Cytochrome P450 enzymes (P450s) play essential roles in the biosynthetic pathway of brassinosteroids (BRs), a class of plant hormones that regulate various aspects of growth and development. Among them, CYP90B1 functions as the first and rate-limiting enzyme in BR biosynthesis. In 2014, we reported several crystal structures of CYP90B1 and elucidated its stereoselective hydroxylation mechanism responsible for the formation of 22(S)-hydroxycampesterol.1 The biological activity of BRs is significantly enhanced during the final two enzymatic steps, both catalyzed by members of the CYP85A family.2,3 The initial CYP85A-mediated transformation involves two successive hydroxylations of 6-deoxocastasterone (6dCS) by compound I, followed by spontaneous dehydration to yield castasterone (CS). Subsequently, a Baeyer-Villiger oxidation introduces an oxygen atom between C6 and C7 of CS, converting it into brassinolide—the most potent natural BR identified to date. Although the involvement of a peroxo-iron species has been suggested, direct experimental validation has been lacking, and how this enzyme performs two distinct oxygenation reactions remains to be elucidated.
To better understand how this enzyme works, we determined the crystal structure of CYP85A3 in complex with the inhibitor uniconazole. Interestingly, the overall fold of CYP85A3 closely resembled that of cholesterol-bound CYP90B1, with an RMSD of 1.79 Å. We also performed substrate (6dCS and CS) docking simulations using the uniconazole-bound crystal structure of CYP85A3, and the potential substrate docking conformations can explain the two-successive hydroxylation of 6dCS by compound I and the Beyer-Villiger oxidation of CS using peroxo-heme as the oxidizing species. These substrate-binding models are consistent with our alanine-scanning mutagenesis results, which identified key residues likely involved in substrate recognition and positioning.