Many physiologically important proteins are synthesized as large inactive polypeptides. Activation commonly occurs by limited proteolytic cleavage at mono-, di or oligobasic residues, a reaction catalyzed by members of the
kex2/subtilisin-like proprotein convertase family (kexins) of serine endoproteinases. In vertebrates at least seven different family members have been identified and include furin, PC1/3, PC2, PC4, PC5/6, PC7 and PACE4. The combined efforts of ourselves, the genome sequencing consortium and the laboratory of Ian Dickerson (University of Miami School of Medicine), have identified four C. elegans kexin family members including
bli-4, Celfurin, CelPC2 and F32A7.6. We are interested in determining the developmental role of each of the kexins, and their substrate specificity. Originally we had shown that
bli-4 encoded four gene products that arise via alternate splicing in which the first 12 exons are joined to different downstream exon(s). A comprehensive genomic comparison with the C. briggsae
bli-4 locus now suggests that the gene encodes at least nine isoforms. Molecular and genetic studies of
bli-4 mutants suggests that one role of the convertases is the processing of cuticle collagens during development. To further our understanding of how the convertase gene accomplishes this feat, we are using epitope-tagged transgenes that encode candidate substrate proproteins to examine their spatial and temporal cleavage pattern in wild-type animals and
bli-4 mutant strains. Candidate genes which may encode BLI-4 substrates include
bli-1,
bli-2 and dpy- 5, all of which encode cuticle collagens that contain the Arg-X-X-Arg kexin cleavage motif toward the N-terminus.
dpy-5 mutations dominantly suppress blistering of the adult cuticle suggesting that DPY-5 may be involved in a complex interaction with BLI-1 and BLI-2, both of which localize within the cuticular structures called struts which are defective in blistered animals (J. Kramer pers. comm). F32A7.6 is expressed as part of an operon with
unc-54 suggesting that the role of this convertase is restricted to muscle tissue. The kexin is trans-spliced to SL2 consistent with it being the downstream gene in the operon. We recently isolated a deletion mutant,
h1557, which removes exons 6 to 10 of
unc-54 and the first exon of the kexin gene. Subsequently, the Unc-54 mutant phenotype was rescued using a transgene (kindly provided by A. Fire). The resulting transgenic animals were thin, slow growing, weak Egl and severely constipated. The constipated phenotype prompted us to examine whether the
unc-54-linked defecation mutant
aex-5 and F32A7.6 were the same gene. To this end, we found that
h1557 failed to complement
aex-5(
sa23). Furthermore, sequence analysis of the kexin gene from
aex-5(
sa23) homozygotes identified a missense mutation (Cys510Trp), providing further evidence that F32A7.6 is
aex-5. Together, this data suggests that proprotein convertase activity is required for defecation. This work was supported by a grant from the MRC of Canada.