Molecular characterization of
sma-2 Cathy Savage, Scott Townsend, Alyce L. Finelli,and Richard W. Padgett. Waksman Institute and Department of Molecular Biology and Biochemistry, Rutgers University, Piscataway, NJ 08855. We are interested in cell signalling by TGF-beta-like ligands in C. elegans.
daf-4 has been shown to be a ser/thr kinase receptor that binds human BMP-2, a TGF-beta-like ligand (Estevez et al., Nature 365: 644). In addition to its dauer constitutive phenotype,
daf-4 mutants show several other phenotypes, suggesting that this receptor has multiple roles in nematode development. The genes
sma-2,
sma-3, and
sma-4 share two phenotypes in common with
daf-4: small body size (Sma) and specific male tail ray and spicule defects (Mab) (S. Baird, personal communication). The striking similarity in mutant phenotypes implicates
sma-2,
sma-3, and
sma4 in this TGF-beta-like signal transduction pathway. To characterize these genes molecularly, we have begun by focusing on the molecular cloning of
sma-2.
sma-2 maps to the portion of chromosome m that has been completely sequenced (Wilson et al. Nature 368: 32). We have examined the predicted open reading frames (ORFs) in the interval containing
sma-2. Since no homologies to known components of TGF-beta signalling pathways were found in this interval,
sma-2 is likely to encode a novel protein involved in TGF-beta signal transduction. Using the genetic map position of
sma-2, we targeted several predicted ORFs that may be encoded by the
sma-2 gene. We used PCR to clone fragments containing these ORFs from
sma-2 mutants and DNA sequence analysis to identify any alterations in the mutants. In this way, we have identified a point mutation in sma- 2(eS02), suggesting that the predicted ORF may be the product of the
sma-2 gene. We are currently sequencing other
sma-2 alleles to confirm this result. Further confirmation will come from transformation rescue with genomic DNA containing
sma-2. We used a probe from the predicted ORF to screen Andy Fire's mixed stage and embryonic and Stuart Kim's cDNA libraries. Each of these libraries gave positive clones. The longest cDNA of 1.6kb was sequenced. This cDNA is mostly co-linear with the ORF ZK370.2, but diverges significantly 3' to this ORF. The predicted protein product is about 400 amino acids long, and shares no significant homology to any protein in the database. There are also no hydrophobic stretches, suggesting that the protein may be cytoplasmic or nuclear. lnterestingly, the genome sequencing has identified another predicted ORF, in the cosmid R13F6, that is homologous to ZK370.2 [WBG 13(#2)], indicating that
sma-2 may be a member of a gene family. Using the putative
sma-2 cDNA sequence, we examined the genomic sequence of R13F6. In this way, we have identified protein coding sequence in R13F6 that is related (47% identical) through the full extent of the
sma-2 cDNA. In addition, the amino acid alteration in
sma-2(
e502) maps to a glycine residue that is also present in the R13F6 sequence, suggesting that this residue may be functionally irnportant. Finally, R13F6 maps very near the location Or
sma-3. All intriguing hypothesis is that
sma-2 and
sma-3 encode related signalling molecules required for TGF-beta-like signal transduction. We are currently injecting R13F6 and nearby cosmids to test for rescue of
sma-3.