Although sexual dimorphism manifests most obviously in the gonads (testis and ovary), it can be observed in the other parts of the entire animal body. For instance, it is well known that several tissues such as the external genitalia, muscle, and brain exhibit sexual dimorphisms in terms of their structures and functions. This process of sex differentiation can be divided into three steps. The first occurs at fertilization, during which the sexes of fertilized eggs are determined genetically by combination of sex chromosomes. As the second step, the individuals carrying XY and XX sex chromosomes develop the testis and ovary, respectively. Sex differentiation of the gonads usually proceeds during fetal stages. Thereafter, the gonads control the sexes of the whole body through the function of sex steroids synthesized in the sexually differentiated gonads. Therefore, the gonadal sexes are quite important for sex differentiation of the animals.

It has been established that a number of transcription factors play crucial roles in the process of gonadal differentiation. Some of these factors, such as SRY, WT1, DAX-1, SOX9 and ARX were identified as the products of genes responsible for human diseases that display structural and functional defects in tissues including the gonads. The functions of the other transcription factors such as Ad4BP/SF-1, Emx2, M33, and Lhx9 were defined by the phenotypes of the gene disrupted mice. In addition, the expression profiles with respect to their distribution and sexual dimorphism strongly suggested the functional significance at the early stage of gonadal differentiation. However, it remains to be clarified how the transcription factors above regulate their target gene transcription and how the genes encoding the transcription factors are regulated by upstream regulators. Studies from above two directions are quite important to define the gene regulatory cascade and the molecular mechanisms that supports sex differentiation of the gonad.

I. Function of AhR (dioxin receptor) in the ovary.

It has been well documented that dioxins acts as an endocrine disruptor through estrogenic action. However, the molecular mechanisms underlying this action have been largely unknown. We have clarified it by focusing on the reproductive defects displayed in the AhR gene disrupted mouse. The gene disrupted female mice showed significantly reduced fertility probably due to disordered estrus cycle, low concentration of ovarian estradiol, and reduced numbers of ovulated eggs. Since estradiol is a representative sex steroid of female mainly produced in the ovary by successive reactions of steroidogenic enzymes, the expression of these steroidogenic genes were investigated, and found that the expression of cytochrome P450 aromatase catalyzing the final step of estrogen synthesis is affected significantly in the gene disrupted mouse. Showing a good correlation, estrogen treatment rescued the reduced ovulation capacity. This remarkable phenotype suggested that AhR regulates the P450 aromatase gene transcription. Indeed, reporter gene analysis indicated that the mouse and human P450 aromatase genes were activated by AhR.

Fig. 1, During maturation process of eggs, estrogen (E2) stimulates follicular growth and rapture. AhR together with Ad4BP/SF-1 were elucidated to regulate P450 aromatase gene expression.

Our previous study revealed that Ad4BP/SF-1, a member of nuclear receptor, is expressed specifically in the steroidogenic tissues including the ovary, and implicated in regulation of the steroidogenic genes. Therefore, it was investigated whether AhR activates the P450 aromatase gene transcription synergistically with Ad4BP/SF-1. Interestingly, reporter gene assays showed synergy between the two factors. Showing a good correlation, these two factors were indicated to interact mutually, and moreover chromatin immunoprecipitation assays with anti-AhR and anti-Ad4BP/SF-1 antibodies revealed that both factors bind their recognition sequences on the aromatase promoter and formed a protein complex in the ovarian cells. Thus, together with the AhR KO phenotype, it was concluded that AhR regulates the P450 aromatase gene transcription with Ad4BP/SF-1, and thereby regulates estrogen synthesis in the ovary (Fig. 1). This finding clearly elucidated that dioxins through binding to AhR exert estrogenic action by upregulating P450 aromatase gene expression in the ovary. (This study was performed as collaboration with Prof. Fujii-Kuriyama (Tsukuba University).)

II. Characterization of factors interacting with Ad4BP/SF-1

In order to define the function of Ad4BP/SF-1, we have screened a two-hybrid library prepared form mouse embryonic gonads. UBC9, PIAS1 and PIAS3, all of which have been indicated as components for sumoylation reaction (conjugation of SUMO (small ubiquitin-like modifier)), were isolated as the interacting proteins. Thus, we investigated whether Ad4BP/SF-1 is sumoylated in vivo and in vitro. Immunoprecipitation study from cultured cells showed that at least a part of Ad4BP/SF-1 is sumoylated in cultured cells. In vitro reconstitution study using purified components revealed clearly that sumoylation occurs at two lysine residues, K119 and K194, of Ad4BP/SF-1. Moreover, interestingly, sumoylation at K194 is essential for the second sumoylation at K119, suggesting that conjugation of SUMO induces conformational alteration of Ad4BP/SF-1.

With respect to the physiological function of SUMO, we have investigated whether sumoylated Ad4BP/SF-1 differentially activates transcription, since sumoylation consensus sequence is overlapped with that of synergy control region originally defined as a suppression domain for synergistic transcriptional activation. Ad4BP/SF-1 carrying mutation at the sumoylation site enhanced synergistic transcription with reporter gene driven by multiple Ad4 binding sites. It was established that Müllerian inhibiting substance (MIS) gene is regulated by SOX9 synergistcally with Ad4BP/SF-1. Interestingly, SOX9 has a sumoylation consensus sequence and in fact sumoylated as well. Therefore, we asked whether sumoylation has effects on the synergy between the two transcription factors. Expectedly, unsumoylated forms activated the synergistic transcription of the MIS gene more than wild type. To define the molecular mechanisms underlying the synergy enhancement, we asked with in vitro sumoylated Ad4BP/SF-1 whether sumoylation affects DNA binding activity and interaction with SOX9. However, sumoylation was found to have no effect on both aspects, strongly suggesting the presence of a putative synergy control factor (Fig. 2). Although the physiological significance of sumoylation has been controversial, identification of factors interacting preferentially with the sumoylated Ad4BP/SF-1 such as the synergy control factor will probably give us quite valuable information. (This study has been performed as collaboration with Profs. Kikuchi (Hiroshima Univ.), Shirakawa (Yokohama City Univ.), and Yamanaka (Kyushu Univ.).)

Fig. 2, Sumoylation (conjugation of SUMO) consists of three step reactions successively catalyzed by UBA2-AOS1 (E1), UBC9 (E2), and PIAS (E3). Reporter gene assays indicated that unsumoylated form of Ad4BP/SF-1 enhances synergistic transcription. This observation suggested the presence of a putative synergy control factor (SCF).

III. Function of growth factors during early stages of gonad differentiation.

The gonad as well as the reproductive tracts, kidney, and adrenal cortex are derived from a part of the intermediate and lateral plate mesoderms. In addition, the mesonephros essential for the early gonadal development also originates from the intermediate mesoderm. Although the mesonephros was elucidated as the source of certain cell types constituting the testis, its contribution to gonad formation has been largely unknown. We examined in detail the expression profiles of Wnt4, Fgf8, and Fgf9 in the developing mesonephros of chick embryos, and found that the expressions of these factors are spatially and temporally correlated with those of the marker genes for the gonad and/or adrenal cortex such as Ad4BP/SF-1 and Dmrt-1 (doublesex- and mab-3-related transcription factor). As was shown previously with the rat fetuses, it was confirmed with the chick embryos that these two tissues are derived from a single cell population, the adreno-gonadal primordium. Thereafter, the primordium divided into two cell populations, one of which is the gonadal primordium and the other of which is the adrenal primordium. All of these cells are identified as immunoreactive cells for Ad4BP/SF-1. The expression of the growth factors, Wnt4, Fgf8, and Fgf9, are detectable in the nephroginous mesenchyme (feature mesonephros) and developing nephric tubules at the stage of the early gonad and adrenocortical development. Interestingly, the expressions of Wnt4 and Fgf9 are correlate with those of Ad4BP/SF-1 and Dmrt1. Therefore, it was assumed that these growth factors regulate the expression of the gonadal and adrenocortical marker gene expressions.

In order to define it, Wnt4 and Fgf9 were misexpressed with replication competent virus system in the gonad and adrenal cortex of the early developmental stages. Misexpression of Wnt4 expanded the expression of Ad4BP/SF-1 as the marker for the adrenogonadal primordium while failed to expand the expression of Dmrt-1. Interestingly, misexpression of Fgf9 expanded the expression of Dmrt-1 and probably thereby resulted in additional gonad formation. In fact, the expressions of marker genes for the gonad were observed even in the additional gonad (Fig. 3). Moreover, chemical inhibitor for Fgf receptor downregulated the Dmrt-1 expression. Both gain-of-function and loss-of-function treatments above induced no obvious structural alteration of the adrenal cortex, indicating that Fgf9 induced gonad development probably through stimulating cell proliferation or alternatively through trans-determining cell fate into the gonad. This study demonstrated that Wnt4 and Fgf9 expressed in the developing mesonephros are involved in the early stages of the gonad and adrenocortical development through distinct functions.

Fig. 3, Misexpression of Fgf9 induced additional gonad formation as indicated by arrows. Ad4BP/SF-1 and Mis are both expressed in the additional gonads.

In the case of mammals, a gene disruption study indicated that Fgf9 is essential for testicular differentiation. However, in the chick the misexpression of Fgf9 was unable to induce the expression of the testicular marker gene in the female gonad. This clear discrepancy might be due to the difference of experimental conditions, one of which is gene disruption (loss-of-function) study and the other is misexpression (gain-of-function) study, or due to difference of animal species. (This study has been performed as collaboration with Dr. Yoshioka in Hyogo Univ. of Teacher Education.)

In addition to the studies described above, we have investigated tissue-specific enhancers localized in the Ad4BP/SF-1 gene. Recently, we identified a fetal adrenocortical-specific, pituitary-specific, and ventro-medial hypothalamic nucleus-specific enhancers, and their fine structures have been clarified. Investigation of the factors to bind the enhancers will be essential to clarify the mechanism underlying the tissue differentiation.

It has been known that several gene disruption studies revealed their implication into gonad differentiation and gonad sex differentiation. However, functional and genetic correlation among the genes have remained to be investigated. As the investigation of the genetic correlation, analyses for genetic interaction have been performed with a certain lines of gene disrupted mice, of which genetic backgrounds have been fixed into C57B/6 and FVB in addition to the original 129J. Some of the analyses already indicated the presence of the genetic interaction, and the molecular basis for the interaction is under investigation.