In C. elegans most posteriorly directed cell and growth cone migrations require
vab-8, a gene that encodes at least two protein products known as VAB-8L and VAB-8S. VAB-8L is a 1066 amino acid protein that contains an N-terminal kinesin-like motor domain and functions in
vab-8-dependent growth cone migrations. VAB-8S is colinear with the C-terminal half of VAB-8L, and lacks the kinesin-like motor domain. VAB-8S is necessary for certain
vab-8-dependent cell migrations. To identify VAB-8-interacting proteins, we conducted a yeast two-hybrid screen using full length VAB-8L as bait. One protein identified was UNC-51, a serine/threonine kinase required for proper axon outgrowth (Ogura et al., 1994). In yeast and in vitro, a region between the kinesin and shared domains of VAB-8L interacts with the C-terminal half of UNC-51, which lacks the kinase domain. Ogura et al. (1997) have shown that the C-terminal half of UNC-51 also interacts with UNC-14. Using yeast two-hybrid, we found that a portion of VAB-8L interacts with UNC-14, but full length VAB-8L does not. These observations suggest that UNC-14 interaction sites are masked in the full length VAB-8 protein. Several observations suggest that VAB-8, UNC-14 and UNC-51 also interact in C. elegans. First,
vab-8,
unc-14 and
unc-51 mutants display axon outgrowth defects. Second, all three genes are expressed in neurons that require
vab-8 function. Finally, misexpression of the UNC-51-binding domain of VAB-8L under control of the
ceh-23 promoter results in highly penetrant CAN cell and growth cone migration defects, presumably by interfering with UNC-51 binding with wildtype VAB-8L. We are in the process of determining whether this misexpression phenotype could be suppressed by simultaneous misexpression of
unc-51. Our results and those of Ogura et al. (1997) argue that VAB-8, UNC-14 and UNC-51 form a complex that functions in the outgrowth of certain axons. We have observed that overexpression of
vab-8 can suppress axon outgrowth defects of an
unc-51 mutant, suggesting a positive regulatory relationship between the two genes. Two possibilities we are testing are whether VAB-8L is an UNC-51 target and whether VAB-8L can regulate UNC-51 kinase activity. Ogura et al. (1994) Genes & Dev. 8: 2389-2400. Ogura et al. (1997) Genes & Dev. 11: 1801-1811.