Exposure to aristolochic acid I and II (AAI and AAII) has been implicated in aristolochic acid nephropathy and urothelial carcinoma. The toxicological effects of AAs are attributed to their ability to form aristolacatam (AL)-purine DNA adducts. Among these lesions, the AL-adenine (ALI-N6-A and ALII-N6-A) adducts cause the “signature” A → T transversion mutations associated with AA genotoxicity. To provide the currently missing structural basis for the induction of these signature mutations, the present work uses classical all-atom molecular dynamics simulations to examine different (i.e., preinsertion, insertion, and postextension) stages of replication past the most abundant AA adduct (ALI-N6-A) by a representative lesion-bypass DNA polymerase (Dpo4). Our analysis reveals that, before dNTP incorporation (i.e., preinsertion step), ALI-N6-A adopts a nearly planar conformation at the N6-linkage and the ALI moiety intercalates within the DNA helix. Since this conformation occupies the dNTP binding site, the same planar lesion conformation results in a significant distortion of the polymerase active site at the insertion step and therefore replication will likely not be successful. However, if ALI-N6-A undergoes a small conformational change to introduce non-planarity at the N6-linkage during the insertion step, minimal distortion occurs in the Dpo4 active site upon incorporation of dATP. This insertion and subsequent extension would initially lead to A:A mismatches and then result in A → T transversion mutations during the second round of replication. In contrast, if a large conformation flip of the ALI moiety occurs at the insertion step to reorient the bulky moiety from an intercalated position into the major groove, dTTP (non-mutagenic) incorporation will be favored. Molecular dynamics (MD) simulations on postextension complexes reveal that damaged DNA will likely further rearrange during later replication steps to acquire a base-displaced intercalated conformation that is similar to that previously reported for (unbound) ALI-N6-A adducted DNA, with the exception of slight non-planarity at the lesion site. Overall, our results provide a structural explanation for both the successful non-mutagenic lesion bypass and the preferential misincorporation of dATP opposite ALI-N6-A and thereby rationalize the previously reported induction of A → T signature transversion mutations associated with AAs. This work should thereby inspire future biochemical experiments and modeling studies on the replication of this important class of DNA lesions by related human translesion synthesis polymerases.

Ochratoxin A (OTA) is a ubiquitous food toxin associated with chronic nephropathy in humans and renal carcinogenicity in rodents. The mutational spectra of cells exposed to OTA reveal that one-base deletions comprise the largest percentage (73%) of the total mutations that occur upon OTA exposure. To contribute toward understanding the prevalence of OTA-induced one-base deletion mutations, the present work uses molecular dynamics (MD) simulations to analyze the conformational preferences of one-base deletion duplexes containing OT-G, the major OTA adduct (addition product) at the C8-site of guanine. Specifically, the influence of OT-G in four possible ionization states and three sequence contexts (G1, G2 and G3 in the NarI (5′–G1G2CG3CC–3′), a prokaryotic mutational hotspot sequence) on the structure of the adducted DNA is investigated. Our data reveal that the damaged helices are stable in two (B-type (B) and stacked (S)) conformations that are structurally similar to those adopted by common N-linked C8-guanine lesions. However, the adduct ionization state and sequence context affect the degree of helical distortion and the B/S conformational heterogeneity, which will impact the lesion repair and replication outcomes. This finding correlates with the experimentally reported tissue-specific mutagenicity of OTA exposure. Furthermore, regardless of the adduct conformation, ionization state, or sequence context, more stable lesion-site interactions and lack of disruption of the flanking base pairs in the one-base deletion duplexes compared to the corresponding two-base deletion helices rationalize the greater abundance of OTA induced one-base deletions. Overall, our work provides valuable structural insights that help explain the experimentally observed mutagenicity associated with OTA.