In this project, a novel ReaxFF force field, ProtReaxFF, has been developed, optimized and applied to enzyme catalysis reactions.
Enzymes are biological nano-machines that catalyze almost all chemical reactions within biological cells. It is widely accepted, that enzyme catalysis can be applied in a large number of industrial contexts, spurring a green chemistry revolution, as enzymes are obtained from renewable resources, are biodegradable, and are generally non-toxic. Further, their selectivity is a great advantage in chemical industry batch processes because it will require far less work to refine the end product. However, the complexity of enzymatic reactions constitutes a problem, since it is difficult to model and predict the outcome of novel enzyme-catalyzed reactions by the same methods that are applied to traditional forms of catalysis. This thesis presents a tool for overcoming this barrier.
Ideally, techniques based on Quantum Mechanics (QM) should be applied to enzymatic reactions, since QM applies to conditions at the atomic level. However, for a biological system this would involve far too many variables for our present computer power to handle. One of the major problems in relation to proteins is calculating the electronic energy for a given nuclear configuration to give a potential energy surface.
The project builds on efforts by the Goddard group at the California Institute of Technology. The group has developed a simplified model, the ReaxFF (reactive force-field), which bridges the gap between quantum chemistry methods and the ordinary force-fields of classical molecular mechanics methods. Thereby, chemical reactions can be described properly with bond-forming and bond-breaking events during the simulation time.
In this project, a novel ReaxFF force field, ProtReaxFF, has been developed, optimized, and applied to enzyme catalysis reactions. It is shown that the current version of ProtReaxFF can be used to perform molecular dynamics simulations of biomolecules. The developed method is complementary to the more well-known QM/MM method, the inventors of which were awarded the Nobel Prize in chemistry 2013. Although the preliminary results using the developed ProtReaxFF are encouraging additional work is needed for the method to fully mature as a computational tool.