In the last Newsletter we described the phenotypes and map positions of some new dominant mutations. To explore the nature of genes that can mutate to dominant alleles, we have begun to attempt to determine the null phenotypes of a set of such genes. Since dominance often reflects the synthesis of a gene product altered in function or increased in level (e.g., see Greenwald and Horvitz, Genetics, 96, 147, 1980; also, 110, in press), the introduction of loss of function ( 'null') mutations should generate alleles that are recessive. Thus, the intragenic reversion of dominant alleles offers one approach to the isolation of null mutations. We have attempted to obtain intragenic revertants of mutants in seven genes with dominant alleles. In general, F1 and F2 progeny of EMS-mutagenized homozygous hermaphrodites were screened for revertants ('homozygous reversion'). If this protocol failed to produce tightly- linked (presumably intragenic) revertants, as it must for certain genes (e.g., those with null phenotypes that are either lethal or indistinguishable from the dominant phenotypes), N2 males were mated with EMS-mutagenized homozygous hermaphrodites and wild-type F1 progeny sought ('heterozygous reversion'). Only tightly-linked revertants (less than 0.1% from the original mutation), which we have tentatively assumed to be intragenic, are described below. The following table summarizes the number of revertants and the homozygous phenotypes of the putative intragenic revertants obtained from these protocols. [See Figure 1] Mutants in unc-105, to homozygous viable phenotypically wild-type animals. Mutants in unc( n498) and unc(n494) 'reverted' from one Unc phenotype to another: n498 animals are almost paralyzed, where as n498 revertants are only mildly uncoordinated; n494 animals are very sluggish and assume coiled postures, whereas n494 revertants are kinky and relatively active. unc(n493) reverted only as a heterozygote, and the revertant heterozygotes segregated larvae that failed to develop. All of these revertants are recessive, and all appeared at the frequency expected after 4 EMS-mutagenesis for null mutations in a typical C. elegans gene (10+E-3 to 10+E-4); in addition, spontaneous intragenic revertants, which are likely to reveal null phenotypes (Greenwald and Horvitz, 1980), have been obtained for unc-105, n498). Thus, it is likely that these revertants display the null phenotypes of these genes (although for unc(n498) and unc(n494) it remains possible that the null phenotype is wild-type, and the original strains contain tightly-linked recessive unc mutations). That rol-6 and unc-8 have wild-type null phenotypes is further supported by the observation that wild-type intragenic revertants failed to complement visible recessive alleles (e187 and e49, respectively) of these loci. We have been unable to isolate revertants of n501 by either protocol, suggesting that n501 either itself reduces gene activity or defines a locus that is relatively difficult to mutate with EMS.