Computational Enzyme Engineering: A Novel CSFp Strategy for Optimizing D-Lac Activity

D-Lac holds significant potential for the optical resolution of racemic pantolatone (DL-PL), leading to the production of D-pantothenic acid, a crucial component of the vitamin B5 family (Scheme S1). Vitamin B5 and its derivatives find widespread applications in the food, pharmaceutical, animal feed, cosmetic, and other industries,52 as they enable acetyl-CoA to synthesize the neurotransmitter acetylcholine, which has attracted significant research attention in the field of neurodegenerative diseases.53

However, the evolution of D-Lac has posed a considerable challenge. Initial efforts focused on error-prone (PCR) and DNA shuffling54 techniques, which lacked a rational design framework due to the absence of crystal structure information for D-Lac. The catalytic mechanism of D-Lac remains elusive, and designing enzymes based on quantum mechanics-derived transition states is not yet feasible. Furthermore, the substrate D-pantolactone lacks substantial or polar groups, and a single amino acid modification in the active pocket results in poor complementarity between the active site and the ligand's overall shape.

To address these challenges, a novel computational design pipeline has been developed, leading to the discovery of combinatorial strategies for ultra-low throughput screening. The proposed CSFp strategy provides a robust, generic solution for minimizing experimental effort while maximizing the exploration of supernumerary effects in terms of additivity and/or synergy between mutant sets. Through this approach, a promising mutant (N96S/A271E/F274Y/F308G) was obtained, demonstrating a 56-fold increase in activity towards D-PL compared to wild-type (WT) D-Lac.

This experimental example underscores the potential of the CSFp strategy to facilitate optimized computational enzyme engineering, rescuing enzymes lacking crystal structure information and complex catalytic mechanisms, and enabling its application in scenarios where enhanced enzyme activity is essential. Complementing existing enzyme engineering methods, this approach offers valuable insights for computational enzyme modification.