The catalytic mechanism of the beta-ketoacyl synthase domain of the multifunctional fatty acid synthase has been investigated by a combination of mutagenesis, active-site titration, product analysis, and product inhibition. Neither the reactivity of the active-site Cys161 residue toward iodoacetamide nor the rate of unidirectional transfer of acyl moieties to Cys161 was significantly decreased by replacement of any of the conserved residues, His293, His331, or Lys326, with Ala. Decarboxylation of malonyl moieties in the fully-active Cys161Gln background generated equimolar amounts of acetyl-CoA and bicarbonate, rather than carbon dioxide, and was seriously compromised by replacement of any of the conserved basic residues. The ability of bicarbonate to inhibit decarboxylation of malonyl moieties in the Cys161Gln background was significantly reduced by replacement of His293 but less so by replacement of His331. The data are consistent with a reaction mechanism, in which the initial primer transfer reaction is promoted largely through a lowering of the pKa of the Cys161 thiol by a helix dipole effect and activation of the substrate thioester carbon atom by binding of the keto group in an oxyanion hole. The data also indicate that an activated water molecule is present at the active site that is required either for the rapid hydration of carbon dioxide, prior its release as bicarbonate or, alternatively, for an initial attack on the malonyl C3. In the alternative mechanism, a negatively-charged tetrahedral transition state could be generated, stabilized in part by interaction of His293 with the negatively charged oxygen at C3 and interaction of His331 with the negatively charged thioester carbonyl oxygen, that breaks down to generate bicarbonate directly. Finally, the carbanion at C2, attacks the electrophilic C1 of the primer, generating a second tetrahedral transition state, also stabilized through contacts with the oxyanion hole and His331, that breaks down to form the beta-ketoacyl-S-acyl carrier protein product.

The animal fatty acid synthase is a dimer of identical, multifunctional 272 kDa subunits oriented antiparallel such that two centers for fatty acid synthesis are formed at the subunit interface. In order to clarify the interdomain and intersubunit communications necessary for the operation of the two centers, we have explored the possibility of reassembling catalytically-active fatty acid synthase heterodimers from pairs of inactive dimers carrying mutations in different functional domains. To this end, rat fatty acid synthase mutants, defective in either the beta-ketoacyl synthase, C161T or K326A (KS- FAS), or the acyl carrier protein, S2151A (ACP- FAS), domains, were engineered by site-directed mutagenesis, expressed in insect Sf9 cells using a baculovirus expression system, and purified. A novel procedure was devised to facilitate rapid production and isolation of a population of mixed mutant dimers that had undergone randomization of its constituent subunits. Homodimeric mutants (KS- FAS/KS- FAS and ACP- FAS/ACP- FAS) and KS- FAS heterodimers consisting of paired C161T and K326A mutant subunits were unable to synthesize fatty acids, confirming the essential nature of residues C161, K326, and S2151A. However, KS- FAS/ACP- FAS heterodimers regained partial activity. Formation of these heterodimers necessitated prior dissociation and reassociation of the homodimers, indicating that the rate of spontaneous exchange of subunits in the dimer is negligible. The formation of catalytically-active heterodimers from pairs of inactive, complementary homodimers affords a useful method for testing the validity of the current model for the multifunctional complex.

An in vitro mutant complementation approach has been used to map the functional topology of the animal fatty acid synthase. A series of knockout mutants was engineered, each mutant compromised in one of the seven functional domains, and heterodimers generated by hybridizing all possible combinations of the mutated subunits were isolated and characterized. Heterodimers comprised of a subunit containing either a beta-ketoacyl synthase or malonyl/acetyltransferase mutant, paired with a subunit containing mutations in any one of the other five domains, are active in fatty acid synthesis. Heterodimers in which both subunits carry a knockout mutation in either the dehydrase, enoyl reductase, keto reductase, or acyl carrier protein are inactive. Heterodimers comprised of a subunit containing a thioesterase mutation paired with a subunit containing a mutation in either the dehydrase, enoyl reductase, beta-ketoacyl reductase, or acyl carrier protein domains exhibit very low fatty acid synthetic ability. The results are consistent with a model for the fatty acid synthase in which the substrate loading and condensation reactions are catalyzed by cooperation of an acyl carrier protein domain of one subunit with the malonyl/acetyltransferase or beta-ketoacyl synthase domains, respectively, of either subunit. The beta-carbon-processing reactions, responsible for the complete reduction of the beta-ketoacyl moiety following each condensation step, are catalyzed by cooperation of an acyl carrier protein domain with the beta-ketoacyl reductase, dehydrase, and enoyl reductase domains associated exclusively with the same subunit. The chain-terminating reaction is carried out most efficiently by cooperation of an acyl carrier protein domain with the thioesterase domain of the same subunit. These results are discussed in the context of a revised model for the fatty acid synthase.

The properties of the beta-ketoacyl reductase, dehydrase, and enoyl reductase components of the animal fatty acid synthase responsible for the reduction of the beta-ketoacyl moiety formed at each round of chain elongation have been studied by engineering and characterizing mutants defective in each of these three catalytic domains. These "beta-carbon processing" mutants leak the stalled four-carbon intermediates by direct transfer to CoA. However, enoyl reductase mutants leak beta-ketobutyryl, beta-hydroxybutyryl, and crotonyl moieties, a finding explained, at least in part, by the observation that the equilibrium and rate constant for the dehydrase reaction favor the formation of beta-hydroxy rather than enoyl moieties. In this regard, the type I animal fatty acid synthase resembles its type II counterpart in Escherichia coli in that both systems rely on the enoyl reductase to pull the beta-carbon processing reactions to completion. Kinetic and nucleotide binding measurements on fatty acid synthases mutated in either of the two nucleotide binding domains revealed that the NADPH binding sites are nonidentical, the enoyl reductase exhibiting higher affinity. Surprisingly, NADPH binding is also completely compromised by certain deletions and mutations in the central core region distant from the nucleotide binding sites. Comparable central core sequences are present in the structurally related modular polyketide synthases, except in those modules that lack all three beta-carbon processing enzymes. These findings suggest that the central core region of fatty acid and polyketide synthases plays an important role in facilitating the beta-carbon processing reactions.

The enzyme activity responsible for translocation of saturated acyl chains from the 4'-phosphopantetheine of the acyl carrier protein to the active site cysteine of the beta-ketoacyl synthase in the animal fatty acid synthase has been identified. An enzyme assay was devised that allows uncoupling of the interthiol transfer step from the condensation reaction. Experiments with various fatty acid synthase mutants indicate clearly that catalysis of the transfer of saturated acyl moieties from the 4'-phosphopantetheine thiol to the active site cysteine thiol, Cys-161, is an inherent property of the beta-ketoacyl synthase domain. Catalytic efficiency of the interthiol transferase increases from C2 to C12 and decreases with increasing chain-lengths beyond C12. Malonyl, beta-hydroxybutyryl, and crotonyl thioesters are not substrates for the transferase, whereas the beta-ketobutyryl moiety is a poor substrate. These features of the substrate specificity are exactly as predicted for a transferase that fulfills the proposed role in the fatty acid synthase reaction sequence and indicate that this activity plays an important role in determining the overall specificity of the beta-ketoacyl synthase reaction.