As part of the grand homeobox search (1), we have found nine loci that hybridize with oligonucleotides designed to pick out POU domain genes. The two of these that we have sequenced,
ceh-6 and
ceh-18, are real POU genes. We guess that there are at least 5-10 POU genes in C. elegans.
ceh-6 is located on IC, near
lin-28. Its POU domain sequence is most similar to a class of mammalian genes including Brn-1( 2) (the Brn genes were cloned by PCR, so sequence is available only for a small fragment). Another mammalian gene, Brn-3(2), has a POU domain more closely related to
unc-86 than to other mammalian genes. Thus at least two subfamilies of POU domain genes have members in both worms and mammals. This strongly suggests that the
unc-86 and
ceh-6 subfamilies existed before worms split from mammals.
ceh-18 was first cloned as cDNAs from the Kim library, and later placed somewhere on XL. Based on cDNA frequencies, it seems to be the most highly expressed POU gene. It's a big gene, 30-50 kb genomic, although all of the cDNAs are less than 3 kb.
ceh-18 is the most divergent POU gene found so far the POU domain sequence differs from the previous consensus at seven locations, and it has the longest linker region between the POU- specific region and the homeodomain. Given the relationships mentioned above, we expect that genes related to
ceh-18 will also be found in mammals. [See Figure 1]