Molecular electrostatic potentials of DNA base–base pairing and mispairing

An understanding of why adenine (A) pairs with thymine (T) and cytosine (C) with guanine (G) in DNA is very useful in the design of sensors and other related devices. We report the use of dissociation energies, geometries and molecular electrostatic potentials (MEPs) to justify the canonical (AT and CG) Watson-Crick pairs. We also analyze all mismatches in both configurations—cis and trans—with respect to their glycoside bonds. As expected, we found that the most stable pair configuration corresponds to CG, providing an energy criterion for that preferred configuration. The reason why A gets together with T is much more difficult to explain as the energy of this pair is smaller than the energy of some other mismatched pairs. We tested MEPs to see if they could shed light on this problem. Interestingly, MEPs yield a unique pattern (shape) for the two canonical cases but different shapes for the mismatches. A tunnel of positive potential surrounded by a negative one is found interconnecting the three H-bonds of CG and the two of AT. This MEP tunnel, assisted partially by energetics and geometrical criteria, unambiguously determine a distinctive feature of the affinity between A and T as well as that between G and C.