Harnessing natural evolution and computation towards systems enzymology.

Abstract

Rationally reprogramming enzyme catalysis requires systems-level knowledge of various enzyme mutations, which is extremely challenging. Learning from nature is inescapable to overcome this barrier. We recently distilled the evolutionary information from natural homologous sequences using a maximum-entropy model and established a connection between enzyme evolution and enzyme catalysis. The finding also provides a rational enzyme engineering approach, and about half of the predicted mutations improve enzyme catalytic power in experiment. Furthermore, we utilized natural evolution to systematically rationalize laboratory evolution of designer enzymes; we identified the sequence determinants for the selectivity of kinase covalent inhibitors and confirmed the insight using enzyme modeling. Overall, our studies show the availability of vast protein sequences from nature is promising to advance enzymology to a systems-level, an emerging field termed ‘systems enzymology.’