We are screening for maternal effect lethal mutations that may identify genes important for determination or differentiation of intestinal cells in C. elegans embryogenesis. Specifically, we are looking for mutants in which the arrested embryos possess more than a hundred cells, but show no detectable gut granules under polarized light. (WBG 9[2] 74, 1986; C. elegans meeting abstracts 1987, 3). We have identified 36 such mutants among recent screens of approximately 12,000 F1s from EMS mutagenized
egl-23(
n601) 47) animals. We find that about 15% of the F1s carry a maternal effect lethal mutation; among these Mel mutants, approximately 2% produce embryos lacking gut granules. Our previous work with par mutants suggested that failure to produce gut granules during embryogenesis could result from mutation in the genes
par-1 and
par-4. In fact, among the 36 new gut granule mutations, we have identified 5
par-1 alleles and 7
par-4 alleles. The remaining mutants do not appear to have partitioning defects in the early cleavages. Apparently,
par-1 and
par-4 are the only par genes that can mutate to give a complete failure of intestinal cell differentiation. Most of the remaining mutants produce embryos with about two hundred cells, without obvious cell differentiation. Such mutants may not be gut- specific; other cell types that would normally differentiate later in development may also be affected. The most interesting mutants for our analysis are those in which the embryos contain many other differentiated cells, but no detectable gut differentiation. In these mutants a lack of gut granules could be due to a defect specific to the intestinal cell line. We have found two kinds of such interesting mutants -- one mutant,
it77 X, produces embryos that undergo morphogenesis to produce abnormal-looking larvae that lack gut granules. We have not yet examined this mutant carefully for specific anatomical defects. Four other mutants produce embryos with a terminal phenotype similar to
par-1 and
par-4; the embryos arrest as amorphous masses of many differentiated cells, but no gut granules are detectable. These mutants show a normal early cleavage pattern, and p granule staining of one of them showed a normal distribution. We do not yet know whether the mutations in this class represent one or several loci. (Perhaps these mutations identify a gene(s) that encodes a cytoplasmic factor that must be partitioned by the maternal partitioning system to the E line in the first cleavages???) While performing these screens, we also noticed a class of maternal effect lethal mutants in which only a few of the embryos had gut granules, and all of the embryos arrested as masses of differentiated cells. Because this phenotype is characteristic of known
par-2 mutants, we also saved this type of mutant, and found 8 that have a
par-2-like early cleavage pattern. Five of these thus far have been confirmed as new alleles of
par-2 (Niansheng Cheng, personal comm.). We also isolated two other mutants with par phenotypes (synchronous and symmetric early cleavages) that are not
par-2-like, and which we have not yet assigned to complementation groups. One of these mutants has an early cleavage pattern similar to
par-1 [some
par-1 mutants produce a low frequency of embryos with gut granules (Betty Suh, and our own observations)]. The other mutant does not map to LG IV,V or X, which rules out
pars-1,4,5. This mutation may represent an unusual allele of
par-1 or
par-2, or could identify another gene that can mutate to give a par-like phenotype. Screens for maternal effect lethal mutants with a characteristic terminal embryonic phenotype appear to be efficient for identifying mutations in par genes. By this method we have more than doubled the number of
par-1, s. The frequency with which we have found these new Dar mutations suggests that they are typical for these genes, and are likely to represent loss of function or reduced activity of the gene product. We did not identify mutations in the genes
par-3 or
par-5 in our screens. Both of these mutants produce a greater proportion of embryos containing gut granules than the other par mutants and so would not be picked up in these screens. However, screens for mutants with terminal embryonic phenotypes characteristic of either
par-3 or
par-5 should be feasible. In fact, we have isolated one new
par-3 mutation in this way.