TWO MUTATIONS THAT PERTURB CELL ATTACHMENTS Carolyn Norns and Ed-vard Hedcecock; Johns Hopkins Universitv, Dept. of Biology. One goal of our laboratorv is to understand the genetic control of axonal growth and guidance in C. elegans. Here we describe two genes, unc
(rh38) and exc
(rhl62), that we think function in cell-cell or cell-matrix; adhesion, both in neuronal and non neuronal tissues. Unc
(rh38) was isolated in a screen for mutants with abnorrnal PVPs, neurons that pioneer the ventral nerve cord (R. Durbin, 1985). Subsequent characterization revealed that the
rh38 mutation disrupts the outgrowth of many aYons, including pioneers. The outgrowth of another cellular process, the e:ccretory canal, is frequently stunted in
rh38 mutants. The
rh38 mutation also causes t-vo defects that have not been previously reported in C. elegans. First, the distal tip cell sometimes detaches from its germ cells. Second, the first polar body, which normally stays attached to the anterior end of the egg where it is expelled, often rotates to abnormal positions in
rh38 worms. Taken together we think l~rlc
(rhl38) disrupts cell-cell or cell-matrix attachment.
rh238 is probably a gain-of-function allele.
rh38/rhl38 worms are quite uncoordinated, nearly 100% have amphid and phasmid axon abnormalities, and the excretory canal is short in ~60% of these homozygotes.
rh38/+ is rarely (<1%) Unc, but the canal is short, and the amphid/phasmid axons are abnormal in 86 % and 96%, respectively, of these heterozygotes. However,
rh238/Df [mnDpl l, nDf23, nD.
f25] is not Unc and rarely has any detectable cellular defects. Another gain-of-function allele isolated by non-complementation for the Unc phenotype,
rh240, has properties similar to
rh238. To isolate a null mutant, we mutagened
rh38/rh38 worrns and looked for non Unc's in the first generation. So far we have obtained 4 new alleles that partially suppress the Unc and know that one,
rh38rh272, is an intragenic revertant.
rh38rh2272 is an impenetrant Unc, with frequent amphid and phasmid defects, and short canals in about 40% of the worms. Given that
rh38rh272 onlv partially suppresses the Unc phenotype of
rh238, we think it is unlikely to be a null mutation. The
rh38 locus has not been separated fromtax-2. unc
(rh238) maps to chromosome I between
unc-29 and aph-l, where
tax-2 is. The reason we suspect that
rh38 and
rh240 may be allelic to
tax-2 is that they have the
tax-2(
p691) phenotype of an amphid axon leaving the amphid commissure and travelling posteriorly in the ventral cord, and phasmid axcons that are occasionally too long (Coburn and Bargman, WM '93). However,
rh38/tax-2
(p691) looks like
rh8l+, not
rh38/Df. . And only two of 103 worms of genotype
rh38rh272/tax-2 had the
tax-2 amphid phenotype. So if
rh38 is allelic to
tax-2,
p691 may not be the null. Exc
(rh162) is interesting for two reasons. First, exc(rhl 62) has several phenotypes in common with rlnc
(rh38) including a short excretory canal, and the rare[ly reported] phenotypes of the detached distal tip cell and the floating polar body. Second, even though exc
(rh16') is not unc at all, preliminary experiments indicate exc(rhl 62) is a synthetic lethal with
unc-6. That is exc
(rhl62); 1lnc-6
(ev400 )/+ double mutants arrest early in development. We therefore think that the function of exc(rhll 62) is intimately related to that of Imc-6. This interaction with
lmc-6 is of great interest because UNC-6 is a laminin-lil;e molecule that guides growth cones pioneering dorsal-ventral pathways (Hedgecock et al., 1990 and Ishii and Wadsworth et al., 1992). exc
(rhl62) is on chromosome V to the right of dpy-l l . nDf32, sDf30, and sDJ20 delete
rll162, but nDp8 does not. That puts exc(rhl 62) between
let-337 and kra-l .
rhl62/Df mutnats are not noticeably worse that
rhl62 homozygotes. In summary, because lmc
(rh38) and exc(rhl 62) are the only two genes that are thus far known to cause detachment of the distal tip cell and polar body, we think these genes function similarly in cell-cell or cell-mati,~ attachment. Moreover, given the interaction of exc
(rh162) with
unc-6, these genes may be acting with UNC-6, a known axonal guidance molecule.