The
tra-3 gene promotes female development in XX animals: XX
tra-3 (m-z-) mutants are transformed from hermaphrodite to male, while XO
tra-3 mutants are essentially wild-type males. Barnes and Hodgkin (EMBO J., 15, 4477-4484, 1996) have shown that
tra-3 encodes a predicted homologue of the large subunit of calpain, a calcium-activated regulatory cysteine protease. Vertebrate calpain large subunits have four domains, numbered I-IV. Domain II contains the catalytic active site, and domain IV contains a series of calcium-binding EF-hand motifs. The predicted TRA-3 protein shows extensive sequence similarity to domains I-III of vertebrate calpains but lacks domain IV; instead, TRA-3 has a region called domain T. Two mammalian calpain homologues containing domain T, but not domain IV, have recently been identified; we propose that TRA-3 is the founding member of a new subfamily of calpain proteases. We are investigating how TRA-3 functions in the sex determination pathway and, more generally, how TRA-3 differs from other calpains. As a first step, we have shown that a full-length
tra-3 cDNA driven by its native promoter rescues all aspects of the
tra-3 mutant phenotype. A
tra-3 construct carrying a putative active-site disruption does not rescue the
tra-3 phenotype; this result suggests that TRA-3 is likely to have proteolytic activity. In order to delineate the regions required for TRA-3 activity, constructs expressing truncated forms of TRA-3 have been tested for their ability to rescue XX
tra-3 (m-z-) mutants (see below). We have also overexpressed
tra-3 in XO and XX animals using a heat shock-driven
tra-3 transgene. Our results indicate that XO males display mild somatic feminisation in response to
tra-3 overexpression, whereas XX animals are unaffected. In addition, we find that the
tra-3 promoter is capable of driving the expression of a GFP-lacZ(NLS) fusion in both XX hermaphrodites and XO males. This preliminary observation suggests that the difference in
tra-3 activity between the sexes does not appear to be controlled at the transcriptional level. Vertebrate calpains are postulated to be regulated by calcium through the EF-hand motifs present in domain IV. The absence of EF-hand motifs in TRA-3 initially suggested that TRA-3 activity might not involve calcium. However, closer examination of the TRA-3 domain T has revealed the presence of a putative C2 domain, which is a calcium-binding and phospholipid-binding motif. We have determined that a
tra-3 transgene lacking the C2 domain rescues XX
tra-3 (m-z-) mutants. However, a weak dominant negative phenotype is observed when the same transgene is expressed in N2 animals. This dominant negative phenotype is mimicked by a
tra-3 allele with a nonsense mutation within the C2 domain. We are presently examining whether calcium plays a role in regulating TRA-3 activity.