High frequencies of 5-methylcytosine (5meC) to thymine mutations dominate spectra of spontaneous mutations in genes. Cataloging of sequence change that cause human genetic diseases including beta-thalassemia, hemophilia, hypercholesterolemia and cancer have revealed that a disproportionately high number of mutations are C to T or G to A changes at sites of methylati - CpG. Earlier, similarly high mutation frequencies were noted at sites of methylation in E. coli. We have developed a simple, but sensitive genetic test that allows direct selection of the conversion of cytosines at sites of methylation to thymine. The test can be used to study mutations at methylation sites in E. coli and at CpG sites. Using this assay, we have shown that a bacterial methyltransferase, M.EcoRII, can also deaminate C to U and 5meC to T. We propose here experiments that should confirm this observation biochemically and extend it. The catalytic parameters for the enzyme- mediated deamination reactions will be determined and used to asses the relative importance of the enzyme-mediated and the spontaneous deamination reactions. We will use our genetic system to show that mammalian methyltransferases can also cause such deaminations. We will also construct strains of E. coli that mimic the conditions that are thought to exist during early stages of cancer- high MTase levels and a lower levels of the methyl donor, SAM. These strains will be used to test the hypothesis that at early stages in carcinogenesis, the MTase directly causes C to T mutations. Finally, we will isolate mutants of M.EcoRII will enhanced ability to cause mutations. Together these studies should help one evaluate the role of cytosine methyltransferases in creating C to T mutations and the relative merits of competing hypotheses regarding the mechanism of mutations at sites of cytosine methylation.