A Novel Semi-biosynthetic Route for Artemisinin Production Using Engineered Substrate-Promiscuous P450 _BM3

Production of fine chemicals from heterologous pathways in microbial hosts is frequently hindered by insufficient knowledge of the native metabolic pathway and its cognate enzymes; often the pathway is unresolved, and the enzymes lack detailed characterization. An alternative paradigm to using native pathways is de novo pathway design using well-characterized, substrate-promiscuous enzymes. We demonstrate this concept using P450(BM3) from Bacillus megaterium. Using a computer model, we illustrate how key P450(BM3) active site mutations enable binding of the non-native substrate amorphadiene. Incorporating these mutations into P450(BM3) enabled the selective oxidation of amorphadiene artemisinic-11S,12-epoxide, at titers of 250 mg L(-1) in E. coli. We also demonstrate high-yielding, selective transformations to dihydroartemisinic acid, the immediate precursor to the high-value antimalarial drug artemisinin.