Figure 1. Formation of holo-ACP from coenzyme A and apo-ACP catalysed by 4¢-phosphopantetheine transferase (holosynthase). An invariant serine of the apo-ACP interacts with the transferase and CoASH in the presence of Mg²⁺, allowing the serine hydroxyl group to react with the phosphodiester bond of CoASH, displacing AMP and forming the serine-4¢-phosphopantetheine linkage. The holo-ACP cysteamine thiol is then able to form ACP-thiol esters, normally by exchange with an acyl derivative in the presence of a specific transacylase.

Figure 2. Chemistry of fatty acid synthases, polyketide synthase and nonribosomal polypeptide synthases. In fatty acid synthases and polyketide synthases an electrophilic acyl-CoA starter or 4¢-phosphopantetheine-bound acyl thiol ester is transferred to a reactive cysteine residue of a condensing enzyme (KS). This then reacts with a nucleophilic -carbon atom of malonyl-ACP accompanied by decarboxylation (a). The resulting -ketoacyl thiol ester is then processed according to the specificity of the enzyme (see Figures 3 and 9 text). In some polyketide synthases, methylmalonyl-ACP may be used. The thiol ester also has an important influence during the subsequent manipulations of 4¢-phosphopantetheine-bound intermediates, facilitating protonation at the position of the -ketoacyl, enoyl and -hydroxy intermediates during the reductions and dehydration steps. In nonribosomal polypeptide synthases (b) it is the aminoacyl thiol ester (R¹) attached to a peptidyl carrier protein PCP_n–1 that acts as the electrophile, the nucleophile being the amino acid of the aminoacyl-PCP_n thiol ester (R²). The reaction is mediated by a specific condensation domain (C). The dipeptide, now attached to PCP_n, reacts with a third aminoacyl-PCP_n+1 carrying an amino acid (R³). See also Figure 11.

Figure 3. Generalized scheme for fatty acid synthesis. Initially, in a priming reaction, the acetyl group of acetyl-CoA is transferred to the acyl carrier protein (ACP) from which it is then transferred to a cysteine thiol of -ketoacyl synthase (KS). The malonyl moiety of malonyl-CoA is then linked to the vacant ACP thiol by a malonyltransferase (MT) and condensation occurs, with decarboxylation, to give the acetoacetyl-thiol ester of ACP. All subsequent reactions occur with the intermediates bound through a thiol ester link to the ACP thiol. Reduction by NADPH is catalysed by the ketoreductase (KR) to give a secondary alcohol, followed by dehydration, catalysed by dehydratase (DH). A second reduction by NADPH, catalysed by enoyl reductase (ER) yields the fully saturated butyryl-ACP thiol ester. The butyryl moiety is then transferred to the KS active site thiol to complete the ‘cycle’. Subsequent cycles occur as above, each commencing with the loading of the ACP with a new malonyl moiety. The cycles continue until a C₁₆ chain has been assembled, at which point the fatty acid is released from the ACP by a thioesterase (TE). The thioesterase is specific for a chain length of 16 carbon atoms.

Figure 4. Organization of functionalities within the dimeric chicken liver fatty acid synthase. Proteolytic studies have revealed that the chicken FAS is composed of three main ‘domains’ – I, II and III. Domain I contains the KS and MT/AT functionalities; domain II contains the DH, ER, KR, and ACP functions and domain III is the thioesterase (TE). The enzyme functions as a dimer with the active-site cysteine of one KS functionality being close to the 4¢-phosphopantetheine cofactor of the ACP from the other subunit in a head-to-tail arrangement.

Figure 5. Structure and active site of the -ketoacyl synthase (KS) from Escherichia coli. Ribbon diagram of the -ketoacyl synthase II from E. coli (PDB access code 1KAS). The interrelationship between the side-chains of the active-site residues Cys163, His303 and His340 are shown. See text for details.

Figure 6. Active-site residues of malonyl-CoA:ACP transacylase from Escherichia coli. The three-dimensional arrangement of the catalytic dyad (Ser/His) in the E. coli malonyl-CoA:ACP transacylase (PDB code 1MLA). Gln250 stabilizes the N-1 of His201 and replaces the aspartate, typically seen in Asp/His/Ser triads. The side-arm of Arg117 is also shown, and the N-2 protons of Arg117 hydrogen-bond to the side-chain hydroxyl of the nucleophilic Ser92. Residues Ser92, His201 and Arg117 constitute three out of eight invariant residues in homologues of malonyl-CoA:ACP transacylase.

Figure 7. Solution structures of the fatty acid synthase ACP from E. coli. Two distinct solution structures ((a) and (b)) determined for the Esherichia coli fatty acid synthase ACP. ACP1 shows a solution in which the secondary structure is seen to be largely intact, whereas ACP2 shows significant disruption of the helices. In both cases, the solutions produce a subset of violations of the nuclear magnetic resonance-derived distance data used to derive the final structures. It is thought that significant conformational averaging means there are necessarily several possible solution structures.

Figure 8. Polyketide derived secondary metabolites of commercial importance.

Figure 9. Scheme for classical and reduced processive polyketide synthesis.

Figure 10. Diagramatic representation of the modules of 6-deoxyerythronolide B synthase from Saccharopolyspora erythraea. The enzyme consists of three giant proteins, DEBS-1, DEBS-2 and DEBS-3, each of which is made up of two modules. Each module contains all the necessary functionalities for the condensation and appropriate processing of each methylmalonyl extender unit. Thus module 4 contains a full set of functionalities – KS, AT, DH, ER, KR and ACP – arranged as in fatty acid synthases, with KR out of sequence. Module 1 also contains an additional AT and ACP for loading the starter propionyl unit. Module 6 contains a thioesterase (TE) responsible for releasing the product from the terminal ACP.

Figure 11. Formation of the ‘unnatural’ polyketides SEK4 and SEK4b, in vivo, using genetically engineered PKS from Streptomyces coelicolor. All the components of the act locus from S. coelicolor (see insert) function to synthesize actinorhodin. Use of a minimal PKS (actI) containing only -ketoacyl synthase (KS), chain length factor (CLF) and ACP results in the octaketides SEK4 and SEK4b. Addition of the remaining genes leads to a variety of other unnatural polyketides. For instance, the addition of the ketoreductase (KR) leads to the formation of a different octaketide, mutactin. The presence of the aromatase (ARO) gives rise to SEK34 and the presence of cyclase (CYC) leads to 3,8-dihydroxy-1-methylanthraquinone-2-carboxylic acid (DMAC) and aloesaponarin.

Figure 12. The mechanistic role of 4¢-phosphopantetheine in the peptide carrier protein (PCP) from a nonribosomal polypeptide synthase (NRPS) and the arrangement of the functionalities of surfactin synthase. (a) Stages in the formation of a peptide bond by a nonribosomal polypeptide synthase (NRPS). The carboxyl group of the amino acid, R, is activated as its acyladenylyl derivative and linked to the 4¢-phosphopantetheine arm of a specific PCP_n by a highly specific acyltransferase (A). The free amino group of this aminoacyl-PCP extender then attacks the thiol ester bond of the previous amino acid (or peptide), already bound as its aminoacyl-PCP_n–1 by a similar activation process. The reaction, catalysed by a specific condensing enzyme (C), leads to the formation of the new peptide bond and the displacement of the PCP_n–1. The peptide, still linked to PCP_n, can be further elongated by reaction with another amino group from an aminoacyl-PCP_n+1 (not shown). In this way, the peptide is built in a stepwise process, from N-terminus to C-terminus, passing along the PCP functionalities until the final stage when the peptide-PCP thiol ester bond is hydrolysed to yield the terminal carboxyl group, releasing the product from the enzyme (not shown). Alternatively, the terminal thiol ester may be displaced by the amino group of the N-terminal amino acid to form a cyclic peptide. (b) Arrangement of the functionalities of the NRPS surfactin synthase. The synthase is arranged in modules, each with condensing enzyme (C), acyltransferase (A) and peptide carrier protein (PCP) functionalities and the additional possibility of an epimerase (EPI) functionality. Modules, one for the addition of each amino acid, are arranged in the sequence of amino acids in the final peptide chain. Thus, the first module links glutamate to the C12 starter unit, the second links leucine, and so on. The energy for the final cyclization to form the ester comes from the reaction of the hydroxyl group with the terminal aminoacyl-PCP thiol ester.