We have determined the sequence of a portion of the mRNA encoded by
par-4, a gene required for localization of cytoplasmic components in early C. elegans embryos. We found that an allele of
par-4, (
zu187), isolated in a mutator screen in Jim Priess's lab, carried a novel Tc1-containing band when cut with the enzyme BglII. We isolated a small piece of DNA flanking the insert, but were unable to isolate either a cDNA clone or a additional genomic sequences despite screening over 300,000 plaques from various libraries. Therefore, we used RACE-PCR to amplify 5' sequence of the mRNA and, in total, have obtained about 1200 bases of coding sequence. We used the RACE-PCR product as a probe against other alleles of
par-4 and found that five
par-4 alleles showed polymorphisms in Southern blots:
lw38 (a gamma-induced allele from Jocelyn Shaw) and
it120,
zu160,
zu182 and
zu198 (mutator alleles). The putative
par-4 DNA also hybridizes to the YAC Y59A8, consistent with the position of
par-4 obtained from physical mapping data, and hybridizes to a single 3 kb RNA on a Northern blot. The sequence that we have obtained includes a sequence similar to the C. elegans SL-2 spliced leader followed by an ATG start and open reading frame extending another 870 bases. A BLAST search reveals strong similarity to a C. briggsae cDNA, and both sequences encode a serine-threonine protein kinase. Within the kinase domain, these two proteins share 82% identity. A Xenopus ovary kinase shares 46% amino acid identity with the
par-4 sequence and a rat calcium/calmodulin kinase shares 34% amino acid identity. We generated antibodies to 164 N-terminal amino acids upstream of the kinase domain in two rabbits, and found that antibodies from both rabbits recognize an 86 kD protein that is absent in
par-4 mutant worms. Indirect immunofluorescence shows that PAR-4, like the other PAR proteins, becomes concentrated at the cortex of cells in early embryos. However, unlike PAR-1, PAR-2 and PAR-3 proteins, PAR-4 is not asymmetric; PAR-4 protein is evenly distributed around the periphery of all cells up to about the 100 cell stage. This protein distribution is not altered by
par-1,
par-2,
par-3,
par-5 and
par-6 mutations. These observations are consistent with other data suggesting that
par-4 may be acting independently of the other par genes.