Because it controls the majority of polyketide stereocenters, the ketoreductase (KR) is a central target in engineering polyketide synthases (PKSs). To elucidate the mechanisms of stereocontrol, the structure of KR from the first module of the tylosin PKS was determined. A comparison with a recently solved erythromycin KR that operates on the same substrate explains why their products have opposite alpha-substituent chiralities. The structure reveals how polyketides are guided into the active site by key residues in different KR types. There are four types of reductase-competent KRs, each capable of fixing a unique combination of alpha-substituent and beta-hydroxyl group chiralities, as well as two types of reductase-incompetent KRs that control alpha-substituent chirality alone. A protocol to assign how a module will enforce substituent chirality based on its sequence is presented.

With the recent structural characterization of each of the component enzymes of type I polyketide synthases, scientists are coming tantalizingly close to elucidating the overall architectures and mechanisms of these enormous molecular factories. This review highlights not only what has been revealed about the structures and activities of each of the domains but also the mysteries that remain to be solved.

The structure of the ketoreductase (KR) from the first module of the erythromycin synthase with NADPH bound was solved to 1.79 A resolution. The 51 kDa domain has two subdomains, each similar to a short-chain dehydrogenase/reductase (SDR) monomer. One subdomain has a truncated Rossmann fold and serves a purely structural role stabilizing the other subdomain, which catalyzes the reduction of the beta-carbonyl of a polyketide and possibly the epimerization of an alpha-substituent. The structure enabled us to define the domain boundaries of KR, the dehydratase (DH), and the enoylreductase (ER). It also constrains the three-dimensional organization of these domains within a module, revealing that KR does not make dimeric contacts across the 2-fold axis of the module. The quaternary structure elucidates how substrates are shuttled between the active sites of polyketide synthases (PKSs), as well as related fatty acid synthases (FASs), and suggests how domains can be swapped to make hybrid synthases that produce novel polyketides.

SUMMARY The dehydratases (DHs) of modular polyketide synthases (PKSs) catalyze dehydrations that occur frequently in the biosynthesis of complex polyketides, yet little is known about them structurally or mechanistically. Here, the structure of a DH domain, isolated from the fourth module of the erythromycin PKS, is presented at 1.85 Å resolution. As with the DH of the highly related animalian fatty acid synthase (FAS), the DH monomer possesses a double hotdog fold. Two symmetry mates within the crystal lattice make a contact that likely represents the DH dimerization interface within an intact PKS. Conserved hydrophobic residues on the DH surface indicate potential interfaces with two other PKS domains, the ketoreductase (KR) and the acyl carrier protein (ACP). Mutation of an invariant arginine at the hypothesized ACP docking site in the context of the erythromycin PKS resulted in decreased production of the erythromycin precursor 6-deoxyerythronolide B. The structure elucidates how the α-hydrogen and β-hydroxyl group of a polyketide substrate interact with the catalytic histidine and aspartic acid in the DH active site prior to dehydration.

Malonyl-CoA:ACP transacylase (MAT), the fabD gene product of Streptomyces coelicolor A3(2), participates in both fatty acid and polyketide synthesis pathways, transferring malonyl groups that are used as extender units in chain growth from malonyl-CoA to pathway-specific acyl carrier proteins (ACPs). Here, the 2.0 A structure reveals an invariant arginine bound to an acetate that mimics the malonyl carboxylate and helps define the extender unit binding site. Catalysis may only occur when the oxyanion hole is formed through substrate binding, preventing hydrolysis of the acyl-enzyme intermediate. Macromolecular docking simulations with actinorhodin ACP suggest that the majority of the ACP docking surface is formed by a helical flap. These results should help to engineer polyketide synthases (PKSs) that produce novel polyketides.