Biocatalysis based on enzyme catalysis is becoming increasingly popular. However, few enzymes can directly catalyze the desired reaction, it is necessary to modify the catalyst to meet the practical needs. Virtual screening combining with molecular dynamics simulation and wet experiments is an attractive approach for obtaining target proteins. Here, based on the traditional conformational dynamics analysis method, we proposed an iterative strategy to quickly screen for nucleoside kinase variants with increased activity. Guided by the simulation results, we obtained a total of 21 candidate residues, of which 5 were selected for in-depth analysis, resulting in two key residues (A220 and A248) and several variants. All variants were evaluated experimentally and computationally; A220L showed a 7.3-fold increase in activity compared to the wild-type enzyme, whereas A248S showed a 3.5-fold increase in the kcat/Km ratio. Using an in vitro dual-enzyme coupling assay, a 400 mM cytidine substrate could be almost completely converted to cytidine 5ʹ-monophosphate within 5 h. Finally, we modified another enzyme in the coupling reaction under the guidance of the same strategy to change its substrate preference, thus successfully halving the requirement of one substrate. Our method can simplify the protein engineering process by assisting the efficient and accurate identification of key residues, thereby expanding the scope of biocatalyst applications.
Read full abstract