Chromosomes and sex differentiation in eutherians

In eutherians, the male is the heterogametic (XY) and the female the homogametic (XX) sex; primary sex differentiation (i.e. the onset of gonadal development) is genetically determined and independent of environmental factors. The Y chromosome determines the male sex, and its absence the female sex. One Y chromosome is sufficient for testicular development even in the presence of supernumerary X chromosomes as in human individuals with 47,XXY; 48,XXXY; or 49,XXXXY karyotypes and Klinefelter syndrome. Conversely, one X chromosome is not enough for normal ovarian development in all eutherians. In humans with a 45,XO karyotype and Ullrich-Turner syndrome, Streak gonads occur after premature gonadal regression. In other orders of eutherians, e.g. rodents, one X is compatible with ovarian development and function. This species difference may be ascribed to different X-inactivation patterns. While X-inactivation seems to take place in all female eutherians for dosage compensation, its degree may vary. X-chromosomal genes involved in ovarian differentiation may operate in double dosage and escape inactivation in man but not in some other species.

The examples indicate that genes on both sex chromosomes are necessary for gonadal differentiation in eutherians. Little is known about the nature of these genes and their products. One candidate of a Y-dependent gene product involved in testis differentiation is the male-specific, serologically detectable H-Y antigen. In fact, testicular tissue is only detected in H-Y positive mammals independent of their genetic sex (for review see Müller, 1984). There is evidence of a gonad-specific receptor of H-Y (Müller, Wolf, Siebers & Günther, 1979; Nagai, Ciccarese & Ohno, 1979; Wachtel & Hall, 1979), and experiments using XX gonads or gonadal cells and testis or Daudi cell supernatant as the source of soluble H-Y antigen indicate virilization of the female gonad in vitro (Zenzes, Wolf & Engel, 1978; Nagai et al. 1979; Ciccarese, Orsini, Massari & Guanti, 1983; Müller & Urban, 1981; Benhaim et al. 1982). H-Y antigen, however, is presumably not the only and primary inducer of the mammalian testis. In the following I will therefore discuss approaches towards a more thorough understanding of the role of sex chromosome-dependent proteins in gonadal development.

While over 15 structural genes (and over 115 genetic loci) have been assigned to the human X chromosome (McKusick, 1984), only one structural gene was found on the Y. This gene codes for 12E7, an antigen defined by a monoclonal antibody (for review see Goodfellow, 1983). Other genes that were assigned to the Y by deletion mapping may well be regulatory, e.g. genes for H-Y antigen, testis differentiation, spermatogenesis, stature, tooth size, etc (for review see Bühler, 1980). All sex chromosome-dependent gene products known, apart from H-Y antigen, are apparently not related to gonadal differentiation. In order eventually to detect sex chromosome-dependent gene products involved in gonadogenesis it might be feasible to assign polypeptides/proteins of unknown function to the sex chromosomes and later investigate their possible role in gonadogenesis. One approach for mapping genes of polypeptides of unknown function is the investigation of total cellular proteins from human-rodent somatic cell hybrids by two-dimensional gel electrophoresis. Two-dimensional gel electrophoresis methods (O’Farrell, 1975; Klose, 1975) allow the analysis of large numbers of proteins by the combination of two separation techniques. First the polypeptides are separated according to their isoelectric point (IEP; first dimension) and subsequently according to their relative molecular mass (M_r; second dimension). If somatic cell hybrids containing one of the human sex chromosomes are analysed, it should become possible to locate genes for some polypeptides on these chromosomes. We have applied this approach to a Chinese hamster-human somatic cell hybrid that contained the Y as the only human chromosome. Approximately 500 polypeptides were detected in both the hybrid and its revertant that contained hamster chromosomes only (Fig. 1). Side-by-side comparison of four gels of each hybrid and parental line revealed two spots in the hybrid that were absent or only faintly visible in the hamster cells (Fig. 2). Furthermore they were not detected in normal human male fibroblasts. Evidently the human Y chromosome induced the expression of rodent genes, i.e. it exerted a regulatory effect on the hamster genome. The modifying effect seemed to be specific for the Y. In similar hybrids that contained the X (Müller & Voiculescu, 1984) or chromosome 21 (Scoggin, Paul, Miller & Patterson, 1983) as the only human chromosome an increase in intensity of these spots was not found. The Y-containing hybrid cell did not reveal any Y-chromosomal human structural gene. This is in agreement with data of Taggart & Francke (1982), who also failed to detect Y-chromosomal structural genes in a large panel of human-rodent somatic cell hybrids by two-dimensional gel electrophoresis. On the other hand we (Müller & Voiculescu, 1984) and two other groups (Cox, Francke & Epstein, 1981; Taggart & Francke, 1982) found indications of X-chromosomal structural genes in hybrid cells. The polypeptides found by the three groups are listed in Table 1. Only one putative human X-chromosomal structural gene was comparable in all three investigations. Others were identical only in two investigations or differed completely with respect to M_t and/or isoelectric point. Thus the reproducibility of two-dimensional gel findings in somatic cell hybrids is limited. Several causes may be considered.

1. Different rodent fusion cells may affect gene expression of human polypeptides differently (repression, induction).

2. Discrepancies even in one of the molecular characteristics (M_r and IEP) of a given polypeptide in different investigations preclude the demonstration of whether the polypeptides are polymorphic variants of one protein or not.

3. The human origin of an unknown polypeptide is hard to establish unequivocally.

4. Human polypeptides might be subject to different post-translational modification in different ‘environments’ (rodent cell lines, additional human chromosomes).

Some of these problems may be avoided by analysing genetically identical cell lines that only differ in one sex chromosome. Such cell lines may be obtained from human Turner mosaics of the type XO/XY. We were able to study 45,XO and 46,XY fibroblasts from a monozygotic twin pair by two-dimensional gel electrophoresis (Müller, Maier, Gimelli & Fraccaro, 1984a). Analysis of the polypeptide patterns revealed two, out of approximately 450 polypeptides, that were present in the XY line only. These polypeptides had M_rs and lEP’s of 38,000/6·3 and 30,000/5·4. It cannot be decided, however, whether they are regulated or coded for by the Y chromosome. Interestingly Wachtel (1983) studying epithelial cells from male and female inbred mice also found a male-specific protein of relative molecular mass 30,000.

If some of the above sex chromosome-dependent polypeptides were related to sex differentiation, they should be present in the gonads of one sex. In addition an important function in sex differentiation would suggest evolutionary conservation of such polypeptides. Thus in order to learn more about the role of sex chromosome-dependent polypeptides in gonadogenesis of eutherians, it is necessary to analyse sex-specific polypeptides during gonadal differentiation.

We have investigated the gonadal protein patterns during development in isogenic rats by two-dimensional gel electrophoresis (Müller et al. 1984b). Our investigation started with gonads from 12·5-day-old embryos and was extended to gonads of newborn rats. At embryonic day 12·5 the gonads of both sexes cannot be distinguished either by light or by electron microscopy; this is the morphologically indifferent stage. Therefore the genetic sex of the embryos had to be determined by karyotyping of foetal cells. Applying banding techniques, the sex chromosomes could be easily distinguished. The minute size of rat gonads at the indifferent stage and at early stages of differentiation makes it practically impossible to analyse their protein patterns by standard two-dimensional gel electrophoresis macromethods. In order to circumvent this problem we have applied a two-dimensional microgel electrophoresis method (according to Poehling & Neuhoff, 1980). Using this technique, up to 250 polypeptides could be separated on the gels after application of only 4–6 μg protein.

Polypeptides that were present in gonadal preparations of one sex only and absent in the other were found at all developmental stages (see Table 2). This indicates expression of several specific genes in both gonads throughout gonadal differentiation. At the indifferent stage two polypeptides were detected in the female gonadal anlage and one in the male. Since gonads and mesonephros cannot be separated mechanically at this stage we do not know whether the sexspecific polypeptides are part of the gonad or of the mesonephros. Considering, however, that the mesonephros is involved in gonadal development, one might speculate that they have a function related to sex differentiation. In fact, Taketo & Koide (1981) provided evidence that the morphologically indifferent gonad is already committed, i.e. its further development into ovary or testis is already determined. It may be that the sex-specific proteins detected in our investigation are a biochemical correlate of this commitment.

At embryonic day 13·5, gonads may be distinguished morphologically. Thus the ultrastructural investigations of Jost, Magre & Agelopoulou, (1981) at this developmental stage indicate testicular Sertoli cell differentiation, while the ovaries still appear indifferent. Two sex-specific gene products were found in each of the gonads, male and female. These polypeptides remained expressed until birth and thus could be investigated in more detail. Analysis of germ cell-free gonads after intrauterine busulphan treatment (for details see Müller et al. 1984b) indicated that the two early appearing testicular polypeptides were in the somatic part of the testis. Furthermore, one testicular protein was primarily (or exclusively) found in preparations of the tunica albuginea. The early expression of a tunica-specific protein is consistent with the early formation of a tunica albuginea in the developing testis of various mammalian species. In fact, several authors (Nagai et al. 1979; Moon & Hardy, 1973; Black & Erickson, 1965) use tunica formation as an indicator of the onset of testicular differentiation. Thus our tunica protein may be used as a biochemical marker of the functional start of testicular development. Of the ovarian polypeptides one was in germ cells, the other in somatic cells. Both occurred at a time in the gonad when morphologic differentiation is not yet detected in ordinary sections of ovarian preparations by light microscopy. In alkaline-phosphatase-stained sections, however, signs of differentiation may be detected at the indifferent stage already (at least in bovine gonads, Gropp & Ohno, 1966). Thus cellular rearrangements become visible that are quite different from those observed in the testis. Perhaps the somatic cell-specific ovarian protein is required for these early processes of ovarian differentiation? This notion is consistent with the finding that germ cells are not involved in early ovarian development (Coulombre & Russell, 1954).

Comparison of the gonadal proteins with the sex chromosome-dependent polypeptides revealed similar molecular characteristics (IEP and M_r) of one polypeptide. The testicular tunica protein (TE 2/Table 2) had practically identical relative molecular mass and IEP as the Y-dependent polypeptide detected in XY fibroblasts of the monozygotic twins described above. Although it is tempting to speculate that the polypeptides are identical, further investigations are required to show this. No similarities were found between the remaining gonadal and the sex chromosome-dependent proteins. Considering, however, that there are no hormonal sex differences in early embryos, all early gonadal polypeptides may in some way depend on the sex chromosomes.

The author was supported by the Deutsche Forschungsgemeinschaft (Mu 668/1–1).