|Home | About | Journals | Submit | Contact Us | Français|
Nuclear receptor subfamily 0, group B, member 1 (Nr0b1; hereafter referred to as Dax1) is an orphan nuclear receptor that regulates adrenal and gonadal development. Dosage-sensitive sex reversal, adrenal hypoplasia congenita, critical region on the X chromosome, gene 1 (Dax1) mutations in the mouse are sensitive to genetic background. In this report, a spectrum of impaired gonadal differentiation was observed as a result of crossing the Dax1 knockout on the 129SvIm/J strain onto the C57BL/6J strain over two generations of breeding. Dax1-mutant XY mice of a mixed genetic background (129;B6Dax1−/Y [101 total]) developed gonads that were predominantly testislike (n = 61), ovarianlike (n = 27), or as intersex (n = 13). During embryonic development, Sox9 expression in the gonads of 129;B6Dax1−/Y mutants was distributed across a wide quantitative range, and a threshold level of Sox9 (>0.4-fold of wild-type) was associated with testis development. Germ cell fate also varied widely, with meiotic germ cells being more prevalent in the ovarianlike regions of embryonic gonads, but also observed within testicular tissue. Ptgds, a gene associated with Sox9 expression and Sertoli cell development, was markedly downregulated in Dax1−/Y mice. Stra8, a gene associated with germ cell meiosis, was upregulated in Dax1−/Y mice. In both cases, the changes in gene expression also occurred in pure 129 mice but were amplified in the B6 genetic background. Sertoli cell apoptosis was prevalent in 129;B6Dax1−/Y gonads. In summary, Dax1 deficiency on a partial B6 genetic background results in further modulation of gene expression changes that affect both Sertoli cell and germ cell fate, leading to a phenotypic spectrum of gonadal differentiation.
The important cell fate decisions that direct early gonadal development center on the somatic cell population and the germ cell lineage. Pre-Sertoli cell commitment is critical for testis development, whereas meiotic entry signifies germ cell fate determination in the ovary. The bipotential gonad remains undifferentiated in the embryo until about midgestation in the mouse, after which point a genetic cascade establishes distinctly male or female sex characteristics by dictating somatic and germ line cell fates. Two male-determining genes, Sry and Sox9, have been identified for their molecular role in Sertoli cell determination and testis differentiation [1, 2]. Transgenic expression of either Sry or Sox9 is sufficient to induce testis development in XX mice [3, 4]. Genetic studies in mice have also sought to elaborate the cellular and morphogenetic outcomes imposed by the network of genes involved in the mammalian sex determination cascade  and genetic disorders of reproduction .
Nuclear receptor subfamily 0, group B, member 1 (Nr0b1; hereafter referred to as Dax1) is an X-linked gene that acts at several levels of the reproductive axis. The 470-amino acid protein encoded by Dosage-sensitive sex reversal, adrenal hypoplasia congenita, critical region on the X chromosome, gene 1 (Dax1) belongs to the nuclear receptor superfamily . Dax1 is expressed in several endocrine organs, including the ventromedial hypothalamus, anterior pituitary, adrenal cortex, testis, and ovary . In a mouse Dax1 knockout model, male mice on the 129SvIm/J background (129Dax1−/Y) were found to have testis abnormalities characterized by dilated seminiferous tubules and failed spermatogenesis . The dilated tubules in the adult reflected impaired testis cord morphogenesis during embryonic development , resulting in the eventual obstruction of tubules and loss of germ cells . Although the expression of Sertoli cell markers, such as Sox9 and Dhh, was not significantly altered in the embryonic gonad of mice with a 129SvIm/J background, peritubular myoid cell number was reduced, possibly accounting for abnormal testis cord formation .
The role of Dax1 in the male sex determination pathway was highlighted using the Mus domesticus poschiavinus mouse, an outbred strain with a “weakened” Sry allele (reduced and delayed mRNA expression of Sry) [12–14]. In the presence of the SryPos allele, all genetically male Dax1 mutant mice were sex reversed, even though the expression of Sry was not altered . However, Sox9 was not expressed, indicating that the molecular action of Dax1 occurs downstream of Sry but before expression of Sox9. A similar phenomenon of sex reversal was observed when breeding the Dax1 knockout allele from the originally reported 129SvIm/J background onto the C57BL/6J background (B6) . On a congenic B6 background (obtained by continuous backcrosses), Dax1 knockout mice fail to upregulate Sox9, even in the presence of full Sry expression from a Mus musculus Sry allele. In summary, Dax1 loss of function in the mouse yields an XY gonadal phenotype that ranges from disordered testis cords in M. musculus 129SvIm/J to complete sex reversal in M. domesticus poschiavinus or congenic M. musculus C57BL/6J.
By taking advantage of the sensitizing effect of the B6 strain, the aim of the current study was to evaluate the intermediary phenotype of Dax1 deficiency on gonadal and sexual development. The goal was to understand which aspects of testis development are most sensitive to Dax1 deficiency and are susceptible to the B6 genetic background.
All procedures described within this article were reviewed and approved by the Northwestern University Institutional Animal Care and Use Committee, and they were performed in accordance with the Guiding Principles for the Care and Use of Laboratory Animals. Animals were killed by CO2 inhalation at 12–16 wk of age. Genotyping for Dax1 and Sry were performed on all progeny at the time of weaning. External genitalia were assessed by measurement of the anogenital distance. Dissection of internal genitalia was performed, and tissue was fixed in 10% neutral buffered formalin for routine histology. Processed tissue was embedded in paraffin, and 4-μm sections were stained with hematoxylin (Surgipath, Richmond, IL) and eosin (Fisher Scientific, Fair Lawn, NJ).
Total RNA from individually dissected embryonic gonad-mesonephros complex pairs was extracted using Trizol (Invitrogen, Carlsbad, CA). Complementary DNA was synthesized using 200 ng RNA template combined with MMLV reverse transcriptase (Promega, Madison, WI), random primers (Roche Diagnostic, Nutley, NJ), and dinucleotide triphosphates (Promega) in a 20-μl reaction. Real-time RT-PCR product was measured by SYBR Green fluorescence emission using the Bio-Rad iCycler real-time system (Bio-Rad, Hercules, CA). A two-step cycling program (Tm of 55°C or 60°C) used 2 μl of the RT reaction containing cDNA. The threshold cycle number (Ct) was determined for each sample. Samples were tested in duplicate; at least six samples were tested for each genotype group. For each gene of interest, the averaged Ct value in both the reference group (ref) and the experimental group (exp) were normalized to the corresponding Ct value of the internal control gene, RPL19 (ctl). The reference group refers to wild-type control. The relative expression was calculated based on the following equation: 2(ref-ctl)/2(exp-ctl). Values are reported as the average ± SD. Statistical significance was determined by performing an unpaired Student t-test analysis (P < 0.05). Primers used were (5′ to 3′): Sox9, forward: AAGAAAGACCACCCCGATTACA, reverse: CAGCGCCTTGAAGATAGCATT; Ptgds, forward: GCTCTTCGCATGCTGTGGAT, reverse: GCCCCAGGAACTTGTCTTGTT; Stra8, forward: TCGATCTCTCCCACTCCT, reverse: CAGAGACAATAGGAAGTGTC; and the internal control RPL19 forward: CTGAAGGTCAAGGGAAATGTG, reverse: GGACAGAGTCTTGATGATCTC.
A PCR-amplified cDNA fragment of the prostaglandin D2 synthase gene (NM_008963; nucleotides 343–723) was inserted into the pCMV-Sport6 plasmid vector (Invitrogen, Carlsbad, CA). Plasmids were linearized, and digoxygenin (DIG)-labeled riboprobes were generated by T7 RNA polymerase using a DIG RNA labeling mix (Roche, Nutley, NJ). Embryonic gonads were dissected at 13.5 days postcoitum (dpc). Following fixation and tissue processing, hybridization with DIG-labeled riboprobe occurred overnight. Bound probe was detected with an anti-DIG antibody labeled with alkaline phosphatase, followed by an NBT-BCIP enzyme substrate reaction (Roche), which yielded a purple color product where mRNA transcripts were present. A minimum of four gonads were tested for each genotype group.
Paraffin-embedded embryos were cut to 4-μm thickness to obtain tissue sections for immunostaining. Deparaffinized slides were subjected to antigen retrieval by high-temperature sodium citrate buffer prior to blocking and primary antibody incubation. GammaH2AX antibody (mouse host) was available through Upstate Cell Signaling (Charlottesville, VA), and the M.O.M. kit (mouse-on-mouse) from Vector Laboratories (Burlingame, CA) was used for detection. Goat anti-AMH antibody was purchased from Santa Cruz Biotechnologies (Santa Clara, CA). Alexa-conjugated secondary antibodies were obtained from Invitrogen. Confocal image microscopy was enabled by a Zeiss UV Laser Scanning Microscope 510 Meta (Thornwood, NY). Imaging was performed with Zeiss LSM software version 3.2.
Gonadal sections were processed for histochemical analysis as described above. The TUNEL assay kit was a product of Calbiochem (EMD Biosciences, San Diego, CA). Images were captured by confocal microscopy. A minimum of n = 9 embryos were evaluated in the 129;B6 Dax1 mutant genotype group; n = 4 embryos were evaluated for each wild-type control group. The number of apoptotic cells was counted on individual tissue sections (at least four sections per individual embryo), then averaged for each genotype group.
Breeding of 129SvIm/J (129) Dax1 knockout mice onto the C57BL/6J (B6) genetic background resulted in male-to-female sex reversal in as few as two generations of breeding (designated 129;B6). Dax1 mutant animals (129;B6Dax1−/Y) were grouped by external appearance as male or female without ambiguity (by the measurement of anogenital distance in mice). A total of 101 129;B6Dax1−/Y mutants were characterized out of 489 animals of the same breeding generation. The majority of 129;B6Dax1−/Y animals were male in appearance (70 of 101; Table 1). By contrast, 31 out of 101 129;B6Dax1−/Y animals were externally female. Examination of internal reproductive tissues in 101 mutant animals revealed a diverse spectrum of defects related to sex differentiation which are summarized in Table 1. Representative histology of the spectrum of phenotypes observed in 129;B6Dax1−/Y mutants is shown in Figure 1. As a reference, gross anatomy and histology of wild-types of each sex are presented (Fig. 1, A–D). Dissection of Dax1-mutants at 12–16 wk of age revealed that the majority of external males were hypogonadal [61/70; Table 1] (testis weight approximately one third of wild-type) with normal testicular descent (Fig. 1, E and F). Epididymis, vas deferens and other secondary male structures developed normally. External females were also identified in the 129;B6Dax1−/Y genotype group, presenting a range of phenotypes from complete male-to-female sex reversal [11/31] (1G, H) to gonadal dysgenesis [16/31] (1I, J, K). Intersex mice (containing both male and female internal reproductive features- 1L, M, N) were also observed among 129;B6Dax1−/Y animals. Intersex mice could be either male or female in external appearance (9 male; 4 female; Table 1). Sex-reversed mice were infertile.
The molecular basis of sex reversal in 129;B6Dax1−/Y mice was examined by quantitative analysis of the male-determining gene Sox9. In previous studies, Sox9 expression was detectable by in situ hybridization in 129Dax1−/Y gonads that developed (100%) as males, and was not significantly different from wild-type male gonads . In contrast, Sox9 expression was abrogated on a congenic B6 background, which conferred 100% sex reversal . To determine whether Sox9 levels correlate with gonadal sex in the Dax1 mutant mice on a mixed background, Sox9 expression was measured from mRNA taken from individual 129;B6Dax1−/Y mutant gonads isolated at 13.5 dpc (Fig. 2). Gonadal sex was designated at the time of dissection by the presence of embryonic testis cords and by gonad size, two distinguishing features of the male gonad . In Figure 2, Sox9 levels in wild-type male and female controls, respectively, are indicated by separately marked horizontal lines; wild-type males served as the reference group for Sox9 quantitation. Sox9 expression from a total of 31 gonads that were 129;B6Dax1−/Y were plotted individually. Overall, male development in 20 embryos (Fig. 2, individual dashes) was characterized by Sox9 levels between approximately 0.4- and 1.0-fold of that in wild-type males (18 of 20 embryonic males). The Sox9 levels in the majority of embryos with a female morphology (Fig. 2, triangles) were below 0.4-fold of wild-type males (8 of 11 embryonic females). In a few cases (3 of 11), female development occurred at levels of Sox9 above 0.4-fold. These results suggest that Dax1 deficiency on a partial B6 background results in a quantitative modulation of Sox9 expression. Thus, a threshold level of Sox9 mRNA may be required for testis differentiation.
The fate of embryonic germ cells in 129;B6Dax1−/Y animals was also addressed. During embryonic development, female germ cells undergo meiosis; in males, meiosis takes place after birth. Histone H2AFX (H2A Histone Family, member X, or H2AX) is rapidly phosphorylated at the sites of DNA double strand break that occur during meiotic crossover. The modified histone protein (gammaH2AX) is widely detected (Fig. 3) in the nuclei of germ cells of wild-type female gonad tissue at 14.5 dpc (Fig. 3A), and it is absent from wild-type male gonad tissue (Fig. 3F). Using an antibody directed against gammaH2AX, meiotic germ cells were detected in mutant 129;B6Dax1−/Y gonads examined at 14.5 dpc (n = 12). The pattern of distribution of gammaH2AX-positive cells throughout the gonadal field varied considerably, and a summary of representative images of 129;B6Dax1−/Y gonads is shown (Fig. 3, B–E). Meiotic germ cells were present throughout the length of the gonad in some mutants, as seen in Figure 3, B and C, although more sparsely than in wild-type females. Some 129;B6Dax1−/Y gonads exhibiting meiotic germ cells developed with a male morphology (Fig. 3, D and E, compared with wild-type male). It should be noted that meiotic germ cells were never detected in 129Dax1−/Y gonads that are not susceptible to sex reversal (data not shown). In summary, meiotic entry in 129;B6Dax1−/Y gonads is nonuniform, suggesting local interactions between somatic and germ cell microenvironments.
To examine the molecular consequences associated with the modulation of Sox9 expression and the ectopic entry of germ cells into meiosis, the expression of prostaglandin D2 synthase (Ptgds) was compared among Dax1 mutants of different genetic backgrounds. Ptgds is the rate-limiting enzyme in prostaglandin D2 production, and it is a member of the lipocalin family of retinoid transport proteins. Its expression is sexually dimorphic, and in the testis it is linked to Sox9 induction . Whole-mount in situ hybridization was performed on gonads at 13.5 dpc (Fig. 4). Ptgds expression was detected in wild-type male gonads (Fig. 4A) along embryonic testis cords, a pattern consistent with published data . Expression was weaker and less distinctive in 129Dax1−/Y gonads (Fig. 4B). Furthermore, in 129;B6Dax1−/Y gonads (Fig. 4C), as in the female (Fig. 4D), there was no detectable expression by in situ hybridization (n = 4 per group). Additionally, Ptgds expression was quantitated by real-time RT-PCR in gonads dissected from 129 Dax1−/Y (not susceptible to sex reversal) and 129;B6 Dax1−/Y (susceptible to sex reversal) animals (Fig. 4, graph). Wild-type male expression levels served as the reference control (n = 6; Fig. 4, black bar in graph). In agreement with the results of in situ hybridization experiments, Ptgds expression was significantly reduced in 129 Dax1−/Y gonads, even without the influence of the B6 genetic background (n = 11; Fig. 4, gray bar in graph). A further reduction was apparent in 129;B6 Dax1−/Y gonads (n = 14; Fig. 4, white bar in graph), the levels of which were similar to that in female gonads (n = 15; Fig. 4, diagonally striped bar in graph). Because Ptgds expression was already attenuated in mutants on the pure 129 strain, this may explain why there was not a spectrum of expression in 129;B6 Dax1−/Y gonads. In summary, both in situ hybridization studies and quantitative real-time PCR demonstrate that Ptgds expression is attenuated in the absence of Dax1, and this reduction occurs regardless of genetic background.
Prior to meiosis, local retinoic acid (RA) accumulation induces expression of Stra8 (stimulated by RA) in primordial germ cells (PGCs) . Since ectopic meiotic entry of germ cells was observed in 129;B6Dax1−/Y gonads, it was feasible that Dax1 mutant gonads expressed inappropriate levels of Stra8. The expression levels of Stra8 were measured in 129Dax1−/Y and 129;B6Dax1−/Y gonads (Fig. 5) at 13.0 and 14.5 dpc, the time in embryonic development when Stra8 expression is initiated in the female gonad (n = 12; Fig. 5, white bars, relative control group) . Whereas Stra8 transcripts were not detected in wild-type males (n = 6; Fig. 5, diagonally striped bars), nor were they detected to a significant extent in 129Dax1−/Y gonads (n = 6; Fig. 5, gray bars), surprisingly, Stra8 expression was upregulated as early as 13.0 dpc in 129;B6Dax1−/Y gonads (n = 7; Fig. 5, black bars). Thereafter, by 14.5 dpc, 129Dax1−/Y gonad Stra8 levels (n = 6) reached nearly the same level as those in wild-type females (n = 10), whereas the level in 129;B6Dax1−/Y gonads (n = 6) exceeded that of the females. Therefore, upregulation of Stra8 occurs in Dax1 mutants and is further augmented on the B6 background.
Histologic observation of cellular morphology in 129;B6Dax1−/Y gonads that were susceptible to sex reversal indicated that apoptosis was prevalent in the gonad. To clarify which cell types were undergoing programmed cell death, double immunostaining with the Sertoli marker, anti-Müllerian hormone (AMH; red), and the germ cell marker, phosphorylated histone H2AX (gammaH2AX; yellow), combined with TUNEL labeling (green) of apoptotic cells was undertaken (Fig. 6). Anti-Müllerian hormone was localized to the cytoplasm of Sertoli cells located in the wild-type male gonad (Fig. 6A, within the testis cords) and also in 129;B6Dax1−/Y gonads (Fig. 6B). In wild-type females AMH was characteristically absent (Fig. 6C). Labeling of AMH along with TUNEL detection in gonad tissue sections showed that apoptosis was prevalent in the Sertoli cells of 129;B6Dax1−/Y gonads (Fig. 6B). Cell death was not observed in wild-type male or female gonadal tissue. For the wild-type male and female groups, n = 4 animals were tested; for the 129;B6Dax1−/Y group, n = 9 animals were tested. All mutant gonads tested exhibited Sertoli cell apoptosis. On average, 129;B6Dax1−/Y gonads contained approximately 87 ± 28 TUNEL-positive cells per gonadal field (Fig. 6, graph). Nuclei were stained with 4′,6′-diamidino-2-phenylindole (blue) on all sections. Notably, TUNEL-positive Sertoli cells were not detected in the pure background 129Dax1−/Y gonads (data not shown).
In addition, double immunostaining to detect gammaH2AX-positive germ cells (yellow) and apoptotic cells (green) did not overlap in 129;B6Dax1−/Y mutants (Fig. 6E), indicating that the germ cell population was not the major cell type undergoing apoptosis. Wild-type male and female served as negative and positive controls, respectively, for gammaH2AX staining (Fig. 6, D and F). These data indicate an essential role for Dax1 in combination with B6 alleles in Sertoli cell survival.
Humans carrying genetic mutations in DAX1 are characterized by defects in adrenal function and aberrant gonadal differentiation . To begin to understand the basic mechanisms underlying the origins of gonadal defects associated with DAX1 mutation, a model of Dax1 loss of function was used in which a change in the genetic background augmented the phenotype from male hypogonadism to complete male-to-female sex reversal. By examining the intermediary phenotype of Dax1 mutants on a 129;B6 background, we have shown that Dax1 plays an important role in Sertoli cell differentiation and survival in the bipotential gonad. Dax1 promotes the male pathway through modulation of Sox9 expression (directly or indirectly), a factor that is known to be required for Sertoli cell differentiation [23, 24]. In the absence of Dax1, Sox9 expression is sensitive to the B6 genetic background, as revealed by reduced (as shown in this work) or abrogated  expression. Of note, Sox9 was also markedly reduced in the M. domesticus poschiavinus strain in the absence of Dax1 . Although these findings were initially attributed to a weakened Sry allele characteristic of the poschiavinus strain, it is also plausible that autosomal genes, such as the chromosome 4 gene tda1, contributed to sex reversal in the absence of Dax1 .
Here, Dax1 deficiency was shown to result in Sertoli cell apoptosis in developing gonads that were susceptible to sex reversal by the influence of the B6 background. It is not clear whether apoptosis was cell autonomous in the Sertoli lineage due to the loss of Dax1 or, if in combination with modifying factors, paracrine cell-cell signals triggered an apoptotic cascade. The milieu of meiotic germ cells might also have contributed to apoptosis in Sertoli cells of 129;B6Dax1−/Y gonads.
Ptgds was shown to require Dax1 for its full expression, since it was significantly downregulated in 129 pure background mutants. Its role in the testis may explain some of the phenotypic features associated with sex reversal in 129;B6Dax1−/Y gonads. For example, it has been shown that dimeric Sox9 binds to the Ptgds promoter . Moreover, exogenous prostaglandin D2 addition to XX gonad explants in culture can induce ectopic Sox9 expression . Thus, a feed-forward loop is thought to exist between Sox9 and Ptgds. Whether Dax1 plays a role in this feed-forward loop requires further investigation.
Interestingly, Stra8 expression was upregulated in both 129Dax1−/Y gonads (male development) and in 129;B6Dax1−/Y gonads (exhibiting sex reversal), suggesting the possibility that Dax1 is a repressor of factors controlling Stra8 expression. The Stra8 gene is expressed in embryonic germ cells of female gonads  in response to the presence of local RA. Retinoic acid plays an important regulatory role in the control of germ cell meiotic entry . One explanation for this consistent upregulation of Stra8 in Dax1 mutants regardless of genetic background is impaired peritubular myoid development . Peritubular myoid cells and the basal lamina may provide a physical barrier that prevents RA diffusion into the compartment of testis cords, thus protecting male germ cells from RA exposure. Additionally, it is thought that RA exposure is circumvented in the male gonad by the presence of CYP26B1 , an RA-degrading enzyme expressed in Sertoli cells during development  and in adult peritubular myoid cells . On the 129 genetic background, Cyp26b1 mRNA levels (12.5 and 13.5 dpc) were not statistically different in Dax1 mutants and wild-type males (data not shown). Therefore, a reduction in Cyp26b1 alone does not account for the measured increase in Stra8 on the 129 background. In congenic B6 Dax1−/Y mice, the level of Cyp26b1 expression was reduced to that of females . Further developmental studies of Cyp26b1 expression in different cell types will be of interest to illuminate the pathways that govern differential germ cell development in the testis and ovary.
It should be emphasized that not all the defects observed in Dax1 mutant gonads presented herein can be attributed to the B6 strain. As previously noted, both Ptgds and Stra8 were strikingly deregulated in pure 129 mutants. On the other hand, Dax1 mutant gonads on a pure 129 background showed no evidence of Sertoli cell apoptosis or germ cell meiosis. Thus, the B6 background affects the control of gene expression levels in different cell types during gonadal development. In summary, these studies of Dax1 mutants on a mixed genetic background provide insights into the pathways governing cell fate decisions that determine gonadal sex. The first critical factor in male determination is Sry, followed by Sox9, both key elements in Sertoli cell fate determination. Sox9 levels are subject to modulation, and a threshold level appears to be important for the male pathway to proceed. This suggests that Sox9 dosage may be a critical determinant of somatic cell fate, as modulation of Sox9 expression levels on the B6 background correlates with a male versus female gonadal fate. In parallel, germ cell meiosis during development hinges on the absence or downregulation of these male factors. Thus, in the bipotential gonad, the male and female pathways coexist until a developmental “tipping point” is quickly established . Indeed, if a distinct fate choice is not enforced during development—as in the case of Dax1 mutants of a mixed 129;B6 background—a spectrum of phenotypic combinations bearing both male and female features can develop. These features underscore the likelihood that many sex-determining genes function in a dosage-sensitive manner  and that genetic background influences the quantitative expression of these genes. Moreover, gonadal architecture, such as the formation of testis cords, is important to support Sertoli cell maturation in order to create niches for germ cells .
Imaging and microscopy expertise was provided by Dr. Teong-Leong Chew, Dr. Paul Cheresh, and Lennell Reynolds of the Northwestern University Cell Imaging Facility. The authors wish to thank Dr. So-Youn Kim of the Jameson Laboratory for indispensable aid.
1Supported by National Institute of Health grant R01HD04481. S.Y.P. is a recipient of a Dolores Zohrab Liebmann Fellowship.