Genes whose mutant phenotype depends on the number of X-chromosomes may be involved in regulating or interpreting X-chromosome expression. Four genes, the autosomal
dpy-26 and
dpy-21 and the X-linked
dpy-22 and
dpy-23 have such X-dependent mutant phenotypes (Hodgkin, MGG, in press). The autosomal mutants are non-Dpy when the X:A ratio is below about 0.67, Dpy when the X:A ratio is around 1, and inviable when the X:A ratio is above about 1.3 ( Meneely & Wood, Genetics, in press). The parallel between these latter two phenotypes and the phenotypes of X hyperploids suggests that in
dpy-21 and
dpy-26 mutants the level of X expression may be inappropriately high. In contrast, the phenotypes of the X-linked
dpy-22 and
dpy-23 mutants are that 2X animals are sick Dpys and 1X animals are inviable except for occasional
dpy-22 XO escapers which are very Dpy. We postulate that the Dpy phenotype in 2X animals results from inappropriately low X-expression, and that 1X animals are inviable for the same reason. Although a definitive test of these hypotheses will require molecular assays of X-chromosome transcription, we have obtained genetic evidence that supports them, by examining the phenotypes of X- linked hypomorphic mutations in the presence of
dpy-21 and
dpy-22 mutations. By definition, a hypomorph is a mutation whose defective phenotype becomes less defective with increasing dose of the mutant allele; that is, if m is a hypomorphic mutation, the severity of phenotype will vary in the manner m/DF > m/m > m/m/m. We have used the three X-linked hypomorphs
unc-3(
e54), 14), and
lin-15(
n767).In 2X animals homozygous for
dpy-21(
e428) and any one of these mutations, the hypomorph is suppressed; that is, its phenotype becomes less severe. In 1X
dpy-21(
e428) animals carrying any one of the hypomorphs, the hypomorph is also suppressed. Other
dpy-21 alleles as well as
dpy-26 alleles are being tested. Since the hypomorphic mutations themselves do not appear to have a common physiological basis, this finding suggests that the
dpy-21 mutation suppresses them by increasing overall expression of the X in both 1X and 2X animals. Consistent with this view is that
dpy-21(
e428) does not suppress an apparent null (amorphic) allele of
unc-3, nor does dpy- 21
(e428) suppress any of eight autosomal hypomorphs so far tested. Conversely,
dpy-22(
e652, the only known mutant allele) increases the severity of all three X-linked hypomorphs in 2X animals, suggesting that it decreases overall expression of the X. Furthermore,
dpy-22 67) XO escapers have the phenotype of strong
lin-15 alleles in XO animals: they are Dpy males with multiple vulvae. The
dpy-22 mutation does not affect autosomal hypomorphs so far tested. To examine the interaction of
dpy-21 and
dpy-22, we have constructed the double mutant
dpy-21(
e428);
dpy22(e652). Its phenotype is similar to that of the single mutant
dpy-22(
e652): XX animals are sick Dpy hermaphrodites, and XO animals are inviable. Taken together, our results suggest that
dpy-21+ is a negative regulator and
dpy-22+ is a positive regulator of X-expression in both XO and XX animals. The finding that
dpy-22(
e652) is apparently epistatic to
dpy-21(
e428) in the double mutant suggests further that
dpy-21+ may act by negatively regulating
dpy-22. We have preliminary evidence that the one known
dpy-23 mutation does not affect the three X-linked hypomorphs. We would like to have a larger collection of both X-linked and autosomal hypomorphs, and will welcome any contributions.