Stabilization of radical anion states of nucleobases in DNA.

Trapping of an electron by DNA leads to the formation of radical anion states of pyrimidine bases. Because these states play an important role in biological and chemical processes, their computational treatment is of particular interest. We show that simple electrostatic and quantum chemical models can accurately reproduce the adiabatic electron affinities (EAs) of short DNA stacks recently derived from high-level ab initio calculations (M. Kobylecka, J. Leszczynski, and J. Rak, J. Am. Chem. Soc., 2008, 130, 15683). The electrostatic interaction of an excess electron localized on cytosine or thymine with intra- and inter-strand adjacent nucleobases is found to strongly affect the energy of the radical anions. This interaction is the main origin of the dependence of EA of nucleobases on the nature of neighboring base pairs. In particular, the states XT(-)Y and XC(-)Y, where X and Y = C, T, are, by ca. 0.7 eV, more stable than radical anions GT(-)G and GC(-)G. We find that second-neighbor effects can also significantly modulate EAs, although being smaller than the effects of adjacent bases. The strongest destabilizing effect is found for 5'-GC and 3'-GC, while the 5'-AT base pair stabilizes the radical anion states. Using a combined QM/MD approach, we consider how structural fluctuations of DNA influence the stability of the radical anion states. Despite large dispersions of the stabilization energies due to conformational dynamics of DNA, there are only few thermally accessible structures where GT(-)G and GC(-)G are energetically more favorable than the corresponding pyrimidine triplets. Although stabilization energies calculated for stacks of regular structure are in qualitative agreement with the QM/MD results, structural fluctuations of pi stacks should be taken into account for more accurate description of the excess electron trapped by DNA. The results obtained in this study suggest that simple electrostatic models, in combination with MD simulations, can be very helpful to explore the long time scale behavior of radical anions in DNA.