DNA-mediated acid catalysis: Calculations of the rates of DNA-catalyzed hydrolyses of diol epoxides

DNA catalyzes the reactions of many small molecules and assists enzymes catalyzing modifications of DNA itself. Previous work by the authors showed that hydronium ions constitute an important component of the counterion atmosphere surrounding polyelectrolytes in general and DNA in particular. It was proposed that local regions at the surface of DNA, termed acidic domains, might be responsible for the protonation of epoxides to form reactive intermediates. This acid catalysis is DNA-mediated in the sense that DNA associatively binds the reactant hydronium ions. DNA catalysis of the hydrolysis of the syn- and anti-7, 8-diol 9, 10-epoxides of the procarcinogen benzo[a]pyrene (syn- and anti-BPDE) has been investigated for over a decade. Jerina and co-workers have shown that, in the absence of DNA, the observed rate of hydrolysis (k(obs)) can be represented as the sum of a spontaneous, pH-independent contribution (k(0)) and an acid-catalyzed component (k(H)): k(obs) = k(0) + k(H)[H+], where [H+] denotes the (bulk) hydronium ion concentration. An important aspect of that work is that while the acid-catalyzed reactions of both BPDE diastereomers occur via similar mechanisms, considerable differences are found in their rates of hydrolysis. Later work by Jerina involved the catalyzed hydrolysis of syn- and anti-BPDE by DNA. Typical Michaelis-Menten saturation curves were used to extract apparent rate constants in the presence of DNA. To illustrate the utility of the acidic domains concept in DNA-mediated acid hydrolysis, we have modeled both the intercalative and associative (outside-binding) reaction modes of specific BPDE diastereomers with DNA by combining Poisson-Boltzmann (PB), Monte Carlo, and molecular mechanics techniques. First, the hydronium ion concentration near a B-form model of DNA was obtained by using a previously described PB method for mapping the electrostatic potential in counterion volume elements surrounding DNA. Second, the Metropolis Monte Carlo procedure was used to find the equilibrium distribution of BPDE externally associated with DNA in the PB-calculated electrostatic potential. Finally, the intercalation of a single BPDE molecule into the base-pair stack of DNA was modeled by using AMBER to generate conformations of possible intercalation complexes and then using the PB method to map [H+] near the suspected binding sites. Hydrolysis rates for the intercalated molecules were subsequently calculated and combined with external hydrolysis rates to yield total rate curves for the hydrolysis of BPDE as a function of DNA concentration and bulk pH. These curves were then compared with available experimental results. Agreement with experiment was very good in all cases. While use of experimentally derived rate constants practically guarantees reasonable agreement, the calculated hydrolysis rates can be dissected to provide separate intercalated and associated rates and to see how they are individually affected by bulk pH. Furthermore, the application of the acidic domains model to DNA-catalyzed BPDE hydrolysis points out that the measured apparent rate constant for acid-catalyzed hydrolysis can be expressed as the product of the DNA-independent rate constant and the local increase in the hydronium ion concentration over the bulk value.