Harnessing plant metabolic diversity for human health

Plants make a vast array of chemicals that have been honed by evolution to be bioactives. If the instruction manual needed to make this chemical diversity could be decoded, this would unlock unprecedented opportunities to understand plant natural product biosynthesis, function, and mechanisms of metabolic diversification, and to harness this biosynthetic capability for medicinal applications. We have developed a powerful computational platform and rapid transient plant expression technology that now uniquely position us to address this challenge. We will focus on a large and structurally complex groups of plant natural products, the pharmaceutically important triterpenoids, as our exemplar. By decoding the genetic potential of the Plant Kingdom to make and diversify these compounds, we will gain a comprehensive understanding of enzyme specificity and function, and of the rules governing sequential modification of scaffolds by tailoring enzymes, enabling us to control and direct biosynthesis. We will develop machine learning-based approaches for predicting bioactivity that will enable hypotheses about structure-activity relationships to be iteratively refined and tested, and design molecules with optimised bioactive properties using engineering biology approaches. This project will deliver a step-change in our ability to access, harness and engineer new enzymes, pathways and chemistries with potential therapeutic applications.