By providing the strand-discrimination signal for DNA mismatch repair, bacterial adenine methylation at GATC sites plays an indispensable role in maintaining genome stability. In mammals, cytosines of CpG dinucleotides are frequently methylated, but no direct evidence exists which would implicate DNA methylation in DNA repair processes. Alterations in the patterns and levels of this methylation are among the most common changes associated with neoplasia, but little is known about the underlying mechanisms for these changes. A recent observation by the applicant that mutational rate is significantly elevated in mouse embryonic stem cells deficient in DNA methylation has raised the possibility that mammalian DNA methylation may also play an important role in DNA repair. This proposal will study this possibility by first revealing the types of mutations that are favored in a DNA methylation deficient background in an attempt to understand the mechanism responsible for the increased mutations. Then, a possible involvement of mammalian DNA methylation in mismatch repair will be studied by determining the effects of DNA methylation deficiency on microsatellite instability and promiscuity of DNA recombination, the two best-studied phenotypes associated with mismatch repair defects. Finally, the possibility of mammalian DNA methylation as a genetic component of the mismatch repair pathway will be directly tested through analyzing the phenotypes of cells defective in both DNA methylation and mismatch repair.