Computational Design-Enabled Divergent Modification of Monoterpene Synthases for Terpenoid Hyperproduction

Enzymes' catalytic promiscuity enables the alteration of product specificity via protein engineering; yet, harnessing this promiscuity to achieve desired catalytic reactions remains challenging. Here, we identified HCinS, a monoterpene synthase (MTPS) with a high efficiency and specificity for 1,8-cineole biosynthesis. Quantum mechanics/molecular mechanics (QM/MM) simulations, which were performed based on the resolved crystal structure of HCinS, revealed the mechanistic details of the biosynthetic cascade reactions. Guided by these insights, in silico HCinS variants were designed with fine-tuned transition-state energies and reaction microenvironments. Three variants (T111A, N135H, F236M), each with one amino acid substitution, exhibited high specificity in the production of monocyclic (R)-α-terpineol, (R)-limonene, and acyclic myrcene, respectively, maintaining over 55% efficiency of native HCinS. These designed HCinS variants surpassed naturally evolved isozymes in catalytic capacity and enabled yeast to achieve the highest microbial titer of each corresponding terpene. Furthermore, the single mutation of four functional equivalent amino acids in other four identified TPSs, respectively, resulted in the expected shifts on product specificity as HCinS variants. This research offers insights into the mechanisms controlling the TPS's product promiscuity and highlights the universal applicability of computational design in reshaping the product specificity of TPSs, thereby paving innovative avenues for creating enzymes with applications in chemistry and synthetic biology.