Abstract
Terpenes are the most diverse class of natural products on the planet, and their diversity arises from a small and versatile toolbox of prenyl diphosphate substrates. These substrates, which increase in five-carbon increments, are cyclized by terpene cyclases to form complicated scaffolds. The Oprian group studies a class of terpene cyclases called sesquiterpene cyclases (SCs), which all utilize the fifteen-carbon substrate farnesyl pyrophosphate (FPP). Select SCs facilitate an anti-Markovnikov (AM) cyclization between carbon 1 and carbon 11 (C1-C11) which places a positive charge on C10, a secondary carbon. This project combines structural and phylogenetic analyses to better understand how these SCs perform an AM cyclization.Pentalenene synthase (PS) is an SC that facilitates an anti-Markvonikov (AM) cyclization of FPP to form the sesquiterpene pentalenene. The initial AM cyclization places a positive charge on C10 (humulyl cation A), and a subsequent hydride shift relocates the positive charge to C9 (humulyl cation B). Analysis of the crystal structure of PS complexed with 12,13-difluorofarnesyl pyrophosphate (DFFPP), an unreactive analog of FPP, revealed that DFFPP adopts a curved, reaction-ready orientation in the active site. Phenylalanine at position 76 (F76) was found to be positioned such that the face of its aromatic side chain was placed directly above C9, the location of the positive charge in humulyl cation B. The proximity of F76 suggests a possible stabilization of the secondary carbocation through π-cation interactions, allowing for the AM cyclization and hydride shift to occur. Indeed, the PS F76A mutant had reduced activity and did not produce pentalenene. However, the mutant was still capable of performing an AM cyclization - its major product was caryophyllene, a sesquiterpene resulting from an AM cyclization and no subsequent hydride shift.
Isohirsutene synthase (IHS), found in the same evolutionary clade as PS and containing an aromatic tyrosine at the analogous position to F76 in PS (Y87), exhibited a similar trend: a Y87A mutation abolished isohirsutene synthesis and produced hedycaryol, a Markovnikov product, as well as two AM products, caryophyllene and humulene. These findings established a common role for aromatic residues at this position in PS and IHS.An evolutionarily distinct SC, caryolan-1-ol synthase (CS), facilitates the AM cyclization of FPP, but not the subsequent hydride shift, forming the sesquiterpenes caryolan-1-ol and caryophyllene through humulyl cation A. CS contains an alanine at its analogous position (A79). This is in accordance with what has been observed with PS and IHS, since CS does not contain an aromatic residue at that position and does not facilitate the subsequent hydride shift. With this logic, the CS A79F mutant was designed to attempt to induce a hydride shift and in turn introduce humulyl cation B as an intermediate. However, the mutation did not alter the product profile, leading to the conclusion that this position does not play a significant role in the CS mechanism. With these established differences, attention was turned to the active site architecture of CS to further elucidate its mechanistic properties.
CS differs from PS in its initial loading of substrate. A crystal structure of CS complexed with its substrate analog, 2-fluorofarnesyl pyrophosphate (2FFPP), revealed that the substrate was linear in the active site, and not curled in a reaction-ready confirmation, as was observed in PS. The linearity of the substrate suggests a stepwise mechanism for CS. Unexpectedly, an electron density was also observed in the active site of the ‘apo’ CS structure. This was fitted to tetraethylene glycol - presumably recruited from the crystallization medium - which was observed in a curled up conformation. This density was also well-fit to farnesene, the dephosphorylated form of FPP, and was accompanied with dramatic active site rotamers, setting the stage for studying the role of additional residues in facilitating CS mechanism. The most notable of these residues was W56, which undergoes a 90º rotation, appearing to coerce the curled-up conformation of tetraethylene glycol. Indeed, the W56L mutation in CS abolished the enzyme’s ability to cyclize FPP, producing solely linear products. Additional residues were studied, such as H309, a conserved residue in both PS and CS. While H309 was previously shown to not be an essential residue in the PS mechanism, it was speculated that it may play a role in CS. With this in mind, the CS H309G mutant was designed to remove any functionality at the 309 position. Surprisingly, this mutation completely altered the product specificity towards pentalenene, along with a drastic decrease in activity.
This work sheds light on the catalytic properties of SCs, as well as how minimal active site mutations can drastically alter mechanism and product profile. Mutagenesis studies with PS and IHS suggest that closely-related enzymes perform remarkably similar mechanisms; however studies with CS suggest that these mechanistic trends may be confined to evolutionary clades. This will be the focus of future studies.