Enzyme functionality is a critical component of all life systems. Whereas advances in experimental methodology have enabled a better understanding of factors that control enzyme function, critical components of the reaction space such as highly unstable intermediates and transition states are best accessed for evaluation through computational simulations. Similarly, computational methodology continues to provide a key resource for probing excited-state processes such as bioluminescence. Combined ab initio quantum mechanical molecular mechanical (ai-QM/MM) simulations are, in principle, the preferred choice in the modeling of both processes. But ai-QM/MM modeling of enzymatic reactions is now severely limited by its computational cost, where a direct ai-QM/MM free energy simulation of an enzymatic reaction can take 500,000 or more CPU hours. Meanwhile, ai-QM/MM modeling of firefly bioluminescence is also hindered by the computational accuracy, where it has yet to produce quantitatively correct predictions for the bioluminescence spectral shift with site-directed mutagenesis. The goal of this proposal is to accelerate ai-QM/MM simulations of enzymatic reaction free energy and to improve the quality of ai-QM/MM-simulated bioluminescence spectra, so that ai-QM/MM simulations can be routinely performed by experimental groups. This will be achieved via a) using a lower-level (semi-empirical QM/MM) Hamiltonian for sampling; b) an enhancement to the similarity between the two Hamiltonians by calibrating the low-level Hamiltonian using the reaction pathway force matching approach, in conjunction with several other methods. The expected outcomes of this collaborative effort include: a) advanced methodologies for accelerated reaction free energy simulations and accurate bioluminescence spectra predictions, which will be released through multiple software platforms; b) a fundamental understanding of reactions such as Kemp elimination and polymerase-eta catalyzed DNA replication; c) a deeper insight into the role of macromolecular environment in the modulation of enzyme catalytic activities or bioluminescence wavelengths, which can further enhance our capability of designing new enzymes and bioluminescence probes.
This project aims to develop quantum-mechanics-based computational methods to more quickly model enzymatic reactions and more accurately model bioluminescence spectra. It will lead to reliable and efficient computational tools for use by the general scientific community. It will facilitate the probe of enzymatic reaction mechanisms and the computer-aided design of new bioluminescence probes.
