Last modified: October 18, 2001

UNC-44 Interaction with UNC-33

The unc-44 and unc-33 phenotypes are similar at a gross level (Hedgecock, 1985). One or more of the UNC-33 isoforms is localized to neural processes and to a few cell bodies in wild type C. elegans (Li, 1992). However in unc-44 backgrounds, the UNC-33 proteins are mis-localized from the nerve processes to the cell bodies. From these observations, Li (1992) proposed that a wild type unc-44 product(s) is required for proper axonal localization of UNC-33. It is possible that many of the neural defects in unc-44 mutants are due to the failure to localize UNC-33 protein. It has been proposed that UNC-33 is part of a signal transduction pathway that responds to collapsin (Goshima et al., 1995). It is known that UNC-33 fails to localize along axons in unc-44 mutants including unc-44(mn259, mn339, st200, e362, e781, and e949) (Li, 1992). We have shown that the mn259, mn339, and st200 alleles affect only the AO13 ankyrin isoform (Boontrakulpoontawee, 1995). Thus, it is clear that the large AO13 ankyrin interacts directly or indirectly with UNC-33, and is required for axon guidance.

Recently, vertebrate proteins related to UNC-33 have been shown to mediate collapsin related events (Goshima et al., 1995), suggesting a role for UNC-33 in axon retraction. UNC-44 may localize UNC-33 so that it can perform its function in axon outgrowth and guidance, and mis-localization of UNC-33 may contribute to some of the neuronal defects in unc-44 mutants.

There are three UNC-33 isoforms which all contain the same carboxyl terminal domain, but which differ in the length of the amino terminal region (Li et al., 1992). We hypothesize that AO13 ankyrin interacts with at least one of the larger UNC-33 isoforms from the following observations: UNC-33 protein fails to be localized to axons in unc-44(mn259), mn339, and st200 backgrounds (Li, 1992). These three DNA insertion alleles do not affect the conventional ankyrin isoforms, but they do alter the AO13 mRNA (Boontrakulpoontawee, 1995; Otsuka et al., 1995). We would expect that the other unc-44 alleles (e362, e781, and e949) with a similar inability to localize UNC-33 properly (Li, 1992) also affect AO13 ankyrin.

In terms of the UNC-33 portion of the interaction, it is likely that one or both of the larger UNC-33 proteins interacts with UNC-44 because the smallest isoform is produced in approximately normal amounts in a DNA insertion allele (Li et al., 1992), yet staining in the axons is very weak in comparison to wild type (Li, 1992). This suggests that the larger UNC-33 isoforms are localized in nerve processes. The three forms of UNC-33 may have a similar function because they all contain a common carboxyl terminal domain that has sequence similarities to dihydropyrimidinase. The amino terminal tails of the larger forms may allow their preferential localization within different parts of the cell. It seems somewhat unlikely that the interaction between UNC-33 and UNC-44 is via a coiled-coil structure, because the only region of UNC-33 predicted to be a coiled-coil is at the carboxyl end common to all three isoforms of UNC-33 (aa 788-807 in the largest isoform). Taken together, the results suggest that the UNC-33/UNC-44 interaction requires one or more non-conventional ankyrin domains of AO13 ankyrin (that is, the STEP repeats, hydrophobic regions, and/or the carboxyl terminus) and the amino terminal tails of one or both large UNC-33 isoforms. These results do not rule out the possibility of an indirect interaction between UNC-33 and UNC-44 through additional unknown molecules. Recently, two-hybrid screens with UNC-33 have detected an interaction with filamin, but no UNC-44 clones were recovered (Ken-ichi Ogura, Yokohama City University School of Medicine, personal communication). This result suggests that the interaction may be indirect.

Other laboratories have carried out two-hybrid screens to find interactions with known axonal guidance proteins. The UNC-51 serine/threonine kinase was found to interact with UNC-73 (a Trio homolog) and VAB-8 (a kinesin-related protein) (Ken-ichi Ogura, personal communication). UNC-73 is a guanine nucleotide exchange factor (GEF). UNC-51 was previously shown to bind to UNC-14 (Ogura et al., 1997). Two-hybrid screens with the cell surface receptor UNC-5, identified APR-1 and UNC-14 (Ingele Roelens, Universiteit Gent, Gent, Belgium, personal communication). How UNC-44 ties into these types of interactions is yet to be determined.

References

Boontrakulpoontawee P. 1995. The Caenorhabditis elegans unc-44 ankyrin gene: Wild-type, mutant, and revertant gene structures and transcripts. Ph. D. Thesis, Illinois State University. 211 p.

Goshima Y, Nakamura F, Strittmatter P, Strittmatter SM. 1995. Collapsin-induced growth cone collapse mediated by an intracellular protein related to UNC-33. Nature 376:509-514.

Hedgecock EM, Culotti JG, Thomson JN, Perkins LA. 1985. Axonal guidance mutants of Caenorhabditis elegans identified by filling sensory neurons with fluorescein dyes. Dev. Biol. 111:158-170.

Li W. 1992. Characterization of the Caenorhabditis elegans axonal guidance and outgrowth gene unc-33 and a variant Tc4 transposon. Ph. D. Thesis. University of Minnesota. 126 p.

Li W, Herman RK, Shaw JE. 1992. Analysis of the Caenorhabditis elegans axonal guidance and outgrowth gene unc-33. Genetics 132: 675-689.

Ogura K-I, Shirakawa M, Barnes TM, Hekimi S, Ohshima Y. 1997. The UNC-14 protein required for axonal elongation and guidance in Caenorhabditis elegans interacts with the serine/threonine kinase UNC-51. Genes & Dev. 11:1801-1811.

Otsuka AJ, Franco R, Yang B., Shim K-Y, Tang LZ, Zhang YY, Boontrakulpoontawee P, Jeyaprakash A, Hedgecock E, Wheaton VI, Sobery A. 1995. An ankyrin-related gene (unc-44) is necessary for proper axonal guidance in Caenorhabditis elegans. J. Cell Biol. 129:1081-1092.