A Method for In Vivo Targeted Mutagenesis

Directed evolution protocols often rely on in vitro targeted mutagenesis methods capable of directing changes to a single gene or portion of a gene. However, they are not easily scalable to improve complex multigenic phenotypes that are often encountered in cellular and metabolic engineering. Classical in vivo mutagenesis techniques for improving such phenotypes elevate the overall global mutation rate. In the absence of any knowledge of the genotype—phenotype relationship of the trait in question, this may be the only viable strategy. But when key genes are known that influence this trait, in vivo mutagenesis is highly inefficient as most mutations that occur will not be within the genes of interest and can have a deleterious effect on the organism.

Here we describe a strategy for in vivo targeted mutagenesis of a defined region (10-20 kb) of DNA which results in an increase of local at the targeted DNA sequency by O(10³) over wildtype rates while limiting the increase in global mutation rates to untargeted DNA sequences to O(10). The method relies on the localization of chimeric fusions of DNA repair enzymes with a DNA binding domain. Localization is achieved through the introduction of repeats of the cognate DNA binding site near the genomic region to be mutated. We demonstrate the viability of this strategy for two different classes of DNA repair enzymes which should allow for control of the mutational spectrum. We suggest two possible mechanisms for localized mutagenesis consistent with the degree of mutation observed. Finally, we discuss on-going applications requiring the mutagenesis of multiple genes where we believe this method will lead to the rapid and directed evolution of a desired trait.