An improved method for the synthesis of analogs of coenzyme A (CoA) and its thioesters, which are modified in the thiol or thioester moiety, has been developed using a combination of chemical and enzymatic reactions. The enzymes catalyzing the last two steps of CoA biosynthesis were used to prepare a CoA analog (lc) in which an amide bond is replaced by a thioester bond and the thiol group is replaced by a methyl group. Reaction of lc with a primary amine in aqueous solution results in aminolysis of the thioester linkage to form the desired CoA analog. Reaction with different amines permits the introduction of a variety of functional groups in place of the normal thiol or thioester group. This methodology has been used in the synthesis of five new analogs of acetyl-CoA in which the thioester sulfur is replaced by a methylene group and the acetyl group is replaced by carboxylate (14a), nitro (14b), carboxamide (14c), methyl sulfoxide (14d), and methyl sulfone (14e) groups. 14a-c were designed to mimic the possible enolate or enol intermediate in the reaction of citrate synthase and related enzymes. 14a and 14c are potent inhibitors of citrate synthase, with K-, values 1000-and 570-fold lower than the Km for acetyl-CoA, respectively. CD titrations indicate that 14a and 14c have low affinity for citrate synthase in the absence of oxaloacetate, consistent with their recognition as enol or enolate analogs. 14b is a poor inhibitor of citrate synthase, with affinity slightly lower than that for acetyl-CoA. These results are consistent with generation of the enol form of acetyl-CoA as the nucleophilic intermediate in the reaction of citrate synthase. 14d and 14e were designed to mimic the tetrahedral intermediate or transition state in the reaction of chloramphenicol acetyltransferase and related acetyl-CoA-dependent acetyltransferases. Both compounds are poor inhibitors of chloramphenicol acetyltransferase, with affinities slightly lower than that of acetyl-CoA, indicating that these compounds are not good mimics of the enzyme-bound tetrahedral intermediate or transition state.

A general approach to the synthesis of enantiomerically pure spirocyclic alpha,beta-butenolides is presented where the fundamental framework is rapidly elaborated by acid- or bromonium ion-induced rearrangement of the carbinol derived by addition of 2-lithio-4,5-dihydrofuran to cyclobutanone. Subsequent resolution of the resulting ketones by either sulfoximine or mandelate acetal technology has been applied effectively. The availability of these building blocks makes possible in turn the acquisition of the enantiomers of dihydrofurans typified by 17, 35, and 38 and lactones such as 25 and 31, as well as the targeted title compounds. Complementary reductions of the early intermediates provide the added advantage that the alpha- and beta-stereoisomeric carbinol series can be obtained on demand. These capabilities have been coordinated to allow the crafting of any member of the series in relatively few steps.

We have previously reported a general synthetic approach to analogues of coenzyme A (CoA) which involves enzymatic synthesis of a general CoA analogue synthon having a thioester linkage in place of the amide bond nearest the thiol group (Martin et al. J. Am. Chem. Soc. 1994, 116, 4660). We report here the synthesis of a second CoA analogue synthon 1c which has the amide bond more distant from the thiol group replaced with a thioester. This analogue was prepared by nonenzymatic synthesis of a racemic phosphopantetheine analogue followed by enzymatic conversion to the corresponding CoA analogue. Stereochemical analysis showed that the natural enantiomer of the phosphopantetheine analogue was selectively converted to product by the enzyme phosphopantetheine adenylyltansferase, yielding a product that possessed the desired stereoconfiguration. Reaction of the new synthon 1c with a primary amine results in amide bond formation to form the CoA analogue of interest. This new methodology provides access to an even broader array of CoA analogues modified in the beta-alanylcysteamine moiety. This has been demonstrated in the synthesis of an analogue having an extra methylene group in the beta-alanine moiety and two analogues in which the amide bond nearest the thiol group is replaced with a pair of methylene groups.