The maternally-expressed gene
mei-1 has mutations that affect either meiosis or mitosis. Loss-of-function mutations prevent the formation of the meiotic spindle while a gain-of-function allele disrupts the first mitotic cleavage. These two classes of mutations show a complex pattern of genetic interactions with each other as well as with mutations in several interacting genes. We have cloned
mei-1 using the usual tricks involving the physical map and transformation rescue; both genomic and cDNA clones have been sequenced. The sequence shows some curious similarities to other genes, but their significance is not clear (identities are about 35% over 250 amino acids). Perhaps the most interesting similarity is to the CDC48 gene of Saccharomyces. Mutations of this gene block division prior to spindle-pole body separation; perhaps this is analogous to the failure of meiotic spindle formation of
mei-1 (lf) alleles. Other genes similar to
mei-1 ,also from Saccharomyces, are SEC 18, involved in vesicle trafficking, and PAS1 ,implicated in peroxisome formation. Perhaps there is a role for vesicles in spindle formation. However, the human gene TBP-1 ,which binds the TAT gene of HIV and regulates transcription, shows the same degree of similarity. Genes, of unknown function, from pig and Xenopus, are also similar to
mei-1 . One common feature these proteins are likely to share is nucleotide binding. Even though consensus sequences proposed for ATP binding sites are very loose, the
mei-1 -likegenes are strongly conserved in this region, well beyond that of the proposed consensus sequences. Indeed, the purified Xenopus protein shows in vitro ATPase and GTPase activities. Perhaps
mei-1 is a member of a conserved family of ATPases. Two additional C. elegans family members were among the sequences reported by the cDNA sequencing project (WBG 12(2): 26); these may represent some of our genetically identified interacting genes. The gene density in the
mei-1 region is fairly high: the beta subunit of casein kinase II is ~3 kb and
lin-10 is ~8 kb from the 5' end of
mei-1 while an uncharacteristic gene (identified by a cDNA isolated with a
mei-1 genomic clone) is immediately 3'. This latter gene is convergently transcribed with
mei-1 and the poly A sites of the two genes are only 10 base pairs apart. There is a hot-spot for recombination in this area; the two-factor distance between
mei-1 and
lin-10 is 0.3 cM, leading to an estimate of about 50 cM/kb, 30 times higher than Sulston et al. (Nature 356:37) have recently reported for the centre of LG III (
mei-1 is in the cluster of LG I). The amount of recombination seems to fall off on either side of this region: to the left (
lin-10 to
unc-13 )it is approximately 1000 kb/cM while to the right (
mei-1 to
unc-120 )it is about 500 kb/cM. Based upon the behavior of deficiencies with endpoints near
mei-1 and
lin-10 ,McKim and Rose also showed that this region is important for recombination