Mutations that disrupt dosage compensation have a paradoxical feminizing effect in some masculinized XX animals. The
sdc-3 gene ( formerly
dpy-29), one of several genes that controls the hermaphrodite modes of sex determination and dosage compensation, provides a striking example of this phenomenon operating in a single gene. This phenomenon was revealed by the fact that even though
sdc-3(null) mutations disrupt both processes, the
sdc-3(null) XX animals are not masculinized. In contrast,
sdc-3 mutations that disrupt only the sex- determination function but not the dosage compensation function result in fully masculinized XX animals. Analysis of
sdc-3 has provided insights into the effects of dosagecompensation disruptions on sex determination.
sdc-3 resembles other sdc genes in that it acts at the same point of the regulatory hierarchy. It is negatively regulated by xol-l and in turn negatively regulates her-l .
sdc-3 differs from the other sdc genes in that its sex-deterrnination and dosage-compensation functions are independently mutable, resulting in three classes of alleles: those that affect only the sexdetermination function (
sdc-3(Tra)), only the dosage-compensation function (
sdc-3(Dpy)), or both functions (
sdc-3(null)). The
sdc-3(Dpy) and
sdc3(null) XX mutants are phenotypically indistinguishable (i.e. hermaphrodites). (DeLong, Plenefisch and Meyer, unpublished) Mutations that eliminate the
sdc-3 dosage-compensation function are not the only alterations that suppresses the masculinizing effect of
sdc-3(Tra) alleles. Mutations in the dosage compensation genes
dpy-26,
dpy-27 and
dpy-28, as well as increases in X-chromosome dose also suppress this masculinization. Our results suggest that suppression acts via regulation of her-l or the sdc genes. Suppression in dpy; sdc- 3(Tra) mutant animals is correlated with reduction in the elevated her- l expression caused by the
sdc-3(Tra) mutations.(see abstract by Klein and Meyer). Suppression is independent of xol-l function, suggesting that it cannot be attributed entirely to changes in the primary sex- determining signal. Additional data are most consistent with suppression resulting from elevated X-linked gene expression, suggesting that suppression results from increased expression of either X-linked negative regulators of her-l or Xlinked positive regulators of the sdc genes. These results suggest that assessing the level of X-linked gene expression might be part of the normal sex-determining process. Altemately, they might indicate that overexpression of genes not normally involved in sex determination have an effect on the process. Several models will be discussed.