We have previously shown that
aex-3 is a regulator of presynaptic activities (1).
aex-3 mutants show various phenotypes that are characteristic of synaptic-transmission defects. As well as other synaptic-transmission mutants,
aex-3 mutants are resistant to the acetylcholinesterase inhibitor aldicarb, suggesting that the acetylcholine transmission is impaired in the mutants. Furthermore, immunostaining assays showed that RAB-3 is mislocalized in neuronal cell bodies instead of presynaptic terminals in the mutant while RAB-3 is predominantly localized at presynaptic terminals in the wild-type animals. RAB-3 is a small GTP binding protein and is implicated in synaptic vesicle release (2). This observation indicates that AEX-3 is necessary for normal subcellular localization of RAB-3 at presynaptic terminals. A molecular characterization of
aex-3 revealed a strong homology to Rab3 GTP/GDP exchange protein (GEP) (3). Taken all together, we hypothesized that
aex-3 regulates presynaptic activities partly through a RAB-3-dependent pathway. In addition to the RAB-3-dependent pathway, a phenotypic characterization of the
aex-3 and
rab-3 mutants implicated that AEX-3 also controls neural activities through a RAB-3-independent pathway. The most obvious
aex-3 phenotype is a defect in the defecation motor program (DMP). The DMP consists of three muscle contraction steps, posterior body wall muscle contraction, anterior body wall muscle contraction, and enteric muscle contraction (or expulsion). In the
aex-3 mutants, the last two steps are frequently missing. However, the
rab-3 mutant does not show this phenotype, suggesting that
aex-3 does have functions more than a
rab3 GEP. Recently, a group of Genetics Institute reported a characterization of the human gene MADD (MAP kinase activating death domain containing protein) [4]. MADD interacts with Tumor necrosis factor receptor through the death-domains (a type of protein interaction domains) and activates the ERK MAP Kinase. Surprisingly, it turned out that MADD is a human homologue of AEX-3 and the rat GEP. A sequence comparison between AEX-3 and MADD showed that their homology is particularly strong around the death domains near the C termini, indicating that AEX-3 also binds to another protein using this domain. Based on these findings, we hypothesized that AEX-3 binds to another protein through its C-terminus and this interaction is responsible for the RAB-3-independent activities. To test this hypothesis, we have isolated a new gene that interacts with AEX-3s C terminus using the yeast two hybrid system. Tentatively we call this gene
cab-1 (the C terminus of AEX-3 binding). By combining between cDNA sequencing and RT-PCR, we determined the
cab-1 gene structure. This gene consists of seven exons and is trans-spliced with the SL1 leader sequence. The gene encodes a protein of 425 amino acids, which has no homology to other known proteins. The CAB-1 contains two strongly hydrophobic domains.
cab-1 lies slightly left from
lin-2. A candidate aex mutant is found in the collection of Erik Jorgensens lab at the University of Utah. Currently, we are testing if a
cab-1 DNA rescues this new aex mutant. Further characterization of
cab-1 will be discussed at the meeting. 1) Iwasaki K, et al. Neuron (1997)18: 613-622. 2) Nonet ML, et al. J Neurosci. (1997) 17: 8061-8073. 3) Wada M, et al. J Biol Chem. (1997) 272: 3875-3878. 4) Schievella AR, et al. J Biol Chem. (1997) 272: 12069-12075.