The glycosyl C1′-N9 bond of deoxyadenosine and deoxyguanosine:: response to electrophilic attacks on the purinic nitrogen atoms

Self-consistent-field computations shed light on two relevant conformations of deoxyadenosine (dA) and deoxyguanosine (dG): one with a pseudoequatorial C1'N9 glycosyl bond and the other, a slightly more stable one, with its C1'N9 bond in a bisectional orientation. In dA, both the N-3 and N-7 nitrogens are plausible sites for electrophilic attack, but only N-7 is a plausible site in dG. The addition of H+, CH3+, C2H5+ or tert-C4H9+ onto N-7 does not provoke notable structural modifications and leaves the base of dA and dG in an antiperiplanar (or nearly antiperiplanar) position with respect to the sugar C1'O4' bond, but N-3 additions cause the base to adopt a synperiplanar or strongly chiral position. This produces strong interactions between the purine and deoxyribose moieties, whose relief could aid the eventual cleavage of the glycosyl bond of dA. addition of a radical cation onto N-7 reduces the dissociation energy of the glycosyl bond by an estimated 8 kcal mol(-1) in dA and 4 kcal mol(-1) in dG - a bond weakening likely to concur to a depurination of DNA induced by radical cations.