Two functional domains of coenzyme A activate catalysis by coenzyme A transferase. Pantetheine and adenosine 3'-phosphate 5'-diphosphate.

Studies of the reactivity of succinyl-CoA:3-keto acid CoA transferase with a small coenzyme A analog, methylmercaptopropionate, have shown that noncovalent interactions between the enzyme and the side chain of CoA are responsible for a rate acceleration of approximately 10(12), which is close to the total rate acceleration brought about by the enzyme (Moore, S. A., and Jencks, W. P. (1982) J. Biol. Chem. 257, 10893-10907). We report here that interaction between the enzyme and the pantetheine moiety of CoA provides the majority of the rate acceleration and destabilization of the enzyme-thiol ester intermediate that is observed with CoA substrates. The role of the adenosine 3'-phosphate 5'-diphosphate moiety of CoA is to provide 6.9 kcal/mol of binding energy in order to pull the pantetheine moiety into the active site. The enzyme-thiol ester intermediate, E-pantetheine, was generated by reaction of pantetheine with the thiol ester of enzyme and methylmercaptopropionate. E-Pantetheine undergoes hydrolysis with khyd = 2 min-1, 140-fold faster than E-CoA, and reacts with acetoacetate with kAcAc = 3 X 10(6) M-1 min-1, only 10-fold slower than E-CoA. However, in the reverse direction acetoacetylpantetheine reacts with CoA transferase (kAcAc-SP = 220 M-1 min-1) 1.6 X 10(6) times slower than acetoacetyl-CoA. The equilibrium constant for the reaction of pantetheine with E-CoA is approximately 8 X 10(-6).

Enzyme catalysis involves the use of intrinsic binding energy to stabilize the transition state for reactions of specific substrates. It is equally important for catalysis that the intrinsic binding energy should not be expressed in the enzymesubstrate complex; instead, it is utilized to overcome destabilization and loss of entropy so that the transition state can be reached easily (1)(2)(3)(4)(5). Enzymes differ from most chemical catalysts in their ability to utilize the intrinsic binding energy of nonreacting groups on specific substrates to bring about large rate increases. For example, noncovalent interactions between the enzyme succinyl-CoA3-ketoacid coenzyme A transferase (3-oxoacid CoA transferase, EC 2.8.3.5) and non-* This work was supported in part by National Institutes of Health Grant GM 20888 and National Science Foundation Grant PCM 81-17816. This is contribution no. 1584 from the Graduate Department of Biochemistry, Brandeis, University. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduer-tisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. reacting portions of the coenzyme A molecule provide a rate increase of 3 X 10' ' that is close to the total rate increase of 6 X 1013 for catalysis by this enzyme (6). We report here that destabilization and the expression of observed binding energy are brought about by different portions of the coenzyme A molecule. Interaction between the enzyme and the pantetheine moiety provides the majority of substrate destabilization and rate acceleration, while the interaction with the 3'-phospho-ADP1 moiety provides binding energy that overcomes this destabilization and permits significant binding of acyl-CoA substrates to the enzyme.

EXPERIMENTAL PROCEDURES
Materials-Succinic acid, DTNB, IAA, and disodium EDTA were recrystallized, and MMP and diketene were distilled before use. Commercial products include the dilithium salt of coenzyme A (95% pure, P-L Biochemicals), disodium adenosine 3',5'-diphosphate (98% pure, Sigma), and pantethine (U. S. Biochemicals). Potassium acetoacetate was prepared by the method of Seeley ( 7 ) and the concentration was determined by the method of Walker (8). Pantetheine was synthesized by reduction of pantethine with a 10-fold molar excess of sodium borohydride in water at 4 "C, with the pH maintained below 8.5, purified by chromatography on a Sephadex G-10 column, and concentrated by lyophilization. The free thiol concentration was measured with DTNB (10). Acetoacetyl-CoA and acetoacetylpantetheine were prepared from diketene and CoA or pantetheine, as described previously (6). Traces of remaining free thiol were removed by reaction with IAA (6). The concentration of acetoacetyl-SR was determined at 310 nm in 67 mM Tris sulfate, 5 mM magnesium sulfate, pH 8.1, 25 "C (9).
CoA transferase was purified from pig hearts to a specific activity of 285 pmol/min/mg (6). For some experiments the protein was concentrated to 10 mg/ml using a Centricon microconcentrator (Amicon Corp.).

RESULTS
Pantetheine catalyzes the formation of active enzyme from the inactive thiol ester derivative, E-MMP, with a secondorder rate constant of kSP = 5.9 f 0.3 "' min-'. This reaction can be completely inhibited by low concentrations of MMP, as shown in Fig. 1. Since MMP does not bind to E-MMP under these conditions (6), this inhibition requires the formation of an E-pantetheine thiol ester intermediate, E-SP. This intermediate can either hydrolyze to form active enzyme or react with MMP to regenerate inactive E-MMP (Scheme 1) with the rate constants khyd and ~M M P , respectively. The ratio of these rate constants, k"p/khyd = (1.3 f 0.1) X IO3 M -~, was obtained from the concentration of MMP necessary to give 50% inhibition of the reactivation rate after correcting for the hydrolysis of E-MMP (dashed line, Fig. 1). The solid line in Fig. 1 was calculated from this ratio and ksP.   for E-CoA it shows that E-SP is even more activated toward nonspecific reactions than E-CoA, which undergoes hydrolysis -50 times faster than acetyl-coA (6) ( Table I).
Rate constants for other reactions of E-SP were calculated from the observed rate constant ratios and the value of khyd = 2 min". These rate constants are compared with those for E-CoA in Table I. this solution was then diluted into 2.5 ml of assay buffer and the enzyme activity and total enzyme activity were assayed as described in Fig. 1. The enzyme activity was corrected for inhibition by MMP in the assay (<5%). The lines are theoretical assuming either a single first-order reaction or two consecutive first-order reactions (14).
lyze the reaction of AcAc-SP and succinate to form succinylpantetheine, although this reaction was not detected previously (9). As predicted, CoA transferase (69 p~, 138 pM active sites) was found to catalyze the disappearance of AcAc-SP, (measured spectrophotometrically at 310 nm with 0.47 mM AcAc-SP, 6 mM potassium succinate, 0.5 mM EDTA, and 0.1 M Tris sulfate, pH 8.1, at 25 "C). The reaction followed firstorder kinetics for >7 half-lives with a rate constant of 0.  Table I, the reactivity of E-SP is similar to that of E-CoA. It reacts with specific substrates, succinate and acetoacetate, 5-10-fold slower and reacts with nonspecific reagents as fast (MMP) or up to 140-fold faster (water) compared with E-CoA. In the reverse direction, AcAc-SP reacts with CoA transferase 2 X 106-fold slower than AcAc-CoA. Thus, the equilibrium constant for the formation of E-SR from E and AcAc-SR is much less favorable for E-SP (Keq = kf/kr = 220/(3 X lo6) E 7 X than for E-CoA (Keq = 9) (15); the equilibrium constant for the reaction of SP with E-CoA is -8

As shown in
x When E-SP is compared to the nonspecific thiol ester E-MMP (6), the reactivity is increased by a factor of >5 X lo7 towards specific substrates and lo3 towards nonspecific reagents. However, E-SP is 1.9 kcal/mol less stable than E-MMP. These facts are summarized in the reaction coordinate diagram shown in Fig. 4A.
The noncovalent interactions between CoA transferase and CoA can therefore be split into two functional domains. The pantetheine moiety provides activation and destabilization, which are used to increase the reactivity of E-pantetheine to a level very close to that of E-CoA.  Table I  and nonenzymic thiol esters, as shown by the less favorable equilibrium constant for its formation (Fig. 4A). The remainder of the CoA molecule, 3'-phospho-ADP, provides 6.9 kcall mol of observed binding energy. This binding energy causes the large increase in kcat/KM for AcAc-CoA compared with AcAc-SP and the large stabilization of E-CoA relative to E-SP at equilibrium. Addition of 3',5'-ADP to the reaction of acetoacetylpantetheine with CoA transferase does not increase the chemical reactivity of E-SR, as measured by the slower hydrolysis rate of E-CoA compared with E-SP and the decrease in the hydrolysis rate of E-SP in the presence of 3',5'-ADP (Fig. 3). Thus, in order for the enzyme to utilize the binding energy of the 3'-phospho-ADP moiety it must be covalently attached to pantetheine. Fig. 4B shows the energy profiles of these reactions normalized to the energy of the transition state, which may have a similar structure for the three substrates. The left side shows that E-CoA and E-SP are activated, or destabilized, relative to E-MMP by 11.1-12.5 kcal mol", so that they can reach the transition state easily. However, on the right side it is apparent that AcAc-SP cannot reach the transition state easily because it lacks the 3'-phospho-ADP group that provides binding energy. The pantetheine group is highly reactive only after it has been forced into the active site; it was attached to the enzyme to prepare E-SP by a chemical reaction. However, the pantetheine group does also provide some binding energy to stabilize the transition state, in addition to destabilization in the ground state, because AcAc-SP is much more reactive than the short chain AcAc-MMP substrate (Fig. 4).
The nature of the substrate-induced rate acceleration in this system differs from that in phosphoglucomutase, which was shown by Ray and coworkers to transfer phosphate from the enzyme to the hydroxyl group of glucose 1-phosphate -lo1' faster than to water. Most of this rate increase can be brought about by binding of two separate molecules, xylose and phosphite, which results in phosphorylation that is 2 X lo9 faster than with water (16,17). In contrast, the different parts of CoA must be covalently linked so that the 3"phospho-ADP moiety can provide the binding energy to drive the activation by the pantetheine group.