Primary sex determination initiates the development of the bi-potent gonad into either a testis with SOX9 expressing Sertoli cells in the medulla or an ovary with aromatase expressing granulosa cells in the medulla. In the chicken, DMRT1 is clearly a key factor in primary sex determination (Ioannidis et al., 2021; Smith et al., 2009), but the exact mechanism underlying this process is unclear. In vivo RNAi mediated knock-down of DMRT1 expression has provided the functional validation of DMRT1 as the dosage sex determinant in chicken (Smith et al., 2009). We now, for the first time, show that DMRT1 expression can be regulated in protein levels during the gonadal sex determination in chicken and that estrogen signalling can regulate its protein levels post-translationally. A recent gene-editing approach has confirmed that two functional copies of the Z chromosome DMRT1 gene are required for testis development in the wild type of male chicken: the DMRT1 heterozygous male developed an ovary (Ioannidis et al., 2021). However, following aromatase inhibitor treatment, the DMRT1 heterozygous male developed testis. Clearly, primary sex-determination in birds is not simply dependent on gene dosage. For this reason, we carried out a detailed examination of DMRT1 RNA and protein expression during gonadal development in the chicken. We now further show that estrogen signalling is involved in gonadal sex determination by suppressing DMRT1 protein expression and preventing the testis formation in the gonadal medulla during ovarian development in birds.
Our analyses using Northern and q-PCR approaches excluded the possibility that the sex determination mechanism depended on expression of different alternatively-spliced DMRT1 transcripts in males and females, and also confirmed that DMRT1 transcription was consistently higher in male than in female gonads (Fig. 1A&B). Western analysis of DMRT1 protein in the developing gonads, confirmed that higher levels of DMRT1 transcription results in higher levels of DMRT1 protein, with the exception of E4.5 gonads where DMRT1 protein levels are similar in male and female gonads (Fig. 1C). However, it is difficult to make definitive conclusions on relative male/female RNA or protein levels, as DMRT1 is expressed in different cell types in the gonads, and even in the initial stages of gonadal differentiation, the male and female gonads are morphologically distinct.
To enable comparison of DMRT1 protein levels in individual cell types, we carried out an IHC analysis of DMRT1 expression in male and female gonads between E4.5 and E12.5 (Fig. 1D). In early gonadal development, DMRT1 protein levels appear to be similar in the somatic cells of both male and female gonads. However, by E8.5 there is an obvious difference in DMRT1 protein levels in male and female gonads: in the somatic cells of the male medulla, DMRT1 is expressed at higher levels than in cells of the female medulla. These differences in male and female DMRT1 protein expression are maintained at E10.5 and E12,5. Between E8.5 and E12.5, male/female difference in protein levels seem to correlate with the male/female difference in DMRT1 transcription levels reported here and elsewhere. However, our results differ from those expected in that the male/female difference in protein levels seems to be due to a decrease in protein levels in female gonads rather than an increase in male gonads i.e. signal intensity is similar in male and female gonads in the early stages and while similar levels are maintained in the male gonads at later stages, levels decrease in female gonads. Analyses of gonads from mixed-sex chimeras supported this conclusion (Fig. 2). In general, it is assumed that a double gene dose of DMRT1 leads to higher levels of DMRT1 transcription and higher levels of DMRT1 protein in developing male gonads than in developing female gonads, and that this induces testicular development in male embryos. Our data would appear to conflict with this assumption: it suggests that DMRT1 protein levels are similar in male and female (even DMRT1 heterozygous male, our unpublished data) medullas in the early stages of gonadal development, and that the suppression of DMRT1 protein by estrogen signalling during the female (even DMRT1 heterozygous male, our unpublished data) gonadal sex determination leads to ovarian development (Fig. 6).
As we had previously shown that loss of a single functional copy of DMRT1 leads to the induction of FOXL2 in the developing gonads of male embryos, we compared the expression of DMRT1 and FOXL2 proteins in the developing left ovary of female embryos (Fig. 3). Our analysis showed that in the early stages of ovarian development, when the somatic cells of the medulla display ‘male-like’ levels of DMRT1 protein, expression of FOXL2 protein is restricted to a narrow area immediately below the epithelium on the ventral side of the gonad. As ovarian development progresses, DMRT1 protein levels decrease in a wave from the ventral aspect towards the dorsal aspect of the medulla, and this is complemented by a ventral-to-dorsal wave of increasing numbers of cells expressing FOXL2 protein. By manipulating the aromatase inhibitor treatment, we generated partial female to male sex-reversed gonads (Fig. 5B). Interestingly, those sex-reversed gonads showed that the ventral part of medulla expressed low DMRT1 protein levels and the female markers, FOXL2 and aromatase, and that the medullary ‘feminization’ process appeared to be blocked in the dorsal region showing the expression of high DMRT1 protein levels and sex cord expressing the male marker, SOX9.
To investigate the relation between DMRT1 protein levels and estrogen signalling, we examined activation of estrogen signalling in the E8.5 mixed-sex chimeric gonads. In these chimeric gonads, aromatase is expressed in the somatic medullary cells of the female portion. Our IHC staining showed that the estrogen receptor was activated (nuclear) in male and female epithelia cells. And more importantly, it showed clearly for the first time that the estrogen receptor was activated in female gonadal medulla with low DMRT1 protein levels, but not activated in male gonadal medulla with high DMRT1 protein level (Fig. 4). This finding not only supports the notions that the epithelium of the male gonad can be induced to form an ovarian cortex in the presence of estrogen (Guioli et al., 2020), but also corelate the activation of ESR1 with the decrease of DMRT1 protein expression.
To further investigate the possibility that estrogen suppresses DMRT1 protein expression and enables ovarian development, we blocked estrogen signalling and examined the effect on DMRT1 expression in the female left gonads (Fig. 5A&B ). Inhibition of aromatase and/or ESR1 activity in the developing ovary led to gonadal sex-reversal with DMRT1 protein levels remaining high and the differentiation of gonadal supporting cells into SOX9 + ve Sertoli cells and the formation of sex cords in the medulla. Both female markers, FOXL2 and aromatase, were expressed only in a limited number of cells in the ventral medulla of sex-reversed gonads. These findings are consistent with reports on other non-mammalian vertebrates, where exposure to estrogen results in male-to-female sex-reversal and inhibiting estrogen leads to female-to-male sex-reversal (Díaz and Piferrer, 2017; Ditewig and Yao, 2005; Nakamura, 2010, 2013; Romano et al., 2020; Wu et al., 2020). And we further showed that estrogen signalling is involved in the feminization of gonad medulla during avian ovariogenesis by suppressing DMRT1 protein (not RNA) expression.
This study provides a detailed characterisation of DMRT1 expression during gonadal development in the chick, and compared DMRT1 expression with FOXL2 expression and estrogen signalling. Our findings suggest that DMRT1 protein levels in female gonadal medulla are equivalent to levels in male during the bipotential stages of gonadogenesis and are reduced during the differentiation stages. The reduction in DMRT1 protein is concurrent with the FOXL2 expression, which occurs in a wave from the ventral to the dorsal aspects of female gonadal medulla. Analysis of sex-reversed gonads and mixed-sex chimeric gonads suggests that during ovarian development, in addition to inducing cortex expansion, estrogen signalling prevents Sertoli cell differentiation and sex cord formation in the medulla by suppressing DMRT1 protein expression. This is supported by the findings of an analysis of male birds with a single functional copy of DMRT1 (Ioannidis et al., 2021). These birds develop ovaries in place of testes. If estrogen signalling is blocked, the DMRT1 heterozygous male gonads develop as testes and express male levels of DMRT1 protein. It seems that DMRT1 holds a key position as the main switch to toggle Sertoli/granulosa cells fate(Lindeman et al., 2015; Matson et al., 2011) and estrogen signalling acts as the operator to activate the differentiation of medullary supporting cells into granulosa cells by suppressing DMRT1 protein expression in the developing ovary in birds. We therefore conclude that DMRT1 protein and estrogen signalling form a sex determination core and determines the developmental fate of the gonads in birds (Fig. 6).
It is well established that the epithelium of the left male gonad expresses ESR1 and has the potential to form an ovarian cortex if exposed to estrogen (Guioli et al., 2020). In the mixed-sex chimeric gonads analysed here, ESR1 was clearly activated in both male and female epithelial cells. However, in the case of the medullary somatic cells, ESR1 was only activated in the female, not in the male. It is noteworthy that, in the mixed-sex chimeric gonads, the female medulla and both the female and male epithelium express low levels of DMRT1 protein, while the male medulla expresses high levels of DMRT1 protein. It is also interesting that while epithelial cells of the male gonad show a paracrine response to estrogen, cells of the male medulla do not respond. Thus, the differentiation of gonadal epithelial cells is hormone dependent, whereas the differentiation of gonadal medullary somatic cells is cell autonomous, which can explain why male and female somatic cells in the medulla retain their own sex identity (CASI) in mixed-sex chimeric gonads (Fig. 4) (Zhao. D.; McBride et al., 2010).
Sex steroid may be the key factor for sex determination in low vertebrates (Nakamura, 2010). Estrogen signalling has been known to be required for sex determination, as well as folliculogenesis, in low vertebrates. The evidence available so far has shown that estrogen effects the DMRT1 RNA level first and then its protein level in fish, reptiles and amphibians. We have found a new mechanism that estrogen is involved in sex determination by suppression of DMRT1 protein level in birds, but not via RNA level as reported in other low vertebrates. It seems now that there are two different mechanisms for estrogen being involved in the sex determination: its effect on DMRT1 transcription and then translation in fish, reptiles ad amphibians and its effect on DMRT1 translation (or maybe post-translation, but not on its transcription!) in birds. Evolutionally, estrogen is no longer involved in sex determination in high vertebrates.
Our data suggests that testes or ovary formation is a result of different levels of DMRT1 protein in the developing gonads of male and female embryos in birds. However, it appears that the sex difference in DMRT1 protein levels is due to a reduction of DMRT1 protein expression via estrogen signalling in female gonads. Further investigations are needed to understand whether the suppression of DMRT1 protein expression by estrogen is involved in the gonadal sex determination in other species and how DMRT1 protein levels are regulated during avian gonadal sex determination. Sex specific stabilization of DMRT1 protein by ubiquitin binding might be one of ways to regulate DMRT1 expression (Bayer et al., 2020; Nakagawa et al., 2017).