The kinetics and mechanism of the citrate synthase from a moderate thermophile, Thermoplasma acidophilum (TpCS), are compared with those of the citrate synthase from a mesophile, pig heart (PCS). All discrete steps in the mechanistic sequence of PCS can be identified in TpCS. The catalytic strategies identified in PCS, destabilization of the oxaloacetate substrate carbonyl and stabilization of the reactive species, acetyl-CoA enolate, are present in TpCS. Conformational changes, which allow the enzyme to efficiently catalyze both condensation of acetyl-CoA thioester and subsequently hydrolysis of citryl-CoA thioester within the same active site, occur in both enzymes. However, significant differences exist between the two enzymes. PCS is a characteristically efficient enzyme: no internal step is clearly rate-limiting and the condensation step is readily reversible. TpCS is a less efficient catalyst. Over a broad temperature range, inadequate stabilization of the transition state for citryl-CoA hydrolysis renders this step nearly rate-limiting for the forward reaction of TpCS. Further, excessive stabilization of the citryl-CoA intermediate renders the condensation step nearly irreversible. Values of substrate and solvent deuterium isotope effects are consistent with the kinetic model. Near its temperature optimum (70 degrees C), there is a modest increase in the reversibility of the condensation step for TpCS, but reversibility still falls short of that shown by PCS at 37 degrees C. The root cause of the catalytic inefficiency of TpCS may lie in the lack of protein flexibility imposed by the requirement for thermal stability of the protein itself or its temperature-labile substrate, oxaloacetate.

The carbon-13 NMR spectrum of oxaloacetate bound in the active site of citrate synthase has been obtained at 90.56 MHz. In the binary complex with enzyme, the positions of the resonances of oxaloacetate are shifted relative to those of the free ligand as follows: C-1 (carboxylate), -2.5 ppm; C-2 (carbonyl), +4.3 ppm; C-3 (methylene), -0.6 ppm; C-4 (carboxylate), +1.3 ppm. The change observed in the carbonyl chemical shift is successively increased in ternary complexes with the product [coenzyme A (CoA)], a substrate analogue (S-acetonyl-CoA), and an acetyl-CoA enolate analogue (carboxymethyl-CoA), reaching a value of +6.8 ppm from the free carbonyl resonance. Binary complexes are in intermediate to fast exchange on the NMR time scale with free oxaloacetate; ternary complexes are in slow exchange. Line widths of the methylene resonance in the ternary complexes suggest complete immobilization of oxaloacetate in the active site. Analysis of line widths in the binary complex suggests the existence of a dynamic equilibrium between two or more forms of bound oxaloacetate, primarily involving C-4. The changes in chemical shifts of the carbonyl carbon indicate strong polarization of the carbonyl bond or protonation of the carbonyl oxygen. Some of this carbonyl polarization occurs even in the binary complex. Development of positive charge on the carbonyl carbon enhances reactivity toward condensation with the carbanion/enolate of acetyl-CoA in the mechanism which has been postulated for this enzyme. The very large change in the chemical shift of the reacting carbonyl in the presence of an analogue of the enolate of acetyl-CoA supports this interpretation.

The infrared spectrum of oxaloacetate bound in the active site of citrate synthase has been measured in the binary complex and in the ternary complex with the acetyl coenzyme A (CoA) enolate analogue carboxymethyl-CoA. The carbonyl stretching frequency of oxaloacetate in binary and ternary complexes is found at 1697 cm-1, a shift of 21 cm-1 to lower frequency relative to that of the free ligand. The line widths of the carbonyl absorption in enzyme complexes differ from that of the free ligand, decreasing from a value of 20 cm-1 for the free ligand to 10 cm-1 in the binary complex and 7 cm-1 in the ternary complex with carboxymethyl-CoA. The integrated absorbance of the carbonyl absorption in these enzyme complexes is significantly increased over that of the free ligand at the same concentration, increasing approximately 2-fold in the binary complex and approximately 3-fold in the ternary complex. These results indicate strong polarization of the carbonyl bond in the enzyme-substrate complexes and suggest that ground-state destabilization is a major catalytic strategy of citrate synthase.

This work reports the relative importance of the interactions provided by three catalytic residues to individual steps in the mechanism of citrate synthase. When the side chains of any of the residues (H320, D375, and H274) are mutated, the data indicate that they are involved in the stabilization of one or more of the transition/intermediate states in the multistep citrate synthase reaction. H320 forms a hydrogen bond with the carbonyl of oxaloacetate and the alcohols of the citryl-coenzyme A and citrate products. Enzymes substituted at H320 (Q, G, N, and R) have reaction profiles for which the condensation reaction is cleanly rate determining. None of these mutants can activate the carbonyl of oxaloacetate by polarization. All these mutants catalyze the necessary proton transfer from the methyl group of acetyl-coenzyme A only poorly, a process which occurs in a structurally separate site. Furthermore, all H320 mutants hydrolyze the citryl-coenzyme A intermediate significantly more slowly than does the wild-type. D375 is the base removing the proton of acetyl-coenzyme A. D375E and D375G have greatly diminished ability to catalyze proton transfer from acetyl-CoA. The D375 mutants polarize the oxaloacetate carbonyl as well as wild-type. For D375E, the hydrolysis of citryl-CoA is rate determining. D375G, having no side chain capable of acid-base chemistry in either the condensation or hydrolysis reactions is nearly completely devoid of activity in any of the reactions catalyzed by the wild-type. H274 hydrogen bonds to the carbonyl of acetyl-coenzyme A but also forms the back wall of the oxaloacetate-binding site. H274G cannot properly activate either oxaloacetate or acetyl-coenzyme A, and the condensation reaction is overwhelmingly rate determining. Nonetheless, hydrolysis of the intermediate is impaired. All the enzymes except H320R and H274G show kinetic cooperativity with CitCoA as substrate, indicating changes in the subunit interactions with these latter two mutants. The energetics of citrate synthase are surprisingly tightly coupled. All changes affect more than one step in the catalytic cycle. Within the condensation reaction, the intermediate of proton transfer must occupy a shallow well between transition states close in free energy so that perturbations of one have substantial effects on that of the other.

The catalytic strategies of enzymes (such as citrate synthase) whose reactions require the abstraction of the alpha-proton of a carbon acid remain elusive. Citrate synthase readily catalyzes solvent proton exchange of the methyl protons of dethiaacetyl-coenzyme A, a sulfur-less, ketone analog of acetyl-coenzyme A, in its ternary complex with oxaloacetate. Because no further reaction occurs with this analog, it provides a uniquely simple probe of the roles of active site interactions on carbon acid proton transfer catalysis. In view of the high reactivity of the analog for proton transfer to the active site base, its failure to further condense with oxaloacetate to form a sulfur-less analog of citryl-coenzyme A was unexpected, although we offer several possible explanations. We have measured the rate constants for exchange, k(exch), at saturating concentrations of the analog for six citrate synthase mutants with single changes in active site residues. Comparisons between the values of k(exch) are straightforward in two limits. If the rate of exchange of the transferred proton with solvent protons is rapid, then k(exch) equals the forward rate constant for proton transfer, and k(exch) values for different mutants compare directly the rate constants for proton transfer. If the exchange of the transferred proton with protons in the bulk solution is the slow step and the equilibrium constant for proton transfer is unfavorable (as is likely), then k(exch) equals the product of the equilibrium constant for proton transfer and the rate constant for exchange of the transferred proton with bulk solvent. If that exchange rate with bulk solution remains constant for a series of mutant enzymes, then k(exch) values compare the equilibrium constants for proton transfer. The importance of the acetyl-CoA site residues, H274 and D375, is confirmed with D375 again implicated as the active site base. The results with the series of oxaloacetate site mutants, H320X, strongly suggest that activation of the first substrate, oxaloacetate, through carbonyl bond polarization, not just oxaloacetate binding in the active site, is required for the enzyme to efficiently catalyze proton transfer from the methyl group of the second substrate.