We report here the identification and characterization of a new X- linked locus important to both the processes of sex determination and dosage compensation in XX animals. This gene has provisionally been named
sdc-2 (sex determination and dosage compensation), by analogy with the gene
sdc-1, also X-linked, in which lesions have been shown to disrupt both these processes. The original allele of
sdc-2,
y15, was isolated as a recessive X- linked, XX-specific lethal mutation. Sex-specific lethality is commonly associated with mutations that disrupt dosage compensation. Consistent with
sdc-2's involvement in dosage compensation is that 21 more alleles have been isolated as suppressors of the XO-specific lethality of mutations in the gene
xol-1. The XO-specific lethality of
xol-1 mutations seems to result from XO animals adopting the XX mode of dosage compensation (see Miller et al., this WBG). Suppression of
xol-1 lethality by
sdc-2 is likely to be by a mechanism in which the XX mode of dosage compensation can no longer be implemented. As a result, X-linked gene expression returns to a normal male level, rescuing XO lethality, but causing a concurrent loss of viability in
xol-1 ls. Thus,
sdc-2 is likely to be important for the hermaphrodite mode of dosage compensation. All 21 alleles fail to complement the XX-specific lethality of
y15. We have divided them into two classes. Class I (13 alleles) are XX specific lethal mutations. They are also lethal to XX animals in trans to the deficiency nDf19. Class II (9 alleles) cause XX animals to display a phenotype of sexual transformation toward the male fate. Homozygous animals range from Egl hermaphrodites to nearly wild-type looking males, but most are intersexual pseudomales with slightly dumpy hermaphrodite-like bodies and imperfect male tails. Homozygous strains cannot be maintained with the strongest alleles of this class since the Egl hermaphrodites produce broods consisting almost entirely of pseudomales. For weaker alleles, homozygous strains may be maintained. Fertile animals are somewhat short, slightly Dpy, Egl, and have protruding vulvae. Similar Dpy, Egl and protruding vulva phenotypes are associated with the dosage compensation defect of
sdc-1 mutants. Surviving animals bearing class II alleles in trans to nDf19 are pseudomales; we are currently quantifying the extent of lethality associated with this interaction. The strong class II allele
y55 fails to complement the XX-specific lethality of a class I allele,
y39. Rare trans-heterozygote survivors are pseudomale. The division between the two classes of alleles is not strict, and is made primarily for convenience in thinking about groups of similar alleles. Some class I alleles produce small numbers of pseudomales as homozygotes and also in trans to nDf19. There is at least one class I allele that is lethal at 20 C but produces some pseudomales at 25 C. XO
sdc-2 animals in all cases appear to be wild type males, both in the presence and absence of
xol-1 mutations. Alleles of both classes have been isolated by several members of this laboratory from both EMS and gamma-radiation induced mutagenesis, and a class I allele has been isolated from a mutator background. Both class I and class II alleles have been mapped by their phenotype of suppression of the XO-specific lethality of
xol-1(
y9). Three- factor crosses place them approximately midway between
dpy-6 and
unc-9. All alleles fall within the deficiency interval defined by nDf19. The XX specific lethality of the
y15 allele has been mapped by three- factor crosses with
dpy-6 and
lin-14.
y15 lies to the left of, or very close to
lin-14, and is not covered by the duplication stDp2. Interactions between the XX transformer phenotype of class II
sdc-2 alleles and the genes of the major sex determination pathway have been explored. In the strains
her-1(
e1520) 70); ) (or (
y54)), and
her-1(
e1520); ) , the
her-1 mutation blocks the masculinizing effects of
sdc-2 mutations. Thus, the XX transforming effect of class II
sdc-2 mutations acts through the major sex determination pathway and requires the wild type
her-1 gene product for it to be expressed. Moreover,
her-1; ) or
her-1; ) individuals still exhibit the slightly Dpy, Egl and protruding vulva phenotypes associated with the dosage compensation defect. In addition, XX animals of genotype
her-1; ry class I (XX lethal) alleles of
sdc-2 are inviable.
her-1 mutations clearly suppress the sex determination effects, but not the dosage compensation effects, of
sdc-2 mutations.
sdc-2 mutations appear to cause inappropriate activation of the XO modes of dosage compensation and of sex determination in XX animals. As a result XX animals are inviable or masculinized (and perhaps both).
sdc-2 mutations appear to be epistatic to both sex determination and dosage compensation effects of lesions in
xol-1: they suppress the XO specific lethality of
xol-1, and
xol-1 utants produce no XO hermaphrodites. This is in contrast to suppression of
xol-1 mutations by mutations in
dpy-26, e epistatic only to dosage compensation effects of
xol-1; these doubly mutant strains produce wild type XO hermaphrodites. All of this is consistent with the notion that
sdc-2 is essential in XX animals for the proper modes of both dosage compensation and sex determination.
sdc-1 mutations are fully epistatic to
xol-1 mutations in a manner similar to
sdc-2, although
sdc-1 acts as a maternal-effect suppressor while
sdc-2 is a simple recessive suppressor (see Miller et al., this WBG). The effects of
sdc-2 mutations on dosage compensation are being further investigated by both genetic and molecular methods. Interactions between
sdc-2 mutations and mutations in the genes
sdc-1,
dpy-26, examined. In addition, X-linked gene expression will be assayed by Northern blot analysis. While the two classes of mutations discussed here have not rigorously been shown to occur at the same locus, complementation tests and initial mapping experiments strongly suggest that they disrupt the same function. Reversion analysis and careful mapping will help to clarify this issue. The formal possibility exists that the two classes of alleles comprise two closely linked loci that exhibit interactions in trans, but this can only truly be resolved at a molecular level. Further experiments will establish whether the null phenotype of
sdc-2 is indeed XX-specific lethality.