mec-3 ,a homeodomain-type transcriptional regulator, is the primary determinant of cell type for three sets of mechanosensory neurons in C. elegans.
mec-3 is of special interest because it is expressed at the end of a stereotyped cell lineage, and only in one of the two daughters or grand-daughters of
unc-86 -expressingcells upon an asymmetric cell division. One property of the stereotyped cell lineages is that with every cell division, there is a change in cell fate. To account for this, we have postulated that in a stereotyped cell lineage, an ordered series of specific transcription factors cooperate with a cell cycle-specific transcriptional regulator to sequentially activate each other. Asymmetric cell division and cell cycle-dependent changes in gene expression are together sufficient to generate a stereotyped cell lineage. Understanding why
mec-3 is well regulated should help us understanding these processes. The promoter region of
mec-3 has been studied by using
mec-3 -lacZfusion and HS-
mec-3 fusion promoter (Way et al., Genes & Dev. 5, 2199-2211). Various regions of the
mec-3 promoter have been fused with a truncated heat shock promoter (leaving one heat shock repeat and TATA) followed by lacZ. The expression of lacZ is then tested after heat shock. When a DNA fragment containing regions I, II, and III (defined by similarity with the C. vulgariensis
mec-3 promoter) is fused with the truncated heat shock promoter, B-galactosidase is expressed only in the ten
mec-3 -expressingcells. No detectable b-galactosidase is expressed without heat shock. When only region I is fused with the truncated heat shock promoter, b-galactosidase is expressed only in the postembryonic
mec-3 -expressingcells (AVM, PVM, and PVD neurons) after heat shock and only when the cells arise, which means region I is sufficient to direct establishment but not maintenance synthesis of
mec-3 .The
mec-3 and heat shock promoter elements are insufficient by themselves to induce transcription but appear to cooperate when both are active. We are dissecting this 71 bp region I and have found that when a 29-bp subregion is fused with the truncated heat shock promoter, lacZ is expressed in all anterior daughters or grand-daughters of
unc-86 -expressingcells, apparently including the PHC cells. PHC is an exception among the anterior descendants of
unc-86 -expressingcells, in that it does not express
mec-3 .However, PHC expresses b-galactosidase in a
sem-4 ;
mec-3 -lacZstrain (M. Basson and H.R. Horvitz, pers. comm.; J-Q. Run and J. Way, unpub.). These results suggest
sem-4 is responsible for the inhibition of
mec-3 in PHC. The second part of the region I might be the
sem-4 binding site.
sem-4 has been cloned by Michael Basson and is a zinc finger-containing protein. Further mutation of the 29bp region leads to significant decrease or elimination of expression. Thus this 29-bp appears to contain targets for 1)
unc-86 action, 2) cell cycle regulation as defined by the fact that only progeny of
unc-86 -expressingcells express the fusion, and 3) asymmetric expression, as defined by the fact that only anterior daughters or grand-daughters express the fusion. This region includes, in the first half, a POU-domain consensus binding site and a core cell cycle box (CGMA) defined by the common sequences from yeast HO and histone H2 Bpromoters, and a novel sequence in the second half. Our first-order speculation about the subdivision of sites in this region is that the POU consensus sequence is the
unc-86 binding site, the CGMA would mediate cell cycle control, and the novel sequence could mediate asymmetric expression of
mec-3 .Other arrangements are possible: for example, since the POU protein Oct 1st mediates cell cycle regulation of histone H2 Bin mammalian cells, the POU consensus site might have this function in
mec-3 .However, because any protein binding to this region should be interesting, we plan to isolate binding proteins from C. elegans extracts using the standard band-shift assay. [See Figures 1&2]