The Cys2-His2 class of zinc finger (Zif) proteins is among the largest class of proteins in nature. A single Cys2-His2 domain is ~30aa long, has a structure, and typically binds a three-base sequence of DNA. Owing to their modularity, such fingers can be joined together to generate polydactyl DNA-binding proteins of high specificity: a single six-fingered protein can recognize a unique sequence among 70 billion unrelated sequences, which is unique in any genome. Work by Carlos Barbas, David Segal and others has focused on identifying specific amino acid residue changes in the -helix to expand the repertoire of available 3-bp target modules. To date, their work has identified zinc fingers that permit access to target sites of the form 5'-GNN-3' and 5'-ANN-3'. We are interested in two broad applications of this technology as additions to the rich repertoire of tools available with the C. elegans system. First is the generation of artificial transcription factors. Fusions of six-fingered proteins with effector (activator or repressor) domains have been shown to function in plants and tissue culture (e.g. Beerli et al., 1998). We are currently constructing 6xZif-effector proteins for testing on the
ges-1 gene. Potential downstream applications of this technology include lineage-specific transcriptional repression of endogenous genes, and the generation of expression systems similar to the GAL4 system used in Drosophila. The second application is the generation of gene knockouts by gene conversion. Pairs of triplet zinc fingers can be joined to endonuclease domains to create a bipartite endonuclease that can be targeted to a single chromosomal locus. In Drosophila, such zinc finger nucleases were shown to mediate generation of small-deletion mutants in the y gene following activation of a nuclease and recovery of germline non-homologous end joining (NHEJ) events (Bibikova et al., 2002). In C. elegans, double-stranded (ds) breaks created by Tc1 excision have been shown to promote gene conversion from multicopy extrachromosomal transgenes (Plasterk and Groenen, 1992). Our goal is to fuse these approaches to generate a system for gene conversion: i.e., we propose to use zinc finger nucleases to generate ds breaks in the presence of homologous repair transgenes. We are currently constructing zinc finger nucleases that will target the
unc-22 locus for gene conversion from a mutated transgene. Progress with these experiments will be described.