Diffusion Approach to Long Distance Charge Migration in DNA: Time-Dependent and Steady-State Analytical Solutions for the Product Yields.

In this study we report analytical solutions for both time-dependent and steady-state problems of unbiased charge transfer through a regular DNA sequence via a hopping mechanism. The phenomenon is treated as a diffusion of charge in a one-dimensional array of equally spaced and energetically equivalent temporary trapping sites. The solutions take into account the rates of charge hopping (k), side reactions (k(r)), and charge transfer to a terminal charge acceptor (k(t)). A detailed analysis of the time-dependent problem is performed for the diffusion-controlled regime, i.e., under the assumption that k(t) >> k, which is also equivalent to the fast relaxation limit of charge trapping. The analysis shows that the kinetics of charge hopping through DNA is always multiexponential, but under certain circumstances it can be asymptotically approximated by a single-exponential term. In that case, the efficiency of charge transfer can be characterized by a single rate constant k(CT) = 1.23kN(-2) + k(r), where N is the DNA length expressed in terms of the number of equidistant trapping sites and k(r) is the rate of competing chemical processes. The absolute yield of charge transfer under steady-state conditions in general is obtained as Y(infinity) = omega alpha sinh(alphaN) + omega cosh(alphaN), where alpha = (2k(r)/k)(1/2) and omega = 2k(t)/k. For the diffusion-controlled regime and small N, in particular, it turns into the known "algebraic" dependence Y(infinity) = 1 + (k(r)/k)N(2). At large N the solution is asymptotically exponential with the parameter alpha mimicking the tunneling parameter beta in agreement with earlier predictions. Similar equations and distance dependencies have also been obtained for the damage ratios at the intermediate and terminal trapping sites in DNA. The nonlinear least-squares fit of one of these equations to experimental yields of guanine oxidation available from the literature returns kinetic parameters that are in reasonable agreement with those obtained by Bixon et al. [Proc. Natl. Acad. Sci. U.S.A.1999, 96, 11713-11716] by numerical simulations, suggesting that these two approaches are physically equivalent.