Fidelity of uracil-initiated base excision DNA repair in DNA polymerase β-proficient and -deficient mouse embryonic fibroblast cell extracts

Uracil-initiated base excision DNA repair was conducted using homozygous mouse embryonic fibroblast DNA polymerase (+/+) and (-/-) cells to determine the error frequency and mutational specificity associated with the completed repair process. Form I DNA substrates were constructed with site-specific uracil residues at U-A, U-G, and U-T targets contained within the lacZ alpha gene of M13mp2 DNA. Efficient repair was observed in both DNA polymerase beta (+/+) and (-/-) cell free extracts. Repair was largely dependent on uracil-DNA glycosylase activity because addition of the PBS-2 uracil-DNA glycosylase inhibitor (Ugi) protein reduced (similar to 88%) the initial rate of repair in both types of cell-free extracts. In each case, the DNA repair patch size was primarily distributed between 1 and 8 nucleotides in length with 1 nucleotide repair patch constituting similar to 20% of the repair events. Addition of p21 peptide or protein to DNA polymerase beta (+/+) cell-free extracts increased the frequency of short-patch (1 nucleotide) repair by similar to2-fold. The base substitution reversion frequency associated with uracil-DNA repair of M13mp2op14 (U-T) DNA was determined to be 5.7-7.2 x 10(-4) when using DNA polymerase P (+/+) and (-/-) cell-free extracts. In these two cases, the error frequency was very similar, but the mutational spectrum was noticeably different. The presence or absence of Ugi did not dramatically influence either the error rate or mutational specificity. In contrast, the combination of Ugi and p21 protein promoted an increase in the mutation frequency associated with repair of M13mp2 (U-G) DNA. Examination of the mutational spectra generated by a forward mutation assay revealed that errors in DNA repair synthesis occurred predominantly at the position of the U-G target and frequently involved a 1-base deletion or incorporation of dTMP.