Using the Ascaris des homeobox AHB-1 as a probe we isolated two genomic clones from C. elegans, y CELHO, see WBG 10(2),1988, p. 36) and ceh- 12. Ceh-11 maps in the region of
egl-5. Ceh-12 maps to the first chromosome, to the right of
dpy-6 (A. Coulson and J. Sulston, personal communication). Out of more than 100 published sequences, the homeodomain with the highest level of similarity with
ceh-11 was X1Hbox2, an Antp class homeodomain from X. laevis. The sequence of the second homeobox clone (
ceh-12) is interrupted by an intron located upstream of the codon for the amino acid 45, inside the putative recognition helix. Most interestingly an intron at the same position, was found in five other known homeoboxes, namely two in C. elegans e in D. melanogaster ( labial, Distalless and proboscipedia). The conservation of intron position between evolutionary quite distant species could reflect some unknown function or selective constraint. On the same genomic clone, we found a tRNA(Arg) gene (
rtr-1) by sequence analysis. This tRNA decodes CGU, the most used Arg codon in C. elegans.In a further approach we used the
ceh-11 homeobox to screen the Ahringer egg cDNA library. In one of the positive clones we found a third homeobox (ceh- 13). The most similar homeodomains to
ceh-13 were
hox-1.6 (a murine homeobox) with 43 residues identical out of 60 and lab with 41 residues identical out of 60. Therefore
ceh-13 seems to belong to the labial subclass of homeoboxes. Furthermore, the similarity between
ceh-13 and lab extends somewhat downstream of the homeobox, the amino acids 61, 62, 65 and 66 being identical. Recently we have isolated the corresponding genomic clone and are currently investigating the structure of the
ceh-13 gene. Our three homeodomains have an Arg at position 9 of their recognition helix. According to the theory proposed by S. Hanes and R. Brent in their recent Cell paper (Cell 57, 1989, p. 1275-1283) we speculate that our C. elegans homeodomains have an Antp type DNA- binding specificity. [See Figure 1]