A TGF-beta signal transduction cascade controls body size and aspects of male tail morphogenesis in C. elegans, as evidenced by the Sma and Mab phenotypes of a TGF-beta-like ligand,
dbl-1 (Yo Suzuki and Bill Wood, personal communication), the TGF! receptors,
daf-4 (Estevez et al., Nature, 1993; Savage et al., 1996) and
sma-6 (Srikant Krishna and Rick Padgett, unpublished data), and the TGF-beta signal transducers
sma-2,
sma-3, and
sma-4 (Savage et al., PNAS, 1996).
sma-2,
sma-3, and
sma-4 are members of a conserved family of TGF-beta signal transducers, the Smads, that function downstream of TGF-beta receptors. Biochemical studies on the vertebrate Smads has led to a model in which Smad proteins are activated by phosphorylation by TGF-beta receptors allowing them to translocate to the nucleus and act as transcriptional activators. Smad proteins feature two conserved domains of ~100 amino acids, separated by a more divergent proline-rich linker. Although they are highly conserved, the Smads contain no recognizable structural motifs. Much work remains to be done to understand what structural features of the proteins are required for their activities. To begin to dissect the functional domains of these proteins, we are analyzing the structure and function of the
sma-3 gene through the analysis of 10
sma-3 mutants identified in genetic screens.
sma-3 mutants fall into two phenotypic classes, 3 weak alleles and 7 strong alleles, based on the penetrance of the male tail sensory ray fusions (Mab phenotype). In general, the severity of the Sma phenotype is also proportional to the penetrance of the Mab phenotype. The following table gives the percent frequency that a given ray is fused with another ray. Allele Ray 4 Ray 5 Ray 6 Ray 7 Ray 8 Ray 9 N
wk20 0 0 6 6 0 0 16
wk23 0 0 0 0 0 0 8
e958* 5 5 8 8 ND ND 118
wk21 14 14 86 86 14 14 7
wk24 10 10 70 70 20 20 10
wk27 12 12 65 65 0 0 17
wk28 9 18 91 91 27 27 22
wk30 0 0 100 100 20 20 5
e491* 28 38 34 37 ND ND 67
e637* 11 11 74 74 ND ND 119 *Savage et al., PNAS, 1996. We are sequencing the molecular lesions in these mutant alleles. Two of the strong alleles (
e637 and
wk27) are nonsense mutations in the linker and in the N-terminal domain, respectively; three others (
e491,
wk21 and
wk28) are missense mutations in conserved amino acids in the C-terminal domain. The C-terminal domain is also the site of most of the identified mutations in other Smad genes. Interestingly, the single weak allele that we have sequenced (
wk20) has a missense mutation in a conserved amino acid in the N-terminal domain. This result suggests that the N-terminal domain may not be completely essential for the function of the Smad protein.