Structure-Guided Engineering of Prenyltransferase NphB for High-Yield and Regioselective Cannabinoid Production

Abstract

Synthetic biology efforts for cannabinoid research have seen a rapid expansion in recent years. This is in response to the increasing awareness and legalization of the secondary metabolites from Cannabis sativa, dubbed the green rush. In transgenic synthetic biology applications, NphB is a promiscuous prenyltransferase from Streptomyces sp. often used as a replacement in the prenylation step producing the cannabinoid cannabigerolic acid (CBGA), the key precursor to many other cannabinoids. However, its application as a CBGA synthase replacement is limited by its nonspecific regioselectivity in producing a side product along with CBGA. Herein, we demonstrated a detailed and extensive computational structure-guided approach in identifying target residues of mutation for engineering NphB for optimal CBGA production. Our comprehensive computational workflow has led to the discovery of several highly regiospecific variants that produce CBGA exclusively, with the best-performing V49W/Y288P variant having a 13.6-fold yield improvement, outperforming all previous work on NphB enzyme engineering. We subsequently investigated the effects of these mutations by X-ray crystallographic studies of the mutant variants and performed molecular dynamics simulations to uncover an interplay of a H-bonding network and an optimal ligand orientation that favors the CBGA production over the side product. Collectively, this study not only recapitulates the utility of computational tools in informing and accelerating experimental design but also contributes to a better understanding of molecular mechanisms that govern enzyme regioselectivity and readily aids in cannabinoid synthetic biology production for future research into maximizing their therapeutic potential.