mec-8 mutants are mechanosensory defective (Mec) (Chalfie and Au, 1989); display dye-filling defects (Dyf) of the amphid and phasmid sensory neurons, owing to improper bundling of the ciliated tips of the neuronal processes (Perkins et al., 1986); and show a variable, cold-sensitive lethality. Arrested
mec-8 animals have disorganized body wall muscle. Furthermore,
mec-8 ;
unc-52 (viable)double mutants exhibit a nonconditional paralyzed arrest at the twofold stage of embryogenesis (Pat), which is characteristic of severe body wall muscle defects including
unc-52 (lethal)mutants(B. Williams and R. Waterston, pers.comm.).
mec-8 was mapped genetically with respect to markers on both the physical and genetic maps. Overlapping cosmids were identified that detected dimorphisms associated with four
mec-8 alleles (2 gamma-ray-induced, 1 EMS-induced, and 1 RW7097 -induced).A 2.5-kb HindIII fragment, contained in cosmid C34A2 ,was affected in all four alleles. An 8.5-1kb Xhol fragment, containing the 2.5-kb HindIII fragment, was subcloned from C34A2 and used in germ line microinjection experiments. Transformation of
mec-8 (
u74)animals with the Xhol fragment rescued both the Mec and Dyf phenotypes. Eight cDNAs were obtained by probing Stuart Kim's lambda-
gt10 library and Bob Barstead's lambdaZAP library with the 2.5kb HindIII fragment. Restriction enzyme analysis indicates that the eight cDNAs are identical except for length differences at the 5' end. Both strands of the longest cDNA (1.5 kb) were sequenced. The longest ORF, starting about 15 nt from the 5' cloning site, is 323 amino acid residues long and leads to a 3' polyadenylation signal and a poly-A tail. The putative peptide encoded by this ORF is similar to a family of RNA-binding proteins. The closest similarity is to the Drosophila gene couch potato which shows 47.3% identity to
mec-8 over 184 amino acids. Both genes contain RNPs. consensus sites for RNA binding. Severe alleles of couch potato cause embryonic lethality, and weaker alleles affect functions of the peripheral nervous system (Bellen et al., Genes and Development 6, #11:2125-2136, 1992). Weakly similar to the
mec-8 peptide are the D. melanogaster loci Sex lethal and Elav, both of which encode putative RNA-binding proteins. Furthermore, the
mec-8 peptide is similar to snRNPs and nucleolins from frog, human, fruit fly, potato and yeast.
mec-8 may affect the processing of mRNAs, possibly the splicing of a subset of mRNAs that affect muscle and neuronal function. The interaction of
unc-52 with
mec-8 suggests that
unc-52 may be a target for
mec-8 action;
unc-52 rnRNAs are in fact differentially spliced (Rogalski et al., 1993). We are collaborating with Don Moerman to explore this possibility.
unc-52 is involved in the attachment of muscle cells to basement membrane (Gilchrist and Moerman, 1992). The muscle defects seen in
mec-8 homozygotes and the Pat phenotype of
mec-8 ;
unc-52 (viable)double mutants could thus be caused by aberrant splicing of
unc-52 mRNAs. Four recessive suppressors of the synthetic lethality of
mec-8 (
u218);
unc-52 (
e669su250)define two loci. The Mec and Dyf phenotypes of weak
mec-8 alleles (
u218 and
mn472 )are well suppressed by each mutation. To a lesser extent, each suppresses the Dyf phenotype of stronger
mec-8 alleles. All four alleles suppress the Unc phenotype of
unc-52 (
e669su250),the weakest
unc-52 allele; sup
(mn416 )11, suppresses the Unc phenotype of
unc-52 (
e669).No suppression of other
unc-52 alleles is seen. By themselves, the suppressors are slightly sick and are uncoordinated at a low penetrance, but show no other phenotype. The suppressors may define loci that are involved in mRNA splicing.