draft, writing – review and editing. Aki Oishi: investigation. Atsumi Tsuji‐Hosokawa: data curation, investigation, project administration, supervision, validation, writing – original draft, writing – review and editing. Shuji Takada: conceptualization, data curation, funding acquisition, investigation, methodology, project administration, supervision, writing – original draft, writing – review and editing. Miho Terao: investigation. Karin Sugiyama: investigation. Yuya Ogawa: investigation, writing – review and editing.

Funding

This study was supported in part by grants from the National Center for Child Health and Development (grant numbers 29‐11 and 2023B‐1) and by a JSPS KAKENHI Grant (Grant Number 24K11036) to S.T.

Ethics Statement

All animal experiments were approved by the Animal Protection and Use Committee of the National Center for Child Health and Development (A2020‐007‐C05‐M04 issued on 6th November 2020, Y2026‐041 issued on 2nd April 2026). This study complied with all relevant ethical regulations for animal use.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Expression levels of genes related to sex differentiation in E11.5 (20 tail somite stage) undifferentiated gonads. Bar plots represent mean TPM values in each sex and strain.

Figure S2: Python script. A. enhance.v2.py B. merge.py.

Figure S3: gtc70131‐sup‐0002‐FigureS3.xlsx.

Table S1: Summary of genotypes and phenotypes.

Tsuchiya, I. , Tsuji‐Hosokawa A., Furuhashi S., et al. 2026. “Strain‐Dependent Functional Variation of a Sex Specific Sox9 Enhancer.” Genes to Cells 31, no. 4: e70131. 10.1111/gtc.70131.

Data Availability Statement

The data that supports the findings of this study are available in the Supporting Information of this article.