Page last updated on 7/28/07. Contact the webmaster if there are problems with this website.
PublicationsResearchLab tools, stock lists and protocolsMembers of our labContact Us!The Biochemistry DepartmentThe Molecular Biology ProgramProtocols and stocks

Research

Our lab works on gene targeting. Our approach exploits the observation, made many times in many organisms, that double-strand breaks in DNA stimulate local mutagenesis and homologous recombination. We make use of enzymes called zinc finger nucleases (ZFNs) to produce targeted breaks in chromosomal DNA. Cells will often repair these breaks inaccurately, generating mutations at the cut site. When we provide a homologous donor DNA, repair by homologous recombination can replace sequences at the chromosomal site with ones engineered into the donor. Thus, it is possible to introduce designed changes in genes of interest.

The key to successful targeting is the ability to design ZFNs for arbitrarily chosen targets. The structure and requirements of these hybrid enzymes are described in more detail in our Current Research section. The important features are: 1) that zinc fingers are DNA-recognition modules, each of which binds primarily 3 base pairs of DNA; 2) the modules bind essentially independently, so they can be linked in new combinations; 3) fingers for many of the 64 DNA triplets have been identified, so a broad range of DNA sequences can be addressed; 4) the cleavage domain must dimerize to cut DNA, so ZFNs must be designed in pairs for any new site; and as a corollary, a monomeric ZFN does not cut DNA.

Working with the fruit fly, Drosophila melanogaster, we have achieved both targeted mutagenesis and targeted gene replacement in the germline at frequencies greater than 10% in the best cases. Both our lab and others’ are working to improve the procedures for ZFN-induced targeting and extend them to other cells and organisms. We have had good preliminary success in the nematode, Caenorhabditis elegans, and there is no reason why this approach shouldn’t work essentially everywhere, as long as methods for delivery of the necessary components are available.