dpy-21 appears to be required in both XX and XO animals for proper dosage compensation. In XX animals, mutations in
dpy-21 result in a Dpy phenotype and elevated X-linked gene expression. XO
dpy-21 animals appear wild-type in length, but they also have altered X- linked gene expression. To date
dpy-21 is the only XX specific dpy mutation that appears to affect both XX and XO animals. We have begun a systematic genetic analysis of
dpy-21. The mutant phenotype of
dpy-21 was originally defined by the alleles
e428 and
e459 (J. Hodgkin, Mol. Gen. Genet. 192: 452 (1983). 12 new EMS induced alleles have been isolated as suppressors of the XO-specific lethality of
xol-1(
y9). Three of these alleles (
y47am,
y59am and
y60am) are suppressible by
sup-7(
st5) X. All alleles isolated to date resemble the original alleles: XX animals are Dpy, frequently have a protruding vulva (rare animals are Bivulva) and occasionally are Egl. XO animals appear wild-type. In five of six alleles tested for viability at 20 C the progeny of homozygous mothers have reduced viability:
e428 (83%),
y47am (86%),
y58 (96%),
y60am (89%),
y87 (93%) ( percent of zygotes that matured to adulthood). The exception is
y88ts, which is slightly Dpy at 15 C, Dpy and fully viable at 20 C (99.8%), and Dpy with reduced viability at 25 C (90%). Despite the isolation of amber suppressible alleles, it is possible that null alleles of dpy- 21 might not have been recovered from the
xol-1 suppression screen. We have devised a complementation screen that has allowed the isolation of deficiencies of
dpy-21 and we are now using this same screen to isolate additional EMS induced alleles of
dpy-21.Using the complementation screen we have isolated 3 gamma ray induced mutations ( yDf4, yDf6 and yDf7) that fail to complement both
dpy-21 and
par-4(
it57), which lies to the right of
dpy-21. Since all three deficiencies complement mutations to the left of
dpy-21, we must use molecular markers to show that the deficiencies do indeed span the dpy- 21 locus. The 5S rRNA contig, which maps just to the left of
dpy-21, will provide the initial probes for mapping the leftmost deficiency endpoints.
dpy-21(
y88ts)/yDf6 hermaphrodites are Dpy and 70% viable ( compared to 99.8% viability for
y88/y88). In contrast, the viability of
dpy-21(
y88ts)/yDf4 and
dpy-21(
y88ts)/yDf7 hermaphrodites is reduced to 14% and 20%, respectively. The severely reduced viability in these two stains may be explained in part by the additional mutant phenotypes associated with these two deficiencies. yDf4/unc-76 and yDf7/unc-76 strains produce 35% and 41% dead eggs respectively and only 17% of zygotes mature to become Unc-76 hermaphrodites. Since both deficiencies have been multiply backcrossed these results may indicate that yDf4 and yDf7 have some dominant maternal-effect lethality. In addition, Df/unc-76 hermaphrodites are slow growing ( evidence of haploinsufficiency). Our working hypothesis is that all three deficiencies span the
dpy-21 locus and that the dominant effects of yDf4 and yDf7 are due to removal of some other locus or loci. To determine if the available
dpy-21 alleles are null we compared the viability of
y88ts/yDf6 (70%) to that of
y88ts/y60am (94%). In this situation
y60am does not act like a null allele, in that it does not resemble a deficiency when placed in trans to a hypomorphic allele. Two possible interpretations of these data are: 1) we have not yet isolated a null allele of
dpy-21 or 2) yDf6 removes a locus that acts synergistically with
dpy-21. This question will be settled once we know the nature of mutations obtainable from the complementation screen.