Engineering Distant Allosteric Networks To Remodel the Substrate Access Channel for Efficient β-Arbutin Biosynthesis

Efficient production of β-arbutin via whole-cell biocatalysis is constrained by glycosyltransferase catalytic performance. To overcome this, we developed a global dynamic design strategy moving beyond conventional active-site engineering. Integrating dynamic cross-correlation matrix analysis with computational screening, we targeted distal flexible residues, identifying three candidates. The optimal variant, G35L, exhibited substantially improved enzyme kinetics and thermal stability. Supported by principal component analysis, free energy landscapes, and double mutant cycles, mechanistic elucidation revealed that this distal mutation functionally couples with the active site. It stabilizes the transition state via enhanced hydrogen bonding and restructures the substrate tunnel to alleviate steric bottlenecks. Under optimized conditions, the biocatalyst achieved a β-arbutin titer of 8.99 g/L (a 4.14-fold improvement). This establishes distal dynamic network targeting as a powerful paradigm for biocatalyst optimization.