We are using the dauer larva as a model to investigate how animals couple environmental cues to a specific developmental program. Several of the genes involved in the dauer decision are homologs of transforming growth factor beta (TGF-beta) signal transduction proteins. These genes may act during the formation and wiring of the neurons involved in dauer formation or may function in the actual transduction of the pheromone signal. We are interested in the downstream effectors of the TGF-beta signal transduction pathway, so we cloned
daf-3, which suppresses the phenotype of mutations in
daf-1 and
daf-4, the TGF-beta receptors. We mapped
daf-3 to the region between
aex-3 and
unc-1 and identified B0217 as a rescuing cosmid. This region had been sequenced by the Genome Sequencing Project. This sequence allowed us to locate on B0217 a member of the DOT family (downstream of TGF-beta). The first DOT gene identified was the Drosophila gene Mad, which is required for signaling by dpp, a TGF-beta homolog. The DOT family is a family of proteins consisting of two conserved domains ("domains I and II"), separated by a proline-rich domain. Nothing is known about the function of these domains. We have identified two
daf-3 alleles with mutations in the DOT gene:
mg90 is a homozygous viable deletion that removes all of
daf-3, and
mg125 is a missense mutation in a conserved residue of domain I. In contrast, eleven out of twelve mutations identified in the other DOT family mutants are located in domain II. Eight of these mutations lie in a small hotspot of 45 amino acids. We have sequenced this hotspot in nine
daf-3 alleles, and none of them has a mutation in this hotspot. We have constructed a 15 kb rescuing subclone predicted to contain only the DOT gene. We have also characterized full-length cDNAs of
daf-3, and we have found evidence of alternative splicing at the 5' end. Analysis of a
daf-3 expression construct indicates that
daf-3 is expressed in most neurons and in the intestine, but not elsewhere in the animal; thus
daf-3 may be required for proper sensation of pheromone signal or proper transduction of the signal through the nervous system. We are planning to examine the localization and expression of
daf-3 in various mutant backgrounds, as well as in the presence or absence of pheromone to try to understand the mechanism of signal transduction. The fact that
daf-3 suppresses
daf-1 and
daf-4 indicates that
daf-3 acts oppositely to all of the known members of the DOT family, which act as positive transducers of the TGF-beta signal. These positively acting DOTs include
daf-8 and
daf-14, which have a phenotype like that of
daf-1 and act to positively transduce a
daf-7 signal (personal communication, A. Estevez and D. Riddle, T. Inoue and J. Thomas). The prevalence of mutations in domain II in many DOT members, particularly in the hotspot, suggests that this region is important for the function of these positively acting DOT genes. Consistent with this model,
daf-8 mutations affect domain II (personal communication, A. Estevez and D. Riddle), and
daf-14 mutations affect the hotspot in domain II (personal communication, T. Inoue and J. Thomas).
daf-3 mutations, on the other hand, do not cluster in the hotspot, and the location of the point mutation in
daf-3 indicates that domain I is important for the function of
daf-3. The combination of positively and negatively acting DOT genes in a single genetic system will allow us to investigate the biochemical function of the DOTs.