Abstract
Clustered DNA damages are well-established characteristics of ionizing radiation. As a model clustered lesion in the same strand of DNA, we have evaluated the mutagenic potential of 8-oxoguanine (8-oxoG) adjacent to a uracil in simian kidney cells using a phagemid vector. The uracil residue would be excised by the enzyme uracil DNA glycosylase in vivo generating an abasic site (AP site). A solitary uracil in either GUGTC or GTGUC sequence context provided >60% progeny containing GTGTC indicating that dAMP incorporation opposite the AP site or uracil occurred, but a >30% population showed replacement of U by A, C, or G, which suggests that dTMP, dGMP, or dCMP incorporation also occurred, respectively, opposite the AP site. While the preference for targeted base substitutions at the GUG site was T ≫ C > A > G, the same at the GUC site was T ≫ A > C > G. We conclude that base incorporation opposite an AP site is sequence-dependent. For 8-oxoG, as compared to 23−24% G→T mutants from a single 8-oxoG in a TG8-oxoT sequence context, the tandem lesions UG8-oxoT and TG8-oxoU generated ∼60 and >85% progeny, respectively, that did not contain the TGT sequence. A significant fraction of tandem mutations were detected when uracil was adjacent to 8-oxoG. What we found most interesting is that the total targeted G8-oxo→T transversions that included both single and tandem mutations at the TG8-oxoU site was nearly 60% relative to about 30% at the UG8-oxoT site. A higher mutational frequency at the TG8-oxoU sequence may arise from a change in DNA polymerase that is more error prone. Thermal melting experiments showed that the Tm for the 8-oxoG:C pair in the TG8-oxo(AP*) sequence in a 12-mer was lower than the same in a (AP*)G8-oxoT 12-mer with ΔΔG 0.8 kcal/mol (where AP* represents tetrahydrofuran, the model abasic site). When the 8-oxoG:C pair in each sequence was compared with a 8-oxoG:A pair, the former was found to be more stable than the latter. The preference for C over A opposite 8-oxoG for the (AP*)G8-oxoT 12-mer duplex with a ΔΔG of 1.6 kcal/mol dropped to 0.4 kcal/mol in the TG8-oxo(AP*) 12-mer duplex. This suggests that the polymerase discrimination to incorporate dCMP over dAMP would be less efficient in the TG8-oxo(AP*) sequence relative to (AP*)G8-oxoT. Additionally, the efficiency of recognition and excision of A opposite 8-oxoG by a mismatch repair protein may be impaired in the TG8-oxo(AP*) sequence context.