INTRODUCTION

In vertebrates, the Y chromosome gene, SRY, has been identified as the testis-determining gene in mammals (Sinclair et al., 1990; Gubbay et al., 1990), however until recently, no equivalent gene or sex-determining genes had been identified in non-mammalian vertebrates. In the medaka, Oryzias latipes, which has an XX/XY sex-determining system (Aida, 1921), a Y-linked gene, DMY, was found to be a prime candidate for the sex-determining gene using positional cloning. DMY encoded a protein that contains a DM-domain, which was originally described as a DNA-binding motif found in two proteins, DSX and MAB-3, involved in sexual development in Drosophila melanogaster and Caenorhabditis elegans, respectively (Raymond et al., 1998). Vertebrates have also several DM-domain genes, and one of these, DMRT1 (DM-related transcription factor 1), has been implicated in male sexual development in mammals, birds, reptiles and fish (Raymond et al., 1999; Smith et al., 1999; De Grandi et al., 2000; Guan et al., 2000; Kettlewell et al., 2000; Marchand et al., 2000; Moniot et al., 2000). The cDNA sequences of medaka DMY and DMRT1 indicated the high similarity (83%) and DMY appears to have originated through a duplication event of an autosomal segment containing the DMRT1 region (Nanda et al., 2002).

DMY expression was observed only in genotypic (XY) males during gonadal sex differentiation, and a loss-of-function DMY mutation and depressed DMY expression mutation resulted in XY females (Matsuda et al., 2002). When a DNA fragment containing DMY was introduced as a trans-gene, testis developed in chromosomally (XX) female medaka (Matsuda et al., manuscript submitted for publication). These results demonstrate that DMY is both necessary and sufficient in determining testis formation in medaka. DMY is the first sex-determining gene to have been identified in non-mammalian vertebrates.

Sex-reversal mutants are important in revealing the molecular function of DMY and identifying other genes that are involved in sex determination. Analyses of these mutants allow us to extend our understanding of the function of genes and cell differentiation during sex determination. Two XY sex-reversed females with naturally occurring DMY mutations have already been found by screening wild medaka populations (Matsuda et al., 2002). This suggests that other spontaneous gene mutations involved in the sex-determining pathway can be identified from surveys in the wild.

A variety of sex-determining systems can coexist in closely related species or even within a single lower vertebrate species. Both the male-heterogametic XX/XY and female-heterogametic ZZ/ZW systems have been investigated within the Xiphophorus genus (reviewed in Volff and Schartl, 2001) and at the species level in Rana rugosa (Nishioka et al., 1993, 1994). Wild medaka populations comprise four genetically divergent groups (Sakaizumi, 1986; Sakaizumi and Joen, 1987; Sakaizumi et al., 1983), while the presence of DMY has been confirmed only in several medaka strains and wild populations classified in the Northern or Southern Populations (Matsuda et al., 2002). Therefore, it is possible that some populations have a sex-determining system that is not determined by DMY. On the other hand, artificial sex-reversals as a result of exogenous sex steroids (Yamamoto, 1953, 1958) or heat stresses (Gresik and Hamilton, 1977) have been reported in medaka. This suggests that in addition to the role of sex-determining genes, the natural environment might affect the phenotypic sex of wild medaka.

In this study, we conducted a broad survey of the DMY gene in wild medaka populations to locate any new gene mutations involved in the sex-determining pathway. Consequently, we report that approximately 1% of the wild medaka surveyed were sex-reversals and that all XY sex-reversals examined had a DMY or DMY-linked mutation.

MATERIALS 0D METHODS

Fish: We surveyed 2274 wild-caught fish at 40 localities throughout Japan (Table 1), collected from 2001 to 2003, and 730 fish from 69 wild stocks at the Faculty of Science, Niigata University. The wild stocks were collected from 55 localities in Japan (#41 to #95 in Table 2), 12 in Korea (#96 to #107), one in China (#108), and one in Taiwan (#109) between 1986 and 2001, and have been maintained thereafter.

Table 1

Phenotypic and genotypic sexes of the wild-caught medaka

Table 2

Phenotypic and genotypic sexes of the wild stocks

Sexing of the wild fish: Phenotypic sex was judged from secondary sex characters, namely, the shape of the dorsal and anal fins, and papillary processes on the male anal fin rays. Genotypic sex, XY or XX, was determined by the presence or absence of the DMY gene using PCR from caudal fin clip DNA extracted according to Shinomiya et al. (1999). PCR was performed with the following primers for DMY and DMRT1: PG17.5, CCGGGTGCCCAAGTGCTCCCGCTG, and PG17.6, GATCGTCCCTCCACAGAGAAGAGA at an annealing temperature of 55°C. PCR products were analyzed by electrophoresis in a 1% agarose gel. Other primers for DMY, ex3.1, GCAACAGAGAGTTGGATTTACGTCTCA, ex3.2, CTTTTGACTTCAGTTTGACACATCAATG, ex6.1, GTCATTAACACAACGCACAACAACTT, and ex6.2 AAAAACCAGAAGACCCGAGAGGAAG were also used (see Results). The positions of these primers in the DMY gene are shown in Fig. 1A.

Fig. 1

Genotypic sexing with PCR. (A) DMY structure of the Hd-rR.Y^HNI strain and positions of the primers used. Open boxes and horizontal lines indicate exons and introns respectively. Translation start and stop sites are shown by ATG and TGA. Numbers represent nucleotide sequence length (bp). The primer positions are illustrated by solid arrowheads. (B) One percent agarose gel electrophoresis of the DMY and DMRT1 PCR products. Individuals with only the DMRT1 band were judged XX (lanes 2, 4, 6, 8), while those with both DMRT1 and DMY bands were judged XY (lanes 3, 5, 7, 9).

Progeny tests: All XY females from the wild populations were mated with XY males of an inbred strain, from either Hd-rR (Hyodo-Taguchi, 1996) or Hd-rR.Y^HNI (Matsuda et al., 1998). The F₁ progeny from each pair were grown, and their phenotypic and genotypic sexes were determined. The DMY gene of the Northern Population had 21 nucleotide deletions in intron 2 compared to the Southern Population. Therefore, XY females from the Northern Population were mated with XY Hd-rR males (Southern Population), and those from the Southern Population were mated with XY Hd-rR.Y^HNI males, which had the HNI(Northern Population)-derived DMY gene. Four genotypes were distinguished in the F₁ progeny (XX, XY^m, XY^p, Y^mY^p; Y^m, maternal and Y^p, paternal) by separating the DMY PCR products with 10% vertical polyacrylamide slab gels in the buffer system of Davis (1964).

RESULTS

Progeny of the XY females

To clarify the cause of the XY sex-reversals, a total of 16 XY sex-reversals from 11 localities were mated with XY males of an inbred strain, either Hd-rR (females from #3, #8, #9, #17, #23, #26, and #48), or Hd-rR.Y^HNI (#51, #78, #83, and #90). The genotypic and phenotypic sexes of the F₁ progeny were then analyzed (Table 3). All XX individuals in the F₁ progeny from each pair were female and all XY^p and Y^mY^p were male, while all or some of the XY^m were female. The XY^m F₁ from Aomori2 (#3), Aizu-bange (#9), Suzu (#23), Awara (#26), Kurobe (#48), Saigo (#78), Aki (#83) and Oura (#90) were all female, while the XY^m F₁ of Aizuwakamatsu (#8), Shirone (#17) and Kesennuma (#51) contained both males and females. In all crosses, the occur-rence of XY sex-reversals in the F₁ progeny was linked to the maternal Y chromosome, while the paternal Y chromo-some derived from the inbred strains resulted in all F₁ male individuals (XY^p, Y^mY^p).

Table 3

Genotypic and phenotypic sexes of the F1progeny from the XY females

DISCUSSION

In the present study, we conducted a broad survey of the genotypic sex of a total of 3004 wild medaka from 109 localities covering the four genetically divergent groups of this species (Sakaizumi, 1986; Sakaizumi and Joen, 1987; Sakaizumi et al., 1983). Of the 1540 males, 1525 (99%) had the DMY gene (XY), while of the 1464 females, 1438 (96%) did not (XX). This intimate relationship between male and female gender and the presence or absence of the DMY gene strongly suggests that DMY is a common sex-determining gene in wild medaka populations.

Multiple fragment patterns were identified in the PCR products of DMY. In most populations a long fragment was detected, while individuals from Shincheon (#100) and Kunming (#108) had a short fragment. Fifteen out of 16 males from Gwangeui (#101) had both fragment types (Fig. 1B) (one male had no DMY fragments), indicating that the Gwangeui population had two DMY gene on the Y chromo-some. The short fragment was only detected in populations classified in the China-West Korean Population. Furthermore, molecular phylogenetic analysis of the DMY gene among the four major medaka groups demonstrated that the two types of DMY from the China-West Korean Population were most closely related. In XY males from Gwangeui with two types of DMY, only the short type was expressed at the sex-differentiating stage (Sato et al., in preparation).

The frequency of sex-reversed fish in wild-caught medaka was 1.0% (23 out of 2274, Table1), while 18 sex-reversals were found in eight of the 69 wild stocks (Table 2). All XY sex-reversals (n=16) produced XY females in their F₁ progeny, demonstrating that all these sex-reversals resulted from a gene mutation involved in sex-determination. This suggests that XY sex-reversal resulted from environmental factors hardly occurs in natural habitats. All XY F₁ females had maternally-derived DMY genes, indicating that the mutation was linked to the DMY gene. The mutant gene is inferred in DMY itself because DMY is the only functional gene in the Y-specific segment (Nanda et al., 2002).

Two types of XY sex-reversals existed with regards to the sex of the XY F₁ progeny. Twelve out of 16 XY females produced an all-female XY^m progeny, while four yielded both male and female XY^m progenies. One XY female from Awara that produced an all-female XY^m progeny (#26 in Table3) was found to have a DMY mutation causing a frameshift and premature termination of the DMY protein. Another XY female from Shirone that yielded a high proportion of XY^m female offspring (#17) had reduced DMY expression (Matsuda et al., 2002). These results suggest that the XY females found in this study are a promising tool for gaining an understanding of the biochemical function and regulatory regions of the DMY gene, just as SRY mutants have in human studies (reviewed in Koopman, 2001; McElreavey et al., 1995; Vilain and McCabe, 1998). The sequences and expression patterns of the DMY gene of other XY sex-reversals are now under investigation.

In humans, approximately 15% of XY females have mutations in the coding region of SRY (Hawkins, 1995), and XY sex-reversals with mutations in other genes related to testis development have also been reported (reviewed in Ahmed and Hughes, 2002; Vilain and McCabe, 1998). On the other hand, all XY sex-reversals in medaka were linked to the DMY gene. A loss-of-function DMY mutation is thought to induce dominant sex-reversal in XY individuals, and furthermore, to be transmitted from XY females to their offspring because XY medaka females are fertile (Yamamoto, 1953) unlike mammals. DMY mutations might not be harmful and thus survive in natural habitats and wild stocks.

Fifteen XX males were observed in this survey. One of these XX males produced XX males in its backcross progeny, suggesting that XX sex-reversal results from a recessive mutation(s). We are now mapping the gene(s) that causes XX sex-reversal, which is expected to be involved in sex-determination.