The gene
pal-1 is required in males for the generation of rays by the V6 cells. As we have reported,
pal-1 stimulates the production of rays by overcoming cell signals from the T cell that would otherwise inhibit ray production. (The Pal-1 V6 phenotype is suppressed by ablation of T.) The effect of
pal-1 is to stimulate the function of the gene
mab-5 and inhibit the function of
lin-22 in V6 or its descendants (Cell, in press Jan. 12, 1990). We have rescued
pal-1 animals by injection of a 14 kb fragment of DNA that appears to contain
ceh-3 [one of the scores of homeoboxes found in C. elegans, ( Burglin et al. Nature 341:239)].
pal-1 maps very close to
ced-4 on the left arm of linkage group III. Twelve recombinants between
unc-79 and
dyp-17 did not separate
ced-4 and
pal-1.
unc-79 and
dpy-17 are 0.25 map units apart. Thus with 95% confidence we know that
ced-4 and
pal-1 are less than or equal to 0.05 map units apart. This suggested that the genes might be within 50 kb of each other (using the conservative estimate of 1 Mb per map unit).
ced-4 has been cloned and mapped to a contig by Junying Yuan. We therefore sent to England for a series of cosmids in the region (Thank you John and Alan) and began injecting to look for
pal-1 rescue. We injected several of these cosmids (we used the gene
unc-31 and the unc- 31-bearing cosmid C14G10 as a coinjection marker) and obtained heritable but unstable rescue of the Pal-1 phenotype with the cosmid W05E6. According to the cosmid map W05E6 completely spans the cosmid AC10 from which Burglin et al. cloned the homeobox
ceh-3. It didn't take a Ph.D. to realize that this would be a good place to start looking for the
pal-1 gene. Using the published
ceh-3 protein sequence as a guide, we designed a degenerate oligonucleotide that should recognize the
ceh-3 sequence, and used it to probe Southern blots of W05E6. We were thereby able to generate a restriction map of the cosmid and localize the hybridization to a 1.8 kb BamHI/SalI fragment. (Since we have not sequenced the hybridizing sequence we can not say unequivocally that this is
ceh-3). We then subcloned a 14 kb (KpnI partial) fragment that contains the 1.8 kb Bam/Sal fragment (pWK14). We injected this 14 kb subclone into
pal-1 animals and rescued the Pal-1 phenotype. As noted above
pal-1 males are missing rays in their fans. Rescued animals had wildtype fans. In addition we saw one animal (out of 5) with an extra ray located in the body of the animal about halfway between the distal end of the gonad and the tail. Our numbers are still quite small so we cannot yet say how frequent the production of ectopic rays is. The generation of such ectopic rays might be due to negative control element(s) that are missing from the pWK14 plasmid or to overexpression of the gene caused by high copy number of the injected DNA, etc. In conclusion we have localized the
pal-1 gene to a 14 kb fragment with the ability to rescue the Pal-1 phenotype. This fragment probably contains the homeobox
ceh-3. Thus the question remains: Does
pal-1 have a homeobox? We cannot be certain, but it looks like it might. If so this would suggest that the activity of
pal-1 somehow overcomes the effect of cell signals by influencing transcription.