Phosphodiester hydrolysis has been the subject of intense study due to its importance in biology. Despite the numerous significant analyses of phosphodiester cleavage mechansim, comparatively little is known about the nucleophiles in these reactions. To determine whether hydroxide acts as a nucleophile or a general base in the hydrolysis of thymidine-5'-p-nitrophenyl phosphate,we determined solvent deuterium isotope effects (D2Ok), ionic strength effects, and 18O isotope effects on the solvent nucleophile (18knuc). The D2Ok for hydroxide-catalyzed phosphodiester hydrolysis is slightly inverse (0.9 +/- 0.1), suggesting that a proton transfer does not occur in the transition state. A significant alpha effect is observed with hydroperoxide, demonstrating that oxyanions can act as nucleophiles in the reaction. Additionally, the ionic strength dependencies of hydroxide and hydroperoxide catalysis are indistinguishable, suggesting that they act by the same mechanism. Finally, the 18knuc for the hydroxide-catalyzed reaction is 1.068 +/- 0.007, well in excess of the equilibrium 18O isotope effect between water and hydroxide (1.040 +/- 0.003). Together, the data are most consistent with direct nucleophilic attack by hydroxide. From the observed 18knuc and the known equilibrium component, the kinetic component of the isotope effect was calculated to be 1.027 +/- 0.010. This large kinetic component suggests that little bond order to the nucleophile occurs in the transition state.

Heavy atom isotope effects are a valuable tool for probing chemical and enzymatic reaction mechanisms; yet, they are not widely applied to examine mechanisms of nucleophilic activation. We developed approaches for analyzing solvent (18)O nucleophile isotope effects ((18)k(nuc)) that allow, for the first time, their application to hydrolysis reactions of nucleotides and nucleic acids. Here, we report (18)k(nuc) for phosphodiester hydrolysis catalyzed by Mg(2+) and by the Mg(2+)-dependent RNase P ribozyme and deamination by the Zn(2+)-dependent protein enzyme adenosine deaminase (ADA). Because ADA incorporates a single solvent molecule into the product inosine, this reaction can be used to monitor solvent (18)O/(16)O ratios in complex reaction mixtures. This approach, combined with new methods for analysis of isotope ratios of nucleotide phosphates by whole molecule mass spectrometry, permitted determination of (18)k(nuc) for hydrolysis of thymidine 5'-p-nitrophenyl phosphate and RNA cleavage by the RNase P ribozyme. For ADA, an inverse (18)k(nuc) of 0.986 +/- 0.001 is observed, reflecting coordination of the nucleophile by an active site Zn(2+) ion and a stepwise mechanism. In contrast, the observed (18)k(nuc) for phosphodiester reactions were normal: 1.027 +/- 0.013 and 1.030 +/- 0.012 for the Mg(2+)- and ribozyme-catalyzed reactions, respectively. Such normal effects indicate that nucleophilic attack occurs in the rate-limiting step for these reactions, consistent with concerted mechanisms. However, these magnitudes are significantly less than the (18)k(nuc) observed for nucleophilic attack by hydroxide (1.068 +/- 0.007), indicating a "stiffer" bonding environment for the nucleophile in the transition state. Kinetic analysis of the Mg(2+)-catalyzed reaction indicates that a Mg(2+)-hydroxide complex is the catalytic species; thus, the lower (18)k(nuc), in large part, reflects direct metal ion coordination of the nucleophilic oxygen. A similar value for the RNase P ribozyme catalyzed reaction provides support for nucleophilic activation by metal ion catalysis.

No abstract