As described previously (C. elegans CSH Meeting Abstracts, 1987, p . 187), mutations in two genes ,
egl-15 and
egl-17, cause defective sex myoblast (SM) migrations in the hermaphrodite as a result of the repulsion of the SMs by somatic cells in the gonad. The migration of the SMs is sexually dimorphic: hermaphrodite SMs migrate anteriorly, while male SMs migrate posteriorly. The
egl-15 and
egl-17 mutations appear not to affect the male SM migrations. One possibility was that the altered SM migration is a consequence of a confused sexual identity of the SM-gonad interaction. In support of this hypothesis, we have found that the masculinizing, dominant allele of the
her-1 gene,
n695, can cause the SMs to be located in positions similar to those found in
egl-15 and
egl-17 mutants. The observed repulsion in
egl-15 and
egl-17 mutants could be a subtle masculinization of the direction of the SM migration if the normal male gonad propels the SMs posteriorly. However, laser ablation of the gonad primordium in N2 males does not obviously alter the migration of the SMs. The possibility remains that repulsion of the SMs by the gonad in males is a secondary mechanism involved in the correct positioning of these cells and their descendants. Precedent for this type of backup mechanism for positioning the SMs is found in the hermaphrodite (Thomas and Horvitz, WBG 9(2):59-60). We have taken two approaches to determine if the effects of
egl-15 and
egl-17 mutations are due to sexual transformation. First, we have tested these mutations for interactions with sex determination genes. Second, we have sought to isolate new alleles by non-complementation screens to see if they more broadly affect sex determination. We suspected that the null phenotypes of
egl-15 and
egl-17 might not be merely Egl, since Egl mutations in these two genes arise at frequencies at least five times lower than the typical null frequency. One possibility was that the alleles isolated to date specifically affect the sexually dimorphic migration of the SMs, while null alleles might more generally affect sex determination. The weak
tra-2 allele
n1106 causes hermaphrodites to be Egl as a result of the male-like programmed cell death of the HSN neurons that drive egg-laying (Desai and Horvitz, C. elegans CSH Meeting Abstracts, 1985, p. 112). These animals can be more fully masculinized in combination with other mutations that cause subtle masculinization ( Miller and Meyer, C. elegans CSH Meeting Abstracts, 1987, p. 61). We have found that
tra-2 (
n1106);
egl-17(
e1313) animals are infertile pseudomales while
tra-2(
n1106); 84) animals portray no detectable interaction. Similarly,
sdc-1(
n485) shows pronounced masculinization by
egl-17(
e1313) but no strong interaction with
egl-15(
n484) . Thus, it seemed likely that the altered SM migration in egl- 17
(e1313) animals could be due to partial sexual transformation. However, while a recessive allele of
her-1,
e1520, prevents the masculinization of
tra-2(
n1106) by
egl-17(
e1313) (Chad Nusbaum, personal communication), it does not relieve the altered SM positioning. This result suggests that
egl-17 interacts with the sex determination gene
tra-2 in a
her-1-dependent manner, but that the altered SM positioning in
egl-17(
e1313) animals is independent of, or at least requires a much lower level of, the
her-1(+) gene product. One possible model is that the altered SM migration in both
egl-15 and
egl-17 mutants is not the result of partial sexual transformation of the SM-gonad interaction, and that
egl-17 has a second, independent function in determining sexual identity. To determine if we could expect to isolate null alleles by noncomplementation screens using the existing alleles of
egl-15 and
egl-17, we checked the phenotypes of eT2(I)/+;
egl-17(
e1313)/eT2(X) and
egl-15(
n484)/nDf19 (eT2 is missing
egl-17 and the left tip of LGX, and nDf19 deletes
egl-15). Both are Egl, although +/eT2(I);
egl-17(
e1313)/eT2(X) animals are unhealthy. We screened 36,000 EMS- mutagenized haploid genomes for mutations that failed to complement
egl-17(
e1313) and found four new alleles that are similar to the two previously known ones. The low frequency of these alleles suggests that they are not null alleles and that
egl-17(
e1313)/egl-17(null) animals may be too sick to isolate easily. We have also screened 32,000 EMS-mutagenized haploid genomes in a non-complementation screen for
egl-15 alleles, yielding eleven new mutations that fail to complement
egl-15(
n484). Three are similar to
egl-15(
n484), while the other eight are associated with a linked larval arrest phenotype with no obvious cause of the arrest. One of these,
n1477 is Egl and scrawny at 20 C and arrested at 25 C.
n1477 fails to complement all of the other arrested alleles at the restrictive temperature. Another allele,
n1454, maps to
egl-15 ( between
dpy-6 and
lin-14, 6% right of
dpy-6). These data, taken together, suggest that the null phenotype of
egl-15 is larval arrest. Results from these non-complementation screens suggest the involvement of these genes in other functions, although not specifically in sex determination. We plan to study the new
egl-15 and
egl-17 alleles further in the hope of better understanding the roles of these genes.