A structural and data-driven approach to engineering a plant cytochrome P450 enzyme

Functional manipulation of biosynthetic enzymes such as cytochrome P450s (or P450s) has attracted great interest in metabolic engineering of plant natural products. Cucurbitacins and mogrosides are plant triterpenoids that share the same backbone but display contrasting bioactivities. This structural and functional diversity of the two metabolites can be manipulated by engineering P450s. However, the functional redesign of P450s through directed evolution (DE) or structure-guided protein engineering is time consuming and challenging, often because of a lack of high-throughput screening methods and crystal structures of P450s. In this study, we used an integrated approach combining computational protein design, evolutionary information, and experimental data-driven optimization to alter the substrate specificity of a multifunctional P450 (CYP87D20) from cucumber. After three rounds of iterative design and evaluation of 96 protein variants, CYP87D20, which is involved in the cucurbitacin C biosynthetic pathway, was successfully transformed into a P450 mono-oxygenase that performs a single specific hydroxylation at C11 of cucurbitadienol. This integrated P450-engineering approach can be further applied to create a de novo pathway to produce mogrol, the precursor of the natural sweetener mogroside, or to alter the structural diversity of plant triterpenoids by functionally manipulating other P450s.