In C. elegans, sex determination and dosage compensation are coordinately controlled by the X:A ratio. Our lab has identified at least four dose-sensitive regions of X that contain numerator (X chromosome) elements (1, 2), indicating that the signal is multigenic. Duplications of one region alone have no effect on XO animals. Multiple copies of one region, or duplications of two or more regions, increase the perceived X:A ratio and cause XO-specific lethality, indicating that numerator elements act additively. These XO animals die because they inappropriately implement the hermaphrodite mode of dosage compensation. Conversely, deficiencies of numerator elements decrease the perceived X:A ratio in XX hermaphrodites. These deficiency XX animals die or are dumpy and masculinized because they cannot implement the hermaphrodite mode of dosage compensation. Initial suppressor screens of the XO-specific lethality caused by duplicating regions 1, 2 and 3 identified
y303 and
y304 as X chromosome signal element mutations. These alleles both map to region 3, near
fox-1.
fox-1 was hypothesized by Hodgkin et al to be a locus contained in a cosmid that kills XO animals when present in multicopy arrays (3). A cDNA encoding a putative RNA binding protein was isolated with this cosmid. While Hodgkin et al did not attribute the feminizing activity of the cosmid to the cDNA, our analysis of
y303 and
y304 offers genetic proof that the gene encoding the RNA binding protein is a numerator element. Sequence analysis reveals that
y304 is a missense mutation at amino acid 161 and
y303 is a nonsense mutation at amino acid 37 of
fox-1. Thus,
y303 is likely to represent a null allele. We find that neither
y303 nor
y304 alone has any discernible phenotype in XX animals, consistent with the proposed multigenic nature of the signal. In addition, we propose that at least one other signal element is contained in region 3. We predict that if
fox-1 is the only numerator element in region 3, the degree of masculinization of XX animals heterozygous for deficiencies of all three regions should be comparable to that of XX animals heterozygous for deficiencies of regions 1 and 2 and a loss of function mutation in
fox-1. However, XX animals with deficiencies of regions 1, 2 and 3 are significantly more masculinized than XX animals heterozygous for deficiencies of regions 1 and 2 and
fox-1(
y303). Continued suppressor analysis is expected to identify additional signal elements (4). The apparent target of the X:A ratio is
xol-1.
xol-1 functions to repress the hermaphrodite modes of sex determination and dosage compensation and is sex-specifically expressed in XO embryos. Overexpression of
xol-1 rescues XO embryos from the lethality of numerator element duplications while a
xol-1 null allele rescues XX animals from the lethality and masculinizing effects caused by numerator element deficiencies. We have investigated the effects that numerator elements within regions 1, 2 and 3 have on the expression of
xol-1. We find that region 1 regulates the transcription of
xol-1 as assayed from a transcriptional
xol-1::lacZ reporter gene. Duplications of region 1 dramatically decrease the level of lacZ expression in XO embryos while deficiencies of region 1 cause a corresponding increase in the level of lacZ expression in XX embryos. However, regions 2 and 3, as well as
fox-1(
y303), do not affect the normal XO-specific expression of the lacZ reporter gene. Numerator elements in regions 2 and 3 may control
xol-1 expression post-transcriptionally. If so, then the primary sex-determining signal must control
xol-1 expression via two distinct mechanisms to ensure the appropriate sexual fate. We will present genetic and molecular data that implicate
xol-1 as a target of
fox-1 and discuss the use of a dual control mechanism in the fidelity of the sexual fate decision in the early embryo. 1. Akerib and Meyer, 1994. Genetics 138:1105-1125. 2. Carmi et al, Abstract. 3. Hodgkin et al, 1994. Development 120: 3681-3689. 4. Liu et al, Abstract.