Probing sesquiterpene synthase chemistry with non-canonical amino acids

With more than 55,000 known compounds serving myriad functions in all forms of life, the terpenoids are the largest as well as the most structurally and stereochemically diverse family of natural products found on earth. Terpenoids are found in marine and terrestrial plants, fungi, bacteria and insects and have many applications in agriculture as antifungals, herbicides or insecticides, as fragrances in cosmetics, or as antibiotics, anti cancer agents, hormones, or contraceptive agents in medicine. Despite their structural diversity, all terpenoids are biosynthesised from two chemically related molecules, namely delta1- and delta2-isopentenyldiphosphate, an observation which was first made by the German chemist Otto Wallach more than 100 years ago (Nobel prize 1910) and put on the correct biochemical foundation by Leopold Ruzicka in Zurich (Nobel prize 1939). The isopentenyldiphosphates are joined in elongation reactions to form the linear precursors of all terpenoids, which are classified according to the number of isopentenyl-units (C5) they contain as monoterpenes (C10), sesquiterpenes (C15), diterpenes (C20) and higher molecular weight terpenoids. The action of terpene synthases then converts these linear precursors into a large number of cyclic products. It is these cyclisations that are responsible to a large part for the diversity seen in terpenoid natural products. Remarkably all these enzymes share a common three-dimensional fold and hence the complexity and diversity of the final structures is contrasted with the simplicity of the generation of these products. In order to decipher the code that determines the specificity of terpene synthases a much deeper understanding of individual reaction mechanisms is required. We have previously used techniques where the substrates were modified or where individual amino acids within the active site of terpene synthases were replaced with others and the effects on the reaction studied. This latter approach is limited by the fact that only 20 different amino acids are normally observed in naturally occurring proteins. We now wish to use a new technique that allows the incorporation into enzymes of non-natural amino acids that can be chosen in such a way as to obtain answers to specific questions about the reaction mechanisms and hence further our understanding of this fascinating group of enzymes. This work should eventually provide us with the knowledge to generate new terpene-like, but non-natural molecules with many potential applications in basic science and industry.

Terpene synthases catalyse complex, multistep reactions that generate thousands of structurally diverse hydrocarbons of biological and commercial importance. Unlike many other biochemical reactions, terpene biosynthesis is essentially carbocationic in nature. Previous work in our lab has revealed many of the mechanistic details of the conversion of FPP by aristolochene synthase (AS), but the key step, the generation of the positively charged eudesmane cation from the uncharged intermediate germacrene A is only poorly understood. Hence the key unresolved question is how this enzyme achieves the protonation of an isolated double bond to generate what appears to be a high-energy carbocationic intermediate. We propose here to use an approach based on the introduction of natural and non-natural amino acids into the active site of AS to test the proposal that AS uses both an appropriately placed general acid and stabilization of the cationic transition state (and product) through cation-pi interaction to protonate the double bond in a conformationally activated ring system. Synthesis and introduction into AS of amino acids that carry non-canonical aromatic sides chains will be used to test the function of cation stabilisation, while fluorination of potential general acids will modulate the protonation step during AS catalysis. Since most sesquiterpene cyclases follow similar mechanisms the results obtained here will shed light on intricate details of most of these enzymes that all rely on a shared three-dimensional fold for activity. Hence this work will help decipher the mechanistic principles used by terpene cyclases and may eventually allow us to engineer the specificity of these enzymes in a rational way to get closer to the ultimate goal, namely the expansion of the terpenome through the design of novel (unnatural) terpenoids with many potential applications as pharmaceuticals or agrochemicals.