Engineering potassium-activation into a potassium-independent thiolase

Cytosolic thiolase (CT) catalyses the Claisen condensation of two molecules of acetyl-CoA to produce acetoacetyl-CoA. This is the first step in the synthesis of an extremely diverse range of natural products, including isoprenoids in plants and animals, and polyhydroxyalkanoates in many bacteria. The CT reaction cycle proceeds via a ping-pong mechanism involving an acetylated enzyme intermediate and two separate oxyanion holes which stabilize negatively charged reaction intermediates. We have solved crystal structures for Aspergillus fumigatus CT in the presence of substrate, and have serendipitously trapped various stages of the thiolase catalytic cycle, including two tetrahedral reaction intermediates that have previously eluded structural characterisation. We have also shown that afCT is strongly activated by potassium ions, a biochemical property previously thought to apply only to the mitochondrial biosynthetic thiolase. Interestingly, the activity of the bacterial homologue (bCT), which is structurally very similar to afCT, is not dependent on potassium ions. Comparison of the potassium binding site of afCT with the corresponding region of bCT suggests that very few residues are responsible for conferring potassium-activation. We have used the structure of afCT as a guide to design a series of mutations in bCT aimed at defining the key structural determinants of potassium-activation. Only two mutations were together necessary to engineer potassium-activation into bCT: Y218E to increase the overall negative charge of the substrate binding pocket, and Q183Y to complete the potassium coordination sphere. These data provide valuable insight into the thiolase reaction mechanism and allosteric activation of these enzymes by monovalent cations.