Enzyme-Catalyzed Phosphoryl Transfer Reactions: Energetics of Transition States and Vanadate Analogues in Aqueous Solution
Organic phosphate compounds are a key component of biochemical processes.
The transfer of phosphoryl groups is a ubiquitous process in living organisms that drives key cellular activities such as glycolysis and DNA replication. Vanadate is recognized as an effective structural and electronic analogue of phosphate. Solvation effects are significant contributors to the reaction free energies and activation barriers of organic phosphate hydrolysis reactions. The methanolysis of dimethyl phosphate (DMP) and dimethyl vanadate (DMV) as well as acetyl phosphate (AcP) and acetyl vanadate (AcV) are modeled in aqueous solution to address the use of Vanadium(V)-based transition state analogues in the study of DNA Polymerase, β-phosphoglucomutase (β-PGM), and Hexose Phosphate Phosphatase (HPP) -catalyzed phosphoryl transfer reactions. This work seeks to add to the growing body of theoretical work that addresses the cleavage of P-O bonds by offering additional suggestions for the use of vanadium-based transition state analogues to study enzymes that catalyze phosphoryl transfer reactions.
The structures and free energies of phosphate dimethyl ester and vanadate
dimethyl ester have been calculated using the B3LYP/TZVP density functional quantum chemical methods, and the polarized continuum (PCM) and Langevin dipoles (LD) solvation models. These calculations were carried out to obtain fundamental information on the ability of vanadate esters to function as transition state analogs for the nucleotidyl transfer reaction catalyzed by DNA polymerases. Base catalyzed methanolysis of the phosphate and vanadate dimethyl esters were the model reactions examined in this study. The structures of the phosphate and vanadate dimethyl esters and pentavalent intermediates in aqueous solution were optimized and evaluated at the PCM/B3LYP/TZVP level. The three-dimensional free energy surfaces for the studied reactions were determined at the PCM/HF/6-31G*//B3LYP/TZVP level. Comparison with experimental structural data obtained from the Cambridge Structural database and with the observed kinetics of phosphate diester hydrolysis demonstrated that the level of theory chosen for these studies was appropriate. The results showed that structurally and electrostatically the vanadate dimethylester and a five-coordinate near trigonal bipyramidal intermediate were reasonable analogs for the parent phosphorus systems. Despite these similarities in structure, the energetics of the two systems were different
and the transition states of the two model reactions were found on different areas of the potential energy surface with barriers of activation of 45 and 31 kcal/mol, respectively, for the methanolysis of the phosphate and vanadate dimethyl esters in pH 7 solution. The total free energies of the pentavalent intermediates of the phosphate and vanadate compounds were found to be 42 and 18 kcal/mol above the ground state energies, respectively. The pentavalent vanadate intermediate is more stable than the pentavalent phosphate intermediate. When extrapolating the binding energy of a transition state-DNA polymerase complex to a transition state analogue-DNA polymerase complex, the binding of a vanadate-ester nucleotide adduct, pentacoordinated by monodentate ligands, was not found to be favorable. This finding suggests that additional stabilization of this adduct is needed before this type of transition state-analogue will be likely to yield stable adducts with this class of enzymes. New possible candidates for such complexes are suggested.
The B3LYP/TZVP and PCM/B3LYP/6-31G* and LD methods were used to
model the structural, equilibrium and kinetic properties of the methanolysis of acetyl
phosphate (AcP) and acetyl vanadate (AcV) in aqueous solution. The calculated
equilibrium bond distance and force constant of the bond formed between phosphorus
and the bridging oxygen atom (P-Ob) of AcP were compared to experimental IR data. Evaluation of the accuracy of the PCM and LD computational solvation models via experimental pKa constants are presented. Reaction free energies related to AcP methanolysis in solution were calculated using the PCM and LD models and correctly predict the experimentally observed spontaneity of these reactions. The LD model was modified to accept large data sets as input and used in combination with the B3LYP/TZVP method to calculate solvation free energy surfaces for the methanolysis of AcP and AcV in aqueous solution. Calculated ESP charge surfaces for the reactions involving the methoxide attack on AcP and AcV were used to contrast the change in electronic density that occurs along the reaction mechanism pathways. The free energy surface for the reaction involving methanol attack on AcP was calculated to model the reaction for the hydrolysis of AcP in pH-neutral solution. Intermolecular proton transfer was addressed by calculating the free energy barriers for proton transfer at each point along the reaction mechanism pathway. When proton transfer was considered, the calculated free energy of activation was in good agreement with the experimental rate constant for AcP hydrolysis under standard conditions. The three-dimensional free energy surface for the methoxide attack on Mg2+[AcP2-] was calculated to model the effects of Mg2+ on the reaction mechanism for AcP hydrolysis. The results hint that AcP hydrolysis may proceed through a more compact transition state in the presence of Mg2+.
The crystal structure of hexose phosphate phosphatase (HPP) contains a vanadate-based transition state analogue (TSA) in its active site. Using our previously established methods, a structure located on the AcV methanolysis free energy surface was used to calculate the stability of a HPP-vanadate complex in solution and correctly predicted its existence. The controversial assignment of the electronic density map located in the active site of β-phosphoglucomutase (β-PGM) to a stable phosphorane intermediate was addressed by calculating the stability of a β-PGM-phosphorane complex in solution using structures from the AcP methanolysis free energy surfaces. However, our calculations do not support the assignment of the electronic density map of the β-PGM crystal structure to a phosphorane intermediate. This conclusion was further explored by comparing our structures to experimental x-ray data.