Accelerating Enzyme Evolution: Computational Design and the CSFp Strategy for Enhanced Biocatalysis

The screening process in directed enzyme evolution is a significant bottleneck, hindering the rapid development of biocatalysts. Computational design, however, holds immense potential to accelerate the utilization of enzymes in biocatalysis, paving the way for more efficient and powerful applications. This study utilizes the calcium-dependent enzyme D-lactonohydrolase, lacking a crystal structure, as a model system to demonstrate the effectiveness of AlphaFold2-predicted structures as reliable starting points for enzyme evolution.

To enhance the catalytic efficiency of D-lactonohydrolase, we devised a novel, intelligent combination strategy termed CSFp, integrating Steered Molecular Dynamics (SMD), Functional Library (FuncLib), Protein Strain, Unsatisfactoriness, and Frustration findER (pSUFER). This multifaceted approach enables the systematic exploration and optimization of enzyme structure and function. The CSFp strategy yielded promising results, culminating in the identification of the N96S mutant, which exhibited a remarkable 56-fold increase in activity towards DL-pantolactone compared to the wild-type, accompanied by an impressive enantioselectivity of 99%.

This study showcases the transformative potential of the CSFp strategy for rescuing enzymes lacking crystal structure information and obscure catalytic mechanisms, thereby facilitating their application in protein engineering where enhanced enzyme activity is paramount. The ability to accelerate enzyme evolution through computational design and intelligent combination strategies opens exciting new avenues for the development of highly efficient biocatalysts for diverse applications.