Solvent and protein effects on the vibrational frequency shift and energy relaxation of the azide ligand in carbonic anhydrase

Combined QM/MM molecular dynamics simulations have been carried out to investigate the vibrational frequency shift for the azide antisymmetric stretch mode induced by aqueous solvation and by carbonic anhydrase II. In this work the oscillator and the enzyme active site is treated explicitly by quantum mechanics. Thus, the dynamical change of the potential energy surface of the oscillator can be adequately represented. We found that although the average geometry of the azide ion is symmetric in aqueous solution, the instantaneous solute-solvent interactions induce localization of the resonance structure having triple bond character, leading to a blue shift in the observed antisymmetric vibrational frequency in polar solvents. The computed frequency shift of azide ion from water to the active site of carbonic anhydrase is 56 cm ^-1, in good accord with the experimental value of 51 cm ^-1. Analyses of the computational results demonstrate that the origin of the protein-induced blue shift is due to a combination of ligand binding to the zinc ion and protein dynamical interactions. The former makes the dominant contribution by stabilizing the N=N ^--N ^2- ionic state through ligand-metal coordination, whereas the latter attenuates the ligand-metal bonding interactions, recovering some of the N ^-=N ⁺=N ^- valence bond character. Furthermore, the vibrational energy relaxation time has been determined both in water and in the enzyme. Intramolecular vibrational redistribution provides the main doorway for energy relaxation of the azide antisymmetric stretch in the enzyme, which is consistent with that obtained previously for an azide ion in water by Morita and Kato. Following this mechanism coupled with the quantum correction factor suggested by Skinner and Park, we obtained vibrational relaxation times of about 2 ps in water and 6 ps in carbonic anhydrase. Importantly, the change in relaxation time by a factor of 2.5 from water to the enzyme active site is in good accord with experiment.