Terpene synthases including cyclases

The terpene synthases are structurally and mechanistically very closely related to the IPPS. A number of these enzymes have been cloned and overexpressed from plant sources (Chappell 1995) thale cress (Arabidopsis thaliana), whose genome is now fully sequenced, has 32 functional terpene synthase DNA sequences in its genome (Aubourg et al. 2002). Conserved amino acid sequences, particularly DDxxD motifs for binding of Mg + and pyrophosphate in the substrates, are common and allow terpene synthases from a wide variety of sources to be identified by sequence similarity searches (Bohlmann et al. 1998). Sequence analysis of the genes which encode known terpene synthases in plants has enabled them to be grouped into six subfamilies whose members each share 40% sequence identity. 

The major difference between the IPPSs and the terpene cyclases lies in the fact that terpene cyclases do not bind IPP. The first step in the terpene cyclase-catalysed reaction is still in most cases the loss of the pyrophosphate group with the formation of a polyprenyl cation in the active site. This time, however, it is an electron-rich double bond elsewhere in the molecule which serves as an internal nucleophile, with the result being formation of a cychc structure. The active sites of terpene cyclase enzymes are tailored to fold the polyprenyl pyrophosphates into the optimum conformation for intramolecular attack to take place, with hydrophobic residues to force the prenyl chain into the desired conformation, and aromatic residues such as tyrosine to stabilize the positive charge on the intermediate carbocation (Starks etal.1997). Initial isomerization, for example to linalyl or nerohdyl pyrophosphate (vide infra), is often important to present the correct geometry at the electron-accepting end of the molecule to allow cychzation to occur (Bohhnann et al. 1998). 

As with the IPPSs, site-directed mutagenesis of key active site residues can alter the folding of the substrate, with the result that alternative cychzation paths can be foUowed, and novel products obtained (Rising et al. 2000). 

---

Terpene synthases, also known as terpene cyclases because most of their products are cyclic, utilize a carbocationic reaction mechanism very similar to that employed by the prenyltransferases. Numerous experiments with inhibitors, substrate analogues and chemical model systems (Croteau, 1987 Cane, 1990, 1998) have revealed that the reaction usually begins with the divalent metal ion-assisted cleavage of the diphosphate moiety (Fig. 5.6). The resulting allylic carbocation may then cyclize by addition of the resonance-stabilized cationic centre to one of the other carbon-carbon double bonds in the substrate. The cyclization is followed by a series of rearrangements that may include hydride shifts, alkyl shifts, deprotonation, reprotonation and additional cyclizations, all mediated through enzyme-bound carbocationic intermed iates. The reaction cascade terminates by deprotonation of the cation to an olefin or capture by a nucleophile, such as water. Since the native substrates of terpene synthases are all configured with trans (E) double bonds, they are unable to cyclize directly to many of the carbon skeletons found in nature. In such cases, the cyclization process is preceded by isomerization of the initial carbocation to an intermediate capable of cyclization. 

As presented in this chapter, today, much is known about the process of terpene biosynthesis. The accumulated knowledge includes a detailed picture about the biosynthesis of the terpenoid monomers IPP and DMAPP either via the mevalonate or the DXP route and their interconversion by isomerases. Also, the stereochemical courses and enzyme mechanisms of all transformations have been largely elucidated. Especially the recently obtained structural data of prenyltransferases and various kinds of terpene synthases resulted in an evolutionary model that involves six domains (a, P, 7,8, e, and Q for the biosynthesis of linear polyisoprenoids from IPP and DMAPP and their subsequent transformation into (poly)cyclic terpenes. All these insights may open up new chances in controlling terpene biosynthesis, e.g., by directed evolution of terpene cyclases or domain swaps in multi-domain enzymes for the production of new terpenes, reconstitution of terpene biosynthetic pathways in heterologous hosts for production optimization, or targeted inhibitirm of pathways in pathogens for disease control. 

---