Mapping the H-Y gene

The male-specific transplantation antigen, H-Y, is controlled by a gene located on the Y chromosome in both humans and mice. H-Y is a member of a family of minor histocompatibility (H) antigens, each characterized by their ability to stimulate certain immune responses of T lymphocytes (Loveland & Simpson, 1986). At one time, the examination of H-Y expression was limited to grafting experiments but since the advent of methods for generating specific cytotoxic and proliferative T cell responses in vitro and of maintaining these as cloned lines following the introduction of T cell growth factors, H-Y expression can be tested in vitro as well (Simpson, McLaren, Chandler & Tomonari, 1984; Simpson et al. 1987). This approach has been particularly useful for examining the H-Y phenotype of individuals from outbred populations who are not so amenable to the grafting approach. One constraint on such in vitro testing with H-Y-specific T cells is the need to identify the major histocompatibility complex (MHC: HLA in man, H-2 in mouse) alleles of the individual to be typed, since the recognition of H-Y, like other minor H antigens, is MHC restricted (Simpson & Gordon, 1977). T cells recognize H-Y only when it is associated with a particular self-MHC allele, so an appropriate panel of H-Y-specific T cells in necessary to H-Y type individuals of different MHC allotypes.

H-Y typing of mice is simpler than that of man, because of the ease of preparing H-Y-specific T cells restricted by all of the common H-2 haplotypes, using inbred mouse strains (Simpson, 1982). Female mice of inbred strains of appropriate H-2 type can be selected for immunization with H-2 compatible male cells and from these either in vitro bulk cultures of cytotoxic T cells or T cells cloned from these can be prepared for H-Y phenotyping the mice of interest. Examples of the MHC restriction and H-Y specificity of cytotoxic T cells from mixed lymphocyte cultures (MLC) of C57BL/10 (H-2^b) and C57BL/10 × CBA)F₁ (H-2^b/^k) females immunized with (C57BL/ 10(H-2^b) and CBA (H-2^k) male cells, respectively, are given in Table 1 (Simpson, 1982). Table 2 shows the MHC restriction and H-Y specificity of proliferative T cell clones isolated from similar MLC using spleen cells from C57BL/6 (H-2^b) and C3H (H-2^k) female mice immunized with syngeneic male cells (Simpson, 1985). H-Y-specific cytotoxic T cells and clones were used to type cells from a panel of mice carrying the sex-reversing mutation Sxr (Table 3). These include XXSxr males and T16HXSxr females carrying the T16H, X-autosome translocation, which is invariably active, so that the XSxr of paternal origin is inactive. This permits the female development of these individuals, since Sxr is presumably inactive in the majority of cells, at least during gonadogenesis (McLaren & Monk, 1982). The results in Table 3 indicate that each of the XXSxr and XY males were H-Y positive with the cytotoxic T cells and T cell clones appropriate for their H-2 haplotype. These mice are from a noninbred colony in which H-2^k and H-2^b are segregating. Each of the XX females is H-Y negative, whilst of the nine T16HXS.vr females, eight are clearly H-Y positive, indicating that in adult life, at least, the gene controlling expression of the H-Y antigen, Hya, on Sxr is expressed in some spleen cells. The ninth mouse, number 39, was phenotypically H-Y negative: she was subsequently progeny tested (Tl6HSxr females, unlike XXSxr males, are fertile) and since all of the non-XY progeny inheriting her Sxr were H-Y negative, it was clear that a mutation had altered her Sxr fragment. This variant is now designated Sxr¹ (McLaren et al. 1984). XOSxr’ male mice are also H-Y negative when tested by T cells in vitro so that XXSxr’ and T16HX5AT’ mice are not H-Y negative merely because Sxr’ in them is inactivated (Simpson, 1986). XXSxr’ and T16HXSxr’ mice are also H-Y negative when tested for its presence by transplantation, arguing for the identity of H-Y detected by these two methods, one in vitro and one in vivo, and for the absence of H-Y antigen from all cells in the body (Simpson et al. 1986). Sxr’ has lost Hya or the ability to express this gene, but still causes sex reversal in XXSxr’ males, therefore the Y-chromo-some-associated testis-determining gene Tdy on Sxr is clearly separated from Hya by this mutation, although the two genes are closely linked on Sxr and therefore presumably on the portion of the normal Y chromosome, close to the centromere, where Tdy and Hya are normally located (Simpson, 1986). Another mutation which provides evidence for the linkage of Tdy and Hya is Y* described by Eicher & Washburn (1986). Y* is apparently a rearranged Y chromosome in which the pairing region is located close to the centromere: amongst the sperm generated by carrier males is an X^Y, bearing a paternal X to which the greater part of the Y is attached. The XX^Y mice created by the fertilization of a normal X-bearing ovum with such a sperm are H-Y positive and phenotypically male, with aspermatogenic testes (like XXSxr:Simpson et al. 1983).

The investigation of the position of the human H-Y gene on the Y chromosome has produced findings which are in parallel with those of mice, since they clearly separate the testis-determining factor, TDF, from the H-Y gene, but in man the linkage between these two genes, unlike mouse, is not at all close (Simpson et al. 1987).

H-Y typing in man is possible because of the isolation of T cell clones specific for H-Y from transfused spontaneously recovered female aplastic anaemia patients (Goulmy, 1985). Clones currently available are either HLA-A2 or HLA-B7 restricted, so this limits our ability to type cells from individuals carrying one or both of these alleles; fortunately, this includes more than 50% of the population. For the localization of the H-Y gene in man, potentially informative patients are those who have inherited a partly deleted paternal Y chromosome or a translocated Y chromosome fragment. Such patients are in two phenotypic categories: XX males and XY females. The six males described here have inherited variable portions of Yp whilst the two females possess Yq and a variable portion of the Yp. Table 4 shows the results of HLA and H-Y typing lymphoblastoid B cell lines from these patients and appropriate A2- and B7-positive normal male and female controls, with cytotoxic T cells. It is important to confirm serological HLA typing with T cells, since variants of A2 and B7 exist which are not distinguishable serologically but which cannot be recognized by allospecific or MHC-restricted T cells (Horai, von der Poel & Goulmy, 1982). A negative H-Y typing can thus only be interpreted as such in face of a positive allotyping for the restriction element with T cells (A2 or B7 in the case of individuals shown in Table 4).

The deletion map shown in Fig. 1 is based on Vergnaud et al. (1986), Disteche et al. (1986) and Page (1986) and includes the summarized H-Y results of Table 4 as well as unpublished data on class 1 XX males. Since six class 3 males were H-Y negative it is clear that the gene for H-Y does not map to deletion interval 1-3 on Yp (TDF is in interval 1, see also Affara et al. 1986). Likewise the gene for H-Y is excluded from interval 4A, since the class 2 XY female is H-Y positive and lacks this portion of Yp. The H-Y gene thus maps between intervals 4B and 7, far from TDF in interval 1.

In summary, these data, using cytotoxic and proliferative T cell clones specific for H-Y and restricted by MHC molecules to type mice and humans inheriting incomplete portions of the Y chromosome, have allowed us to map the H-Y antigen gene Hya in mouse to a position closely linked with, but separable from, Tdy on the Sxr fragment and thus presumably to a portion of the normal mouse Y chromosome near the centromere. The human H-Y gene maps between deletion intervals 4B and 7, separate from TDFwhich is on interval 1. We are currently testing cells from a number of additional patients who have inherited different portions of the Y chromosome to pinpoint the mapping more closely. It is of interest that in mouse a Y-linked gene, Spy, controlling spermatogenesis maps near Hya (Burgoyne, Levy & McLaren, 1986; for discussion see Burgoyne, this symposium) on the Sxr fragment: they could be the same or closely linked genes. In man, a gene controlling spermatogenesis maps to Yq (Tieopolo & Zuffardi, 1976), and the data so far do not exclude the possibility that it could be coincident with the H-Y gene.