Theoretical comparison of p-nitrophenyl phosphate and sulfate hydrolysis in aqueous solution: implications for enzyme-catalyzed sulfuryl transfer.

Both phosphoryl and sulfuryl transfers are ubiquitous in biology, being involved in a wide range of processes, ranging from cell division to apoptosis. Additionally, it is becoming increasingly clear that enzymes that can catalyze phosphoryl transfer can often cross-catalyze sulfuryl transfer (and vice versa). However, while there have been extensive experimental and theoretical studies performed on phosphoryl transfer, the body of available research on sulfuryl transfer is comparatively much smaller. The present work presents a direct theoretical comparison of p-nitrophenyl phosphate and sulfate monoester hydrolysis, both of which are considered prototype systems for probing phosphoryl and sulfuryl transfer, respectively. Specifically, free energy surfaces have been generated using density functional theory, by initial geometry optimization in PCM using the 6-31+G* basis set and the B3LYP density functional, followed by single-point calculations using the larger 6-311+G** basis set and the COSMO continuum model. The resulting surfaces have been then used to identify the relevant transition states, either by further unconstrained geometry optimization or from the surface itself where possible. Additionally, configurational entropies were evaluated using a combination of the quasiharmonic approximation and the restraint release approach and added to the calculated activation barriers as a correction. Finally, the overall activation entropy was estimated by approximating the solvent contribution to the total activation entropy using the Langevin dipoles solvation model. We have reproduced both the experimentally observed activation barriers and the observed trend in the activation entropies with reasonable accuracy, as well as providing a comparison of calculated and observed (15)N and (18)O kinetic isotope effects. We demonstrate that, counterintuitively, the hydrolysis of the p-nitrophenyl sulfate proceeds through a more expansive pathway than its phosphate analogue. Additionally, we show that the solvation effects upon moving from the ground state to the transition state are quite different for both reactions, suggesting that the enzymes that catalyze these reactions would need active sites with quite different electrostatic preorganization for the efficient catalysis of either reaction (despite which many enzymes can catalyze both phosphoryl and sulfuryl transfer). We believe that such a comparative study is an important foundation for understanding the molecular basis for phosphate-sulfate cross-promiscuity within members of the alkaline phosphatase superfamily.