Enoyl Coenzyme A Hydratase (Crotonase)

The substrate specificity of bovine liver crotonase has been examined with seven A29 3-frans-enoyl-CoA substrates, containing an even number of carbon atoms. The Vmax for this series decreases progressively from a value of about 340,000 moles per min per mole of enzyme for crotonyl-CoA, the Cq derivative, to 2,300 for the (216 derivative. The action of several CoA derivatives on crotonase has been tested. None were found to stimulate the enzyme and only one derivative, acetoacetyl-CoA, was found to be markedly inhibitory. Evidence was obtained that the enolate form of acetoacetyl CoA was the inhibitory species and acted as a competitive inhibitor with a KI of 1.6 X 10V6 M, a value about ten times lower than the Km for the best substrate, crotonyl-CoA. The interaction of acetoacetyl-CoA with crotonase was studied by ultraviolet difference spectroscopy and it was found that 6 molecules of inhibitor were bound per molecule of enzyme, or an average of one per subunit. This suggests that there are six active sites per molecule of native enzyme. The binding constant for the inhibitor was about equal to the kinetically determined value for Kr. The catalytic properties of crotonase have been compared with the turnover numbers and substrate specificities of the other enzymes acting in fi oxidation of fatty acyl-CoA derivatives. This comparison suggests that crotonase, by virtue of its substrate specificity and its sensitivity to feedback inhibition by acetoacetyl-CoA, may play a regulatory role in fatty acid oxidation. The effects of acetoacetyl-CoA on the rate of oxidation of butyric, octanoic, and palmitic acids by heart muscle or liver mitochondria were those expected if crotonase is acting, at least in part, to regulate fatty acid oxidation.


SUMMARY
The substrate specificity of bovine liver crotonase has been examined with seven A29 3-frans-enoyl-CoA substrates, containing an even number of carbon atoms.
The Vmax for this series decreases progressively from a value of about 340,000 moles per min per mole of enzyme for crotonyl-CoA, the Cq derivative, to 2,300 for the (216 derivative. The action of several CoA derivatives on crotonase has been tested.
None were found to stimulate the enzyme and only one derivative, acetoacetyl-CoA, was found to be markedly inhibitory.
Evidence was obtained that the enolate form of acetoacetyl CoA was the inhibitory species and acted as a competitive inhibitor with a KI of 1.6 X 10V6 M, a value about ten times lower than the Km for the best substrate, crotonyl-CoA.
The interaction of acetoacetyl-CoA with crotonase was studied by ultraviolet difference spectroscopy and it was found that 6 molecules of inhibitor were bound per molecule of enzyme, or an average of one per subunit. This suggests that there are six active sites per molecule of native enzyme.
The binding constant for the inhibitor was about equal to the kinetically determined value for Kr. The catalytic properties of crotonase have been compared with the turnover numbers and substrate specificities of the other enzymes acting in fi oxidation of fatty acyl-CoA derivatives.
This comparison suggests that crotonase, by virtue of its substrate specificity and its sensitivity to feedback inhibition by acetoacetyl-CoA, may play a regulatory role in fatty acid oxidation.
The effects of acetoacetyl-CoA on the rate of oxidation of butyric, octanoic, and palmitic acids by heart muscle or liver mitochondria were those expected if crotonase is acting, at least in part, to regulate fatty acid oxidation.
Crotonase, or enoyl-CoA hydratase (EC 4.2.1.17) is the only enzyme in the mitochondrial p oxidation pathway for fatty acids * This work was supported by research grants from the National Science Foundation (GB12676) and the National Institutes of Health (HE-06400).
1 Predoctoral Fellow, National Science Foundation. Portions of this work are contained in a thesis submitted to the Graduate School of Arts and Sciences, Duke University, in partial fulfillment of the requirements for the Ph.D. degree. Present address, Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, Connecticut 06510. which catalyzes the reversible stereospecific hydration of A2a3trans-enoyl-CoA substrates to the corresponding L( +)-fi-hydrosyacyl-CoA derivatives.
We wish to report here studies on the substrate specificity of bovine liver crotonase and its inhibition by acetoacetyl-Cob.
It has been found that the rate of hydration of A2~3-trans-enoyl-CoA substrates decreases markedly with increasing chain length.
Because acetoacetyl-CoA is a, potent competitive inhibitor of crotonase, it is possible that under conditions which allow acetoacetyl-Coa to accumulate, oxidation of long chain fatty acids would be reduced by virtue of inhibition of the crotonase-catalyzed step. Studies with intact mitochondria seem to support this view. For these reasons it is possible that crotonase may play a central role in fatty acid oxidation. Preliminary accounts of portions of this work have been presented previously (1, 2).

EXPERIMENTAL PROCEl)URE
Enzyme and Reagents-Crotouase, five times recrystallized, was prepared a$ reported earlier (3) and was shown to be homogeneous as described elsewhere (4).
Coenzyme A, acety-COB, acetoacetyl-Cob, and DPN were obtained from I' and L Laboratories.
Tetrahydrofuran (Mallinckrodt) was distilled fresh daily over sodium borohydride. Crotonic anhydride and ethyl chloroformate were products of Eastman.
Pure trans.or,/!-unsaturated free fatty acids from 6 to 16 carbon atoms in length were a gift from Dr. Salih Wakil. Bovine serum albumin, ATP, and dl-carnitine were obtained from Sigma.
1-14C-Labeled butyric, octanoic, and palmitic acids were obtained from New England Nuclear. Unlabeled butyric and octanoic acids were obtained from Sigma, and palmitic acid was a product of the Hormel Institute.
Labeled acids were diluted with unlabeled material to achieve a specific activity of approsimately 1 PCi per pmole.
Palmitic acid was solubilized by the addition of a small amount of ammonium hydroside and heating as reported earlier (5).
All other chemicals were reagent grade and were used without further purification.
After 10 min 35 Fmoles of ethyl chloroformate in 0.5 ml of tetrahydrofuran were added, and the reaction was allowed to proceed at 0" 5259 for 20 min. Precipitated triethylamine hydrochloride was removed by filtration through a Pasteur pipette fitted with a glass wool plug. The filtered solution was then evaporated to dryness, and the residue was dissolved in 1.6 ml of tetrahydrofuran.
The mixed anhydride solution was then added to the CoA4 under Ns in 0.5-ml aliquots at 3-to 5-min intervals at room temperature.
The pH was kept at 8 and distilled water was added as necessary to maintain a clear solution.
After 20 min the reaction was complete as judged by disappearance of -SH groups.
The solution was then adjusted to pH 5 with lYc perchloric acid, and most of the tetrahydrofuran was removed by evaporation. The Cl2 to Cl0 acyl-CoA derivatives precipitated on adjusting the solution to pH 3 with 10yc perchloric acid and were freed from residual unesterified acid by ether extraction.
The Cq to Cl0 enoyl-CoA derivatives are soluble in acid, and were purified from residual CoA by ion exchange chromatography on Whatman DE-52 using a lithium chloride gradient as previously described (8).
Preparation of Mitochondriu-Heavy beef heart mitochondria were a gift from Dr. Salih Wakil; they were obtained from the Institute for Enzyme Research, University of Wisconsin, and were kept frozen at -15" in 0.25 M sucrose until used. The mitochondria were then diluted with 0.25 M sucrose to the concentrations desired for assay.
Rat liver mitochondria were prepared by a procedure similar to that described by Johnson and Lardy (9). Rats weighing from 150 to 200 g were decapitated, and the livers were removed and placed in cold 0.25 M sucrose (Mann, special enzyme grade). All subsequent preparative procedures were carried out at O-4". The liver was cut into small pieces and washed several times with cold 0.25 M sucrose until the wash solution remained clear. The minced material was then put through a previously chilled garlic press to remove fibrous tissue and collected in 0.25 M sucrose (10 ml per g). This extract was gently homogenized (3 to 4 passes) and centrifuged at 1200 x g for 2 to 4 min at 0". The pellet was resuspended an recentrifuged.
The first and second supernatants were combined, centrifuged at 9000 rpm for 5 to 7 mm, and the supernatant was discarded.
The pelleted material was resuspended in an equal volume of sucrose, rehomogenized (2 to 3 passes), and recentrifuged.
This washing procedure was repeated twice more. The final mitochondrial pellet was taken up in a minimal volume of 0.25 M sucrose (3 to 4 ml) and was used immediately.
Protein concentrations were determined by the biuret method (IO).
Measurement of Hydrase Activity-The activity of crotonase with each enoyl-CoA substrate was measured spectrophotometrically at either 280 or 263 nm with a Cary model 15 recording spectrophotometer equipped with a water-jacketed cell holder maintained at 25". Assay mixtures contained 0.033 M Tris. HCl, pH 7.5, 5 x lo+ M EDTA, O.lc/; egg albumin, and substrate at the concentrations indicated under "Results." For routine assays, crotonyl-CoA was used at a concentration of 2 x 10e4 M. Initial velocity measurements were made with l-cm cells under conditions where absorbance changes were linear with enzyme concentration.
The decrease in absorbance for a 1 M solution of substrate was 6,700 and 4,400 at 263 nm and 280 nm, respectively (II). Concentrations of all substrates were measured spectrophotometrically assuming a molar extinction coefhcient of 19,200 at 232 nm.
DiJerence spectra with Crotonase and Acetoacetyl-CoA-These measurements were made by methods essentially identical with those reported earlier (12,13), employing tandem cuvettes with an over-all width of 2 cm (1 cm per compartment).
The concentration of acetoacetyl-CoA was measured spectrophotometritally at 260 nm (pH 7) assuming the molar extinction coefficient (e 2y, .,) = 15,400. The concentration of the enolate tautomer of acetoacetyl-Cob was measured spectrophotometrically at 307 nm assuming a molar extinction coefficient of 25,000. This value was obtained by spectrophotometric titration with either alkali or various divalent cations (14,15). According to the data of Stern (15), the enolate tautomer at pH 7.5 represents about 4cJ, of the total acetoacetyl-Cob in solution.

Mifochondrial
Fatty Acid Oxidation-The assay for fatty acid oxidation is similar to that described by Rressler and Friedberg (16). Reactions were performed at 37" with shaking in 25-ml flasks sealed with Kontes serum stoppers fitted with polyethylene center-well inserts. The cups were suspended at a level below that of the water level of the bath to minimize condensation effects.
Reaction mixtures were composed of Krebs-Ringer phosphate (0.2 ml of a fresh, five times aerated solution), 20 mg of bovine serum albumin, 10 pmoles of ATP, 6.25 pmoles of DPN, 4 pmoles of dl-carnitine, 1 pmole of coenzyme A, 100 mpmoles of l-nCfatty acid substrate and varying amounts of mitochondria (10 to 20 mg of protein) per 1.0 ml. Reaction was initiated by iiijection of the mitochondrial solutions into the system and subsequently terminated by the injection of 0.5 ml of 2 s I&SO4 into the reaction mixture.
Following acidification 0.2 ml of 10-X hyamine hydroxide (Packard) was injected into the rup insert.
The r4C02 derived from fatty acid oxidation was then collected for subsequent liquid scintillation measurements by shaking for 1 hour at 37". Control assays were carried out in similar fashion without the addition of mitochondria.

Action of Crotonase on A 2~3-Enoyl-CoA
Substrates- Fig.  1 shows the relationship between the initial velocity and the substrate concentration for seven A2s3-enoyl-CoA substrates, each containing an even number of carbon atoms.
Clearly, the rate of hydration decreases with increasing chain length.
Crotonyl-CoA is the best substrate and the turnover number with this substrate has been calculated to be 340,000 moles per min per mole of enzyme. In contrast the turnover number with hexadecenoyl-Co4 is 2,300 moles per min per mole of enzyme.
It is noteworthy that this latter value falls within the range of turnover numbers displayed by the other enzymes of fatty acid oxidation as will be discussed later.
The V,,, and K, values for each substrate were calculated frorn double reciprocal plots and are summarized in Table I. The relationships between chain length of the substrates and these kinetic parameters are shown in Fig. 2 3. The effect of acetoacetyl-CoA on crotonyl-CoA hydration.
Absorbance changes dlle to enzyme-inhibitor complex formation were negligible at this wave length and protein concentration (Fig. 5) In view of the potent inhibitory effects of acetoacetyl-CoA, the inhibition by equilibrium mixtures of crotonyl-Coh and flhydroxybutyryl-CoA were tested with longer chained enoyl-Cal substrates.
These equilibrium mixtures proved to be good inhibitors as shown in Fig. 4 with octenoyl-CoA as substrate. The apparent K1 for the equilibrium mixture was calculated to be 2 x 10-j M, or about equal to the K, for crotonyl-Cob (Table  I).
Interaction Reaction mixtures were prepared with crotonase as described in Fig. 1  the interaction of crotouase with acetoacetyl-CoA has been esamined by ultraviolet difference spectroscopy. Fig. 5 shows the difference spectrum of the inhibited enzyme, which was obtained by measuring the absorbance of mixtures of acetoacetyl-Coh and crotonase against solutions of iuhibitor and enzyme at the same concentrations but in different cells in the reference beam of the spectrophotometer.
This spectrum, aside from the shoulders at 278 and 292 nm resulting from the perturbation of aronlatic residues, is indistinguishable from that of the enolate form of acetoacetyl-CoA (14). This suggests that the enolate tautomer of acetoacetyl-CoA is the inhibitory species and has a marked affinity for the enzyme. At pH 7.5, 25", the enolate form represents about 4% of the total acetoacetyl-Cob in solution (15). Recalculation of the KI for acetoacetyl-CoA (Fig. 3) based on the concentration of the enolate gives a KI of 1.6 X 1O-6 31, a value about 10 times lower than the K, for the best substrate, crotonyl-CoA ( Table I). Because of the large spectral change associated with the binding of the enolate tautorner of acetoacetyl-Cob, it was possible to titrate crotonase with acetoacetyl-CoA. Fig. 6 shows the titration expressed as the amount of enzyme-iuhibitor complex formed as a function of inhibitor concentration.
The esperimentally determined values correspond very closely to the theoretically calculated curve for a noncooperative reaction with a dissociation constant of 1.7 X 10-j M. Thus, the dissociation constant for acetoacetyl-CoA is very similar to its kinetically determined Kr (3 X low5 M).
After correcting for the concelltration of the enolate tautomer, the dissociation constant for the reaction as calculated from these data is 0.7 x 1O-G M, a value in close agreement with the kinetically determined RI for the enolate form (I .6 X 10m6 M). This shows that crotonase not only has approximat,ely the same affinity for acetoacetyl-CoA over a million-fold range in protein concentration, but also FIG. 5 (left FIG. 7 (right).
Determination of the number of binding sites of crotonasc.
The equilibrium binding data obtained as described in Fig. 0 Fig. 6 can be used to determine the number of molecules of inhibitor bound per molecule of enzyme. When these data are expressed as shown in Fig. 7, it has been found that 6 f 0.1 molecules are bound per molecule of crotonase. This suggests that there is one independent binding site, presumably the catalytic site, on each of the 6 subunit polypeptide chains in the enzyme.
E.ffect of Acetoacetyl-CoA on Mitochondrial Fatty Acid Oxidation--Because acetoacetyl-CoA is a potent inhibitor of crotonase and it is a major end product of fatty acid oxidation, the The conditions are the same as above in Fig. 8 except that freshly prepared liver mitochondria (15 mg per reaction mixture) was used. The symbols used for each substrate are the same as in Fig. 8.
effect of acetoacetyl-CoA on the mitochondrial oxidation of fatty acids was examined. Fig. 8 shows the rate of oxidation of butyrate. octanoate, and palmitate by beef heart muscle mit'ochondria in the presence and absence of acetoacetyl-CoA.
The rates are expressed as the amounts of K!O2 derived from the 14Ccarboxyl labeled acids. Clearly, the rate of oxidation of butyrate and octanoate is unaffected by acetoacetyl-Cob although the rate of palmitate oxidation is markedly depressed by acetoacetyl-CoA. The effect of concentration of acetoacetyl-CoA on palmitate oxidation is shown in Fig. 9. Acetoacetyl-CoA has similar effects in liver mitochondria as shown in Fig. 10, although the amount of inhibition of palmitate oxidation is not as great as found with heart mitochondria. DISCUSSION Substrate Specificity o./ Crotonase-Earlier studies with crotonase suggested that it had a broad substrate specificity.
Crotonyl-CoA, hexenoyl-CoA, and the P-hydroxyacyl-CoA derivatives cont#aining, 4, 6, 8, 9, and 12 carbon atoms were found to be good substrates (21: 22). It appears to have a strict requirement for thiol esters of CoA, and although it can hydrate crotonylpantetheine (22), the rate of hydration is only about 0.01 that of crotonyl-Cob.
Crotonase also displays a broad specificity in terms of the stereochemistry of the A2J double bond, since it hydrates the trans-as well as cis-enoyl-CoA derivatives (23). The studies reported here extend our knowledge of the substrate specificity of crotonase and show that it acts on A213trans.enoyl-Cob substrates ranging in chain length from 4 to 16 carbon atoms, although the rate of hydrat,ion falls markedly as chain length is increased (Fig.  1, Table  I).
In view of the possible importance of this observation it was essential to consider whether the longer chain enoyl-CoA substrates formed micelles which could influence the ' observed kinetic parameters.
Uthouph the critical micelle concentrations for the enoyl-CoA derivatives used here are unknown, it is unlikely that micelles influenced the rate studies for the following reasons. First, micelle formation would be erpected to alter substrate binding, and thus K,, more than the rate of hydration (I',,,).
Just the ol)posite was found; K, varied only slightly from Ce to C16, whereas V,,, decreased progressively with increasing chain length.
Secondly, no abrupt transitions were found in the kinetic data as a function of substrate concentration.
If monomer-micelle transitions occurred they should be reflected by nonlinear double reciprocal plots.
Finally, P-hydroxyacyl dehydrogenase (24) and thiolase (25) act equally well on long or short chained CoA derivatives and if micelles were formed with substrates for these enzymes, they had little apparent effect on ki-lIetics of the enzymes. inhibition by ilcetoacetyl-Co/l-Spectroscopic examination of the reaction of acetoacetyl-CoA with crotonase indicated several important points. First, the difference spectrum obtained in Fig. 6 is almost indistinguishable from that of the enolate form of acetoacetyl-CoA and it can be concluded that this is the species interacting with crotonase. It is noteworthy that the enolate is a potent competitive inhibitor and structurally resembles both crotonyl-and fl-hydroxybutyryl-Coh.
For these reasons, the enolate can be considered a transition state analog for crotonase (26). Thirdly, titration of the enzyme with acetoacetyl-CoA shown that an average of 1 molecule of inhibitor is bound per enzyme subunit.
Since the inhibitor is competit.ive it is likely that there is one active site per subuttit..
In addition, there is no indication of evidence for cooperatirity in the binding of acetoacetyl-('oL4, in accord with the fact that crotonase shows no characteristics of an allosteric enzyme (1, 22). Crotonnse and Acetoacetyl-CoS in Regubation of Fatty Acid Oxi-&ion-When the substrate specificity of crotonase and the in-Iribitory effects of acetoacetyl-Co-4 on crotonase are considered in view of the properties of the other enzymes in fatty acid oxidation, it is evident that crotonase and acetoacetyl-Coh may l)l:ty an important role in the regulation of fatty acid oxidation. 21, consideration of the properties of the enzymes in fatty acid oridation which are listed in Table II illustrates this point. As noted earlier (17), and confirmed in these studies, crotonase appears to be the only hydratase of fatty acid oxidation.
In con- having different chain lengt,h specificities but with equivalent optimal activities. The other two enzymes, the P-hydrosyacyl dehydrogenase and thiolase, exist as a single species just as crotonase, but, unlike crotortase, they act almost equally as well on all substrates irrespective of chain length.
Thus, rrotonase is unique among the enzymes of fatty acid oxidation because its rate of hydration of substrates differing in chain length from Cd to Clc varies about 150.fold.
At present the other enzymes in fatty acid osidation, including the electron transport flaroprotein, are Ilot recognized to possess regulatory properties as evidenced by their catalytic behavior or sensitivity to metabolites. Finally, the uniqueness of crotortase in fatty acid osidat,ion is most evident when the turnover numbers for all of the enzymes of fatty acid oxidation are compared as a function of the chain length for their substrates, as shown in Fig. 11. The turnover numbers for each of the enzymes, except crotonase, fall between about 3.5 to 9 pmoles of substrate per min per g of liver, as in-  From these considerations, it is striking that the turnover for the Cd to Cl4 enoyl-Co.4 substrates is considerably greater than those for all other acyl-CoA substrates of the same chain length.
Only hesadecenoyl-Cob is hydrated at a rate about equal to those for transformation of other acyl-CoA substrates. Fig. 11 also shows the expected t'urnover numbers for each of the substrates of crotonase in the presence of a IO-fold excess of acetoacetyl-Co-4 over the enoyl-Co.4 substrate.
These values are indicated by the solid vertical bars. This shows that accumulation of acetoacetyl-CoA would reduce the rate of hydration of enoyl-CoX substrates with chain lengths between Cd and CIZ, but under these conditions the cro-tonase-catalyzed reactioll is not rate limiting. In contrast, acetoacetyl-Coh accumulation would markedly depress the turnover of Cl4 aud Cl6 elloyl-Coh substrates 2nd under these conditions would limit the rate of fatt,y arid oxidation.
'I'llll~, by a combination of its cascading substrate specificitBy and its marked susceptibility to acetoacetyl-Co& crotonase would be rate limiting in its action on long chain substrates, \yhich are used in initiation of fatty acid oxidation, whereas it would COIItinue to hydrate shorter cshained (C, to C,,) substrates at rates equal to or greater than those for the other reactions in /3 osida-Gon. These conditions would be expected to allow t,he short and medium chain length ncyl-CoL4 intermediates of fatty acid oxidation to be metabolized to acetylLCo;2 at a steady rate and they would not be expected to accumulate in the mitochondria. The proposed regulatory role of crotonase as a consequence of its cascading substrate specificity and sensitivity to feedback inhibition is indicated at different stages of fatty acid oxidation.
by guest on March 24, 2020 http://www.jbc.org/ Downloaded from itrllibit hydratiorr of longer chain er~oyl-Co.1 substrates (Fig. 4) I<I<FEEENCJ~:S about as well as acetoacetyl-Co& Were these CoA intermedi-(>oh, but the rates of butyrate and octanoate oxidation are un-:tffect,ed (Figs. 8,9,and 10). It was impossible to assess whether the wcetoacetyl-CoA was indeed inhibiting oxidation at the level of the crotouase-catalyzed step iu the mitochondria, but the rffect,s observed are in qualitative agreernent with the proposed regulatory scheme (Fig. 12). Secondly, acetoacetyl-Cob is an r11t1 product of fatty acid oxidation and its inhibitory effects may be considered as a type of feedback iuhibition as found in other regulatory lrrocesses ill met~abolism.
Third, the concentration of :tcet,oacetyl-CoA required to inhibit hydration of enoy-CoQL sihstrate,s is very low. It calI be considered a transition state :111alog inhibitor of crotouase since its RI is about 10 times lags than the I<, for crotonyl-CoA and about 500 times less than the T<T,2 for lies:Ldecelioyl-CoA.
Although the normal intrarnito-c~lrorrdrial corrcentratiorl of acetoacetyl-CoA is unknown, its itrlribitory effects on osidation of palmitate are readily apparent n-hen irrta(*t rnitochondria are exposed to levels as low as 0.1 m&l ( Fig. 9) It is reasonable to assume that t,he irrtrarnitochondrial c~oncerrtratiorrs may well be lower than this in the experiments ~lrow~r irr Figs. 8 to 10 :uld are unlikely to elcseed the conceutra-tiorr5 of lralmitate.
Firral!y, the scheme proposed in Fig. 12 c~oultl also possibly account for the fat+ that fatty acids of intertrrrtliate chairr length do rrot :~ccunrulatr in rnitochondria.
The rate of oridatiorl of lorry chain (greater tharr C,,) substrates will IX' c~oritrolled at the crotonase step iii the presence of acetoacetpl-('0.1. 'I'his could limit entry of 1011% chain fatty acids into the ositlation pathway. 13ut ouce fatty acid CoA intermediat,es -1iorter thari Cl4 are formed, they could be expected to be tlr~raded at :I steady state. Further studies will be required to test these lroirrts. .~~1;17oz~ledgmenfs-~~-e thank 1 jr. Robert Barker for his advice a~rtl sugpestious throughout the course of this study. We also \visll t,o thank Dr. S. J. Wakil alIt 1)r. TV. S. La-1111 for their advice