INTRODUCTION

Sex-determining mechanisms can be demonstrated by (1) cytological observation of sex chromosomes, (2) sex-reversal and breeding experiments, (3) segregation analyses of sex-linked genes, and (4) H-Y antigen typing. Sex-determining mechanisms have been assigned in more than 50 species of amphibians by these methods (reviewed by Hillis and Green, 1990; Schmid et al., 1991; Solari, 1994).

Cytogenetic studies have determined the sex-determining mechanisms in the following species: XX/XY type in Triturus alpestris, T. vulgaris and T. helveticus (Schmid et al., 1979), Rana esculenta (Schempp and Schmid, 1981), Gastrotheca riobambae (Schmid et al., 1983), T. cristatus and T. marmoratus (Sims et al., 1984), Hydromantes italicus, H. ambrosii, H. imperalis, H. flavus and H. sp. nova (Nardi et al., 1986), G. ovifera (Schmid et al., 1988), Centrolenella antisthenesi (Schmid et al., 1989) and G. pseustes (Schmid et al., 1990), ZZ/ZW type in Leiopelma hamiltoni (Stephenson et al., 1972), Pyxicephalus adspersus (Schmid, 1980), and Xenopus laevis (Schmid and Steinlein, 1991), and XX/XY and ZZ/ZW types in Rana rugosa (Nishioka et al., 1994). The sex-determining mechanism was demonstrated by analyzing the sex ratio of the progeny of sex-reversed females in Ambystoma (Humphrey, 1942, 1945, 1957), Bufo bufo (Ponse, 1942), Bufo viridis (Ueda, 1990), Xenopus (Gallien, 1953; Chang and Witschi, 1955, 1956), and several Rana species (Kawamura and Nishioka, 1977). Sex-linked loci were also detected in Rana clamitans and R. catesbeiana (Elinson, 1981, 1983), the Rana pipiens complex (Wright and Richards, 1983; Wright et al., 1983), Pleurodeles waltl (Ferrier et al., 1983), P. poireti (Dournon et al., 1984), Xenopus laevis (Graf, 1989), the Rana nigromaculata group (Nishioka and Sumida, 1994) and Rana japonica (Sumida and Nishioka, 1994). The sex-determining mechanism was deduced in Pelodytes punctatus and four other species by the titer of the H-Y antigen found more abundantly in the heterogametic sex (Wachtel et al., 1975; Engel and Schmid, 1981).

The Japanese bell-ring frog, Buergeria buergeri, is an endemic species in which the sex-determining mechanism was confirmed both by cytological observation and by sex-reversal and breeding experiments (Ohta, 1986, 1987; Schmid et al., 1993). The sex chromosomes are of the ZZ/ZW type (female heteromorphic). The Z chromosome has a satellite and a nucleolar organizer at the tip of its long arm, whereas the W chromosome has none. Allozyme analysis of genetic divergence among geographic populations of this species revealed that the allelic frequencies at the AAT-1 locus differed between females and males of several populations (Atsumi et al., unpublished).

In the present study, genetic analysis of AAT-1 variants was carried out to elucidate the mode of inheritance of this locus using field-caught male and female frogs of the Hiroshima population and their offspring produced by artificial crossings.

MATERIALS AND METHODS

A total of 101 adult frogs (51 females and 50 males) of the Hiroshima population (from Ugakyo, Asa-cho, Hiroshima city) of Buergeria buergeri were used in the present study. Twenty males were collected in July, 1996, the other 30 males and 51 females in June, 1997. Skeletal muscle of each individual was used for starch-gel electrophoretic analysis by the method of Nishioka et al. (1980), and aspartate aminotransferase (AAT) was detected by the agar-overlay method outlined by Harris and Hopkinson (1976). Crossing experiments were performed by artificial insemination using two females homozygous and two males heterozygous at the AAT-1 locus. The embryos and tadpoles were reared in a moist chamber at a constant temperature of 20∼22°C; the tadpoles were fed on boiled spinach. The tadpole stages follow those of Taylor and Kollros' (1946) table. The tail-tips of tadpoles at TK stage X∼XIII were fixed in Navashin's fluid and stained with Heidenhain's iron hematoxylin following Ohta's (1986) method. After the sex of each tadpole was determined by counting the number of nucleoli in the nucleus of individual cells (Fig. 1), the AAT-1 phenotypes were analyzed by starch-gel electrophoresis using the tail-tip muscles.

Fig. 1

Photomicrographs of tail-tip epidermal cells of tadpoles. (a) Tadpole having single nucleolus per nucleus (1-nu). (b) Tadpole having two nucleoli per nucleus (2-nu). Scale bars equal 10 μm. The arrows point to the nucleoli.

RESULTS

The electrophoretic analysis of AAT-1 of 101 field-caught female and male B. buergeri revealed six phenotypes controlled by four alleles (Table 1, Fig. 2). The electrophoretic bands corresponding to single alleles at the AAT-1 locus were named A, B, C and D in the order of mobility from fast to slow; the encoding alleles for bands A, B, C and D were named a, b, c and d, respectively. Phenotypes with two bands were given composite names; AB and AC were observed. Of 50 males, 35 were B in phenotype, two were C in phenotype, 12 were BC in phenotype, and one was AB in phenotype. On the other hand, of 51 females, 36 were B in phenotype, 13 were C in phenotype, one was A in phenotype, and one was D in phenotype. None of the 51 females showed phenotypes with two bands. As to the allelic frequencies, allele b was the highest in frequency, followed by allele c in both females and males. The two other alleles were either found only rarely or not at all. There were no significant differences between the actual and expected numbers of phenotypes in males, but there were in females (χ² = 29.5, P < 0.001%). If the AAT-1 locus is supposed to be sex-linked and situated on the ZZ/ZW-type sex chromosomes of B. buergeri, the expected numbers of each phenotype coincide with the actual numbers (Table 1).

Table 1

Phenotypes and allelic frequencies at the AAT-1 locus in the Hiroshima population of B. buergeri

Fig. 2

Electrophoretic patterns of aspartate aminotransferase (AAT) in 11 females and 11 males from the Hiroshima population of Buergeria buergeri. The two loci of AAT were shown by AAT-1 and AAT-2. The phenotypes of AAT-1 were scored for each individual at the upper margin by B, C or BC. The relative positions of bands encoded by alleles b and c at the AAT-1 locus were marked by b and c, respectively. The origin was shown by O.

Two field-caught females (Hir. ♀ 1, Hir. ♀ 2) having phenotype B at the AAT-1 locus were mated with two field-caught males (Hir. ♂ 6, Hir. ♂ 11) having phenotype BC at this locus (Table 2). Most of the eggs (89.1∼98.2% of the total number of eggs) cleaved normally, 85.9∼98.2% of the total number of eggs hatched normally, and 77.9∼91.3% of the total number of eggs became normally feeding tadpoles (Table 2). The sex of each tadpole at TK stage X∼XIII was determined by counting the number of nucleoli in the nucleus of individual cells (Table 3, Fig. 1). Of 664 offspring produced from four matings, 313 (47.1%) had single nucleolus per nucleus (1-nu; females), 342 (51.5%) had two nucleoli per nucleus (2-nu; males), and nine were 3-nu, 1-nu/2-nu mosaics or 2-nu/3-nu mosaics (Table 3). There were no differences between male and female tadpoles in viability. Analysis of the AAT-1 phenotypes (Table 4, Fig. 3) revealed that of 313 1-nu female tadpoles, 151 were B in phenotype, 161 were C in phenotype, and one was BC in phenotype. On the other hand, of 342 2-nu male tadpoles, 185 were B in phenotype, and 157 were BC in phenotype. Almost all females appeared to be b or c homo-zygotes, whereas there were nearly an equal number of bb homozygotes and bc heterozygotes in males. Thus it is evident that alleles on the Z chromosome of female parents are expressed, whereas those on the W chromosome are not.

Table 2

Developmental capacity in the crosses between females homozygous and males heterozygous at the AAT-1 locus

Table 3

Number of nucleoli in the offspring

Table 4

Phenotypes at the AAT-1 locus in the offspring

Fig. 3

Electrophoretic patterns of aspartate aminotransferase (AAT) in nine female and 10 male offspring and their female (F) and male (M) parents of Buergeria buergeri. The two loci of AAT were shown by AAT-1 and AAT-2. The phenotypes of AAT-1 were scored for each individual at the upper margin by B, C or BC. The relative positions of bands encoded by alleles b and c at the AAT-1 locus were marked by b and c, respectively. The origin was shown by O.

DISCUSSION

Chromosome pair No. 7 of Buergeria buergeri are sex chromosomes of the ZZ/ZW type, and Z chromosome is a SAT-chromosome with a secondary constriction that corresponds to a nucleolar organizer (Ohta, 1986, 1987). The nucleolar organizer is situated on the Z chromosome, and a satellite is attached to the end of the long arm. On the other hand, neither the secondary constriction nor the satellite are observed in the W chromosome, making the distinction of W chromo-some from Z chromosome easy. Because sex-determining genes are linked with a secondary constriction, males (ZZ) have two nucleolar organizers and females (ZW) have one nucleolar organizer (Ohta, 1986; Schmid et al., 1993). The present study revealed that the AAT-1 locus is sex-linked, and that its codominant alleles are expressed on the Z chromo-some, but not on the W chromosome. Thus it is evident that the AAT-1 locus of female offspring is hemizygous and that the allele present corresponds to that on the Z chromosome derived from the male parent. The mode of inheritance of the sex-linked AAT-1 locus in B. buergeri is the same as for sex-linked loci in birds (Baverstock et al., 1982) and butterflies (Johnson and Turner, 1979), in which sex-linked genes are carried exclusively on the Z or the X chromosome, respectively, resulting in hemizygous genes in the heterogametic sex. This is the first report of this mode of inheritance in amphibians.

As for sex-linked genes, several linkage relationships between sex-determining genes and enzyme loci have been reported in amphibians. The PEP-A locus in Pleurodeles waltl was the first sex-linked gene detected (Ferrier et al., 1983). This locus was also demonstrated to be linked with sex-determining genes in the closely related P. poireti (Dournon et al., 1984). The LDH-B locus was identified as sex-linked in hybrids between female Rana clamitans and male R. catesbeiana (Elinson, 1981), as was the ACON-1 locus in Rana clamitans (Elinson, 1983). In Rana pipiens, the PEP-C and SOD-1 loci were demonstrated to be linked with sex-determining genes (Wright and Richards, 1983). These loci were also found to be linked with sex-determining genes in hybrids between female R. pipiens and male R. palustris, and the SOD-1 locus was linked with sex-determining genes in hybrids between female Rana sphenocephala and male R. blairi (Wright et al., 1983). In hybrids between female R. sphenocephala and male R. berlandieri, in contrast, the PEP-C and SOD-1 loci were not linked with sex-determining genes, although the ADH-2 locus was found to be linked with the latter. All the loci clearly linked with the ADH-2 locus, including the Alb, PGM-1, F16DP and β-GLU loci, were also linked with sex-determining genes in hybrids between female Rana sphenocephala and male R. berlandieri, and the ACON-1 locus was linked with sex-determining genes in R. sphenocephala (Wright et al., 1983). The mME locus was reported to be sex-linked in Xenopus laevis (Graf, 1989). In the Okayama race of R. porosa brevipoda, sex-determining genes were found to be linked with the MPI, LDH-B and PEP-B loci on chromosome No. 4, whereas they were found to be linked with the ME-1 locus on chromosome No. 3 in the Nagoya race of the same species (Nishioka and Sumida, 1994). The MPI, SORDH, ENO and HK loci on chromosome No. 4 were demonstrated to be linked with sex-determining genes in R. nigromaculata (Nishioka and Sumida, 1994). In R. japonica, the Alb locus was linked with sex-determining genes in the seven western populations, whereas the MPI locus was sex-linked in the two eastern populations. In the northwestern Akita population of R. japonica, none of the Alb, AAT-1, ADA, α-GDH, LDH-B, ME-1, ME-2 and MPI loci were linked with sex-determining genes (Sumida and Nishioka, 1994). In the Rana esculenta group, the MPI, LDH-B, HK-1 and PEP-B loci were reported to be sex-linked (Hotz et al., 1997). The present study demonstrated that the AAT-1 locus is sex-linked in the Hiroshima population of B. buergeri, which has ZZ/ZW-type sex chromosomes. The sex-linked AAT-1 locus has so far not been reported in any of the species of Pleurodeles, Xenopus, or Rana, listed above. Apparently the linkage relationships of sex-determining genes to other loci are evolutionarily unstable and there is no common ancestral or conserved sex-linkage group in amphibians.

Ohta (1986) found that chiasmata were clustered primarily near one end in lampbrush bivalent chromosome No. VII of B. buergeri, whereas they were almost evenly distributed along the total length in other bivalents. In the Hiroshima population of B. buergeri, lampbrush bivalent No. VII had a chiasma frequency of 0.43, whereas the other seven small bivalents had chiasma frequencies of 2.10∼3.10. The homo-logues of each of the bivalents No. VII that lacked a chiasma were joined by one or two terminal fusions or at the small landmark situated at the site of 90.4% from one end of the bivalent. Some of the bivalents having only one chiasma were also joined by one or two terminal fusions or at the landmark. In any case, no chiasmata were observed within the range between the tip and about the middle of each bivalent No. VII (Ohta, 1986).

The present study revealed that of 313 1-nu females derived from four matings, almost all were hemizygous, b or c, at the AAT-1 locus, but there was only one 1-nu female heterozygous at the AAT-1 locus. This 1-nu heterozygous female is considered to be a recombinant produced by crossing-over between the sex-determining genes and the AAT-1 locus at a rate of 0.15%. This result suggests that the AAT-1 locus is situated near the nucleolar organizer on the Z chromosome and that this locus and the nucleolar organizer are recombined at an extremely low frequency, which agrees with the low chiasma frequency of lampbrush bivalent No. VII observed by Ohta (1986). In contrast, codominant AAT-1 alleles of heterozygous females are expressed both on the Z chromosome and on the W chromosome in the Kiriake population of Aomori Prefecture (Atsumi et al., unpublished). Buergeria buergeri from this population also has sex chromosomes of the ZZ/ZW type and the Z chromosome has both a secondary constriction and a satellite, while the W chromosome has neither (Ohta, unpublished). These preliminary data probably suggest that the AAT-1 locus is situated not in the side of the satellite but of long arm.

The reasons why the AAT-1 allele on the W chromosome is not expressed are considered as follows: (1) the W chromosome is entirely heterochromatic and extensive genetic deterioration occurs to the heterochromatic region, (2) the AAT-1 gene is lost because of deletion, or (3) the expression of AAT-1 genes is inactivated by factors such as interaction with other genes. When triploids were produced from fertilized eggs of B. buergeri from the Hiroshima population by suppressing the extrusion of the second polar body, there were nearly an equal number of males and females among the ZZW triploids, whereas all ZWW triploids were females and all ZZZ triploids were males (Ohta, 1988). This suggests that the sex-determining genes on the W chromosome are probably functioning, and thus it is difficult to accept the first reason. Further examination at the gene level will be necessary to answer the question.

Dosage compensation of sex-linked genes is accomplished by random inactivation of one of the two X chromosomes in the homogametic sex in female mammals (Lyon, 1961; Ohno, 1967). According to Cayrol et al. (1983), Schmid et al. (1991) and Solari (1994), there are no indications of dosage compensation mechanisms in the homogametic (XX or ZZ) sex of amphibian species with recognizable, heteromorphic sex chromosomes as in birds, as suggested by Baverstock et al. (1982). It is also necessary to examine quantitatively the difference between males and females in the activity of the AAT-1 gene and to verify the dosage compensation of AAT-1 genes in Buergeria buergeri. This species will serve as an excellent example for examining dosage compensation at the AAT-1 locus in the homogametic sex.