Discovering natural product biosynthetic pathways from medicinal plants is challenging and laborious, largely due to the complexity of the transcriptomics-driven pathway prediction process. Here we developed a novel approach that captures the protein-level connections between enzymes for pathway discovery with improved accuracy. We proved that heterologous protein-protein interaction screening in yeast enabled the efficient discovery of both dynamic plant enzyme complexes and the pathways they organize. This approach discovered complexes and pathways in the monoterpene indole alkaloid metabolism of a medicinal plant, kratom with high success rate. Screening using a strictosidine β-D-glucosidase (MsSGD1) against 19 medium-chain dehydrogenase/reductases (MsMDRs) identified five MsSGD1-MsMDR complexes. Three out of the five interacting MsMDRs were then proven functional, while the remaining 14 non-interacting candidates did not show obvious activities. The work discovered three branched pathways by combining transcriptomics, metabolomics, and heterologous PPI screening and demonstrated a new plant pathway discovery strategy.

Discovering dynamic plant enzyme complexes in yeast for novel alkaloid pathway identification from a medicinal plant kratom