Bromouracil mutagenesis in Escherichia coli evidence for involvement of mismatch repair.

Summary Bromouracil mutagenesis was studied in several strains of E. coli in combination with measurement of incorporation of bromouracil in DNA. For levels below 10% total replacement of bromouracil for thymine, mutagenesis was negligible compared with higher levels of incorporation. Such a nonlinear response occurred both when the bromouracil was evenly distributed over the genome and when a small proportion of the genome was highly substituted. Also, the mutation frequency could be drastically lowered by amino acid starvation following bromouracil incorporation. These observations suggest the involvement of repair phenomena. Studies of mutagenesis in recA ^– and uvrA ^– mutants, as well as studies of prophage induction, did not support an error prone repair pathway of mutagenesis. On the other hand, uvrD ^– and uvrE ^– mutants, which are deficient in DNA mismatch repair, had much increased mutation frequencies compared with wild type cells. The mutagenic action of bromouracil showed specificity under the conditions used, as demonstrated by the inability of bromouracil to revert an ochre codon that was easily revertable by ultraviolet light irradiation. The results are consistent with a mechanism of bromouracil mutagenesis involving mispairing, but suggest that the final mutation frequencies depend on repair that removes mismatched bases.