DFT and MD Studies of Formaldehyde-Derived DNA Adducts: Molecular-Level Insights into the Differential Mispairing Potentials of the Adenine, Cytosine, and Guanine Lesions.

4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanone is a potent nicotine-based carcinogen that generates many DNA lesions, including the HOCH2-C, HOCH2-G, and HOCH2-A hydroxymethyl adducts. Despite all lesions containing an altered exocyclic amino group, which allows the hydroxymethyl group to be directed away from the Watson-Crick binding face, only the most persistent adenine adduct is mutagenic. As a first step toward understanding this differential mutagenicity, density functional theory (DFT) and molecular dynamics (MD) simulations were used to gain atomic-level structural details of these DNA damage products. DFT calculations reveal that all three lesions exhibit conformational diversity. However, regardless of the hydroxymethyl-nucleobase orientation, both DFT and MD simulations highlight that HOCH2-C and HOCH2-G form pairs with the canonical complementary base (G and C, respectively) that are structural and energetically preferred over mispairs. In contrast, depending on the hydroxymethyl-nucleobase orientation, the Watson-Crick HOCH2-A:T pair can become significantly destabilized relative to undamaged A:T. As a result, HOCH2-A mispairs with G, C, and A are energetically accessible and maintain key geometrical features of canonical DNA. Overall, our data directly correlate with the reported differential mutagenicity of the hydroxylmethyl lesions and will encourage future studies to further uncover the cellular impact of the most persistent adenine lesion.

Computational insights into the mutagenicity of two tobacco-derived carcinogenic DNA lesions.

4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone is a potent carcinogen found in all tobacco products that leads to a variety of DNA lesions in cells, including O6-[4-oxo-4-(3-pyridyl)butyl]guanine (POB-G) and O6-[4-hydroxy-4-(3-pyridyl)butyl]guanine (PHB-G), which differ by only a single substituent in the bulky moiety. This work uses a multiscale computational approach to shed light on the intrinsic conformational and base-pairing preferences of POB-G and PHB-G, and the corresponding properties in DNA and the polymerase η active site. Our calculations reveal that both lesions form stable pairs with C and T, with the T pairs being the least distorted relative to canonical DNA. This rationalizes the experimentally reported mutational profile for POB-G and validates our computational model. The same approach predicts that PHB-G is more mutagenic than POB-G due to a difference in the bulky moiety hydrogen-bonding pattern, which increases the stability of the PHB-G:T pair. The mutagenicity of PHB-G is likely further increased by stabilization of an intercalated DNA conformation that is associated with deletion mutations. This work thereby uncovers structural explanations for the reported mutagenicity of POB-G, provides the first clues regarding the mutagenicity of PHB-G and complements a growing body of literature highlighting that subtle chemical changes can affect the biological outcomes of DNA adducts.