rab-3 encodes a small-GTP protein that reversibly associates with synaptic vesicles. Rab3 function is thought to be regulated by its cycling between a soluble GDP-bound form and a vesicle-associated GTP-bound form. Here we characterize the C. elegans mutants of two candidate
rab-3 effectors first identified in vertebrates; rabphilin and RIM. The vertebrate proteins bind
rab3 specifically in the GTP-bound state. These two large proteins contain multiple domains including a zinc finger domain that mediates
rab3 binding and C2 domains that share homology to the calcium binding motifs of synaptotagmin. The C. elegans genes share the general organization of these domains and only exhibit similarity within these regions. Both C. elegans genes are expressed in the nervous system. Vertebrate rabphilin associates with synaptic vesicles; localization of worm rabphilin (RBF-1) in wild type animals and mislocalization of the protein in
unc-104 mutants suggests RBF-1 associates with synaptic vesicles. However, by contrast with the vertebrate homolog, C. elegans RBF-1 is not mislocalized in
rab-3 mutants. Rabphilin deletion mutants exhibit no detectable synaptic defects when examined using behavioral, pharmacological, physiological or ultrastructural assays. Vertebrate RIM is associated with the active zone at vertebrate terminals. C. elegans
unc-10 encodes the vertebrate RIM homolog. A functional worm
unc-10::GFP fusion is localized in tight puncta in synaptic-rich regions, consistent with localization to the active zone. However, this localization has yet to be confirmed with antibodies, and thus remains preliminary.
unc-10 deletion mutants exhibit moderate behavioral, pharmacological and physiological defects. Surprisingly, these defects are stronger than those of
rab-3 mutants suggesting that RIM functions in pathways independently of RAB-3 signaling. We are assessing if dominant
unc-10 mutants (isolated in E. Jorgensen' s lab) function independently of
rab-3 to provide further evidence of a RAB-3-independent pathway transduced through RIM. We are also characterizing the domains required for punctate localization of UNC-10 as a step towards identifying components of the presynaptic active zone.