Understanding the substrate selectivity and the product regioselectivity of Orf2-catalyzed aromatic prenylations.

Orf2, a recently identified prenyltransferase of aromatic natural products, displays relaxed substrate selectivity and interesting product regioselectivity. This gives rise to the opportunity to engineer the active site to tune the functionality of terpenoids for therapeutic applications. The structural basis of substrate binding has been determined, but the source of the observed substrate selectivity and product regioselectivity cannot be completely understood on the basis of the static picture that the crystal structures of Orf2 and its complexes afford. The electron density and B-factors of the substrates, particularly those of 1,6-dihydroxynaphthalene, suggest significant conformational fluctuation in the Orf2 binding site. We thoroughly explored the binding of 1,6-dihydroxynaphthalene and quantitatively evaluated the relative free energies of three binding states that we identified in terms of a two-dimensional potential of mean force. The available experimental orientation, which gives the major prenylated product of 1,6-dihydroxynaphthalene, corresponds to the global free energy minimum. Two alternative binding states were identified on the calculated free energy surface, and both are readily accessible at 300 K. The alternative binding conformations were extracted from the potential of mean force calculation and were subjected to further validation against the experimental X-ray diffraction data using a refinement protocol supplemented with a hybrid quantum mechanical and molecular mechanical energy function. The agreement was excellent as indicated by the R and Rfree factors that were comparable to that obtained for the published orientation using a similar protocol. These binding states are the origin of the selectivity and regioselectivity in Orf2-catalyzed aromatic prenylations. Our analyses also suggest that Ser214 and Tyr288, forming hydrogen bonds with the alternative binding states of 1,6-dihydroxynaphthalene and flaviolin, are good candidates for site-directed mutagenesis, and changing them to, for example, their hydrophobic counterparts would affect the substrate selectivity and product regioselectivity.