Structure-Agnostic Bioactivity-Driven Combinatorial Biosynthesis Reveals New Antidiabetic and Anticancer Triterpenoids

Although combinatorial biosynthesis can dramatically expand the chemical structures of bioactive natural products to identify molecules with improved characteristics, progress in this direction has been hampered by the difficulty in isolating and characterizing the numerous produced compounds. This challenge could be overcome with improved designs that enable the analysis of the bioactivity of the produced metabolites ahead of the time-consuming isolation procedures. Herein, we showcase a structure-agnostic bioactivity-driven combinatorial biosynthesis workflow that introduces bioactivity assessment as a selection-driving force to guide iterative combinatorial biosynthesis rounds towards enzyme combinations with increasing bioactivity. We apply this approach to produce triterpenoids with potent bioactivity against PTP1B, a promising molecular target for diabetes and cancer treatment. We demonstrate that the bioactivity-guided workflow can expedite the combinatorial process by enabling the narrowing down of more than 1000 possible combinations to only five highly potent candidates. By focusing the isolation and structural elucidation effort on only these five strains, we reveal 20 structurally diverse triterpenoids, including four new compounds and a novel triterpenoid-anthranilic acid hybrid, as potent PTP1B inhibitors. This workflow expedites hit identification by combinatorial biosynthesis and is applicable to many other types of bioactive natural products, therefore providing a strategy for accelerated drug discovery.