STRUCTURAL STUDY OF SESQUITERPENE CYCLASES. Starks, C.^*, Back, K.⁺, Chappell, J.⁺, Noel, J.^{*, *}Structural Biology Laboratory, Salk Institute, La Jolla, CA 92037, ⁺Dept. of Agronomy, University of Kentucky, Lexington, KY 40546

Cyclic terpenoids are found throughout nature and comprise a medicinally important class of compounds from plants. The biosynthesis of cyclic terpenes is determined by branch point enzymes referred to as terpene cyclases. The objective of our research program is to understand the structural, functional, and chemical features governing two distinct stereochemically controlled cyclizations of farnesyl diphosphate (FPP) catalyzed by two homologous plant sesquiterpene cyclases that result in two unique bicyclic products. Our current crystallographic and enzymatic studies will provide the foundation for our long range goal that focuses on a rationally and combinatorially based redesign of terpene cyclases for the enzymatically directed syntheses of pharmaceutically important terpenoids. This structure/function analysis should also deepen our understanding of the biosynthesis of the larger class of essential terpenoids including cholesterol, steroid hormones, and lipid soluble vitamins.

Our experimental system encompasses two homologous sesquiterpene cyclases (77% amino acid identity), 5-epi-aristolochene synthase from N. tabacum (TEAS) and vetispiradiene synthase from H. muticus (HVS), which cyclize FPP into products which are structurally quite different. Chemical rationalization of the reaction mechanisms suggests several partial reactions common to both enzymes and at least one final step unique to each. We have grown crystals of TEAS and an active TEAS/HVS chimera in the space group P4₁2₁2 (65% solvent). We have obtained a native chimera data set to 2.8Å and four derivative data sets resulting in our initial 2.8Å MIRAS map. Solvent flattening greatly improved the quality of the initial electron density map; solvent boundaries, connectivity, and secondary structural features are readily apparent. We have begun model building. Once the chimera structure is interpreted and refined, it should provide us with a search model for the structural determination of the wild type enzymes as well as other chemically interesting chimeras. By comparing the structures of these enzymes and their complexes with products and inhibitors we should gain insight into the particular active site residues and surfaces responsible for common and specific reactions among the cyclases.