T H E JOURNAL of BIOLOGICAL CHEMISTRY Vol 255, No. 16, Issue of August 25, pp. 7858-786’2, LYW Prrnted in 1.IS.A.

S-Acetonyl-CoA A NONREACTIVE ANALOG OF ACETYL-CoA*

(Received for publication, December 5, 1979, and in revised form, May 1, 1980)

Peter Rubenstein and Robert Dryer From the Department of Biochemistry, Cotlege of Medicine, University of Iowa, Iowa City, Iowa 52242

We have synthesized S-acetonyl-CoA from CoASH and 1-bromoacetone. This thioether-containing struc- tural analogue of acetyl-coA is a potent competitive inhibitor, with respect to acetyl-coA, of citrate syn- thase, phosphotransacetylase, and carnitine acetyl- transferase. This analog will not activate Escherichia coli phosphoenolpyruvate carboxylase or rat liver py- ruvate carboxylase, two enzymes which require acetyl- CoA as an obligate activator. Furthermore, acetonyl- CoA will not compete with acetyl-coA for binding to these enzymes showing the apparent absolute require- ment of these two enzymes for a thioester group on the activating ligand. S-Acetonyl-CoA should be a useful reagent in the investigation of acetyl-CoA-requiring proces-es.

For current studies on the acetyl-CoA-dependent acetyla- tion of actin in an in vitro protein-synthesizing system we required a means of inhibiting as completely as possible the acetylation of nascent polypeptide chains. Experience with a previously published procedure (1) based on decreasing en- dogenous concent.rations of acetyl-coA by addition of citrate synthase plus oxaloacetate indicated this method was inade- quate for our needs.

Enzymes which require acetyl-coA can be grouped into three classes. The first includes those like citrate synthase and acetyl-coA carboxylase which use the thioester group of the acetyl-coA to activate the acetyl methyl carbon for carb- anion formation and subsequent attack on the electron-defi- cient carbon atoms. The second includes acetyltransferases such as phosphotransacetylase, protein transacetylase, and carnitine acetyltransferase, which transfer the acetyl moiety by means of nucleophilic attack at the thioester moiety of acetyl-coA. The third class is comprised of enzymes such as bacterial phosphoenolpyruvate carboxylase and mammalian pyruvate carboxylase which require acetyl-coA not as a sub- strate but as an obligate activator. We reasoned that an analog of acetyl-coA lacking the thioester group might be extremely useful not only as an agent for blocking nascent polypeptide chain acetylation but also as a probe of the mechanistic properties of the enzymes of these three classes. We, therefore, have synthesized the thioether-containing compound S-ace- tonyl-CoA (Fig. 1). We report here that S-acetonyl-CoA effec- tively blocks acetyl group transfer in several enzyme systems tested, and we also show that this compound is inert in two enzyme systems requiring acetyl-coA as an obligate activator.

* This work was supported by Grant GM24702 from the National Institutes of Health to P.R. and by a grant from the American Diabetes Association, Iowa Affiliate, to R.D. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “adoertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

EXPERIMENTAL PROCEDURES

Materials Pyruvic acid, L-carnitine hydrochloride, acetyl-coA, CoASH, oxal-

oacetic acid, malate dehydrogenase, aspartate aminotransferase, and phosphoenolpyruvate carboxylase were purchased from Sigma Chem- ical Co., St. Louis, MO. ~-[U-’~C]Aspartic acid, 180 mCi/mmol, and [”H]NaBH4, 250 mCi/mmol, were obtained from New England Nu- clear. Cellulose thin layer plates, 100 p thick with fluorescent indicator were purchased from the Eastman Kodak Co., Rochester. NY. Trit- ium-sensitive Ultrafiim was obtained from LKB, Piscataway, NJ.

Monobromoacetone was synthesized by the method of Catch et al. (2). The product was distilled a t 63.5-64.OoC/50 mm as a water-clear liquid and stored in the dark at -20°C. Storage for up to 2 months did not lead to any overt resinification or change in color.

Rat liver pyruvate carboxylase was prepared according to Scrutton et al. (3) through the 28% saturated ammonium sulfate extract of their Stage 4 preparation.

Methods

lished method based on the chromogen, 6,5‘-dithiobis(2-nitrobenzoic Enzyme Assays-Citrate synthase activity was assayed by a pub-

acid) (4). Phosphotransacetylase was assayed by means of the de- creased absorbance a t 232 nm concomitant with acetyl-coA hydrol- ysis (5, 6). Phosphoenolpyruvate carboxylase (7) and pyruvate car- boxylase (8) were assayed in coupled systems with malate dehydro- genase by following the decrease in absorbance a t 340 nm due to reduction of enzyme-generated oxaloacetate by NADH.

The ability of citrate synthase to catalyze formation of a covalent adduct between oxaloacetate and acetonyl-CoA was measured as follows. [14C]Oxaloacetate was generated from ~-[‘~C]aspartate by aspartate aminotransferase. The oxaloacetate was then combined with either acetyl-coA or acetonyl-CoA in the presence of citrate synthase for 30 min a t 37°C. The reaction products were analyzed by autoradiography of cellulose thin layer chromatograms developed in 1-butanol/acetic acid/water solvent (4:1:1, v/v).

Spectrophotometry-Spectral measurements were performed on a Beckman model 35 scanning spectrophotometer.

from initial rate measurements using the HYPER and COMP com- Kinetic Parameters-Kinetic constants K,, and K, were calculated

puter programs of Cleland (9). Synthesis and Characterization of S-Acetonyl-CoA-Coenzyme A

(200 pmol) was dissolved in 15 ml of freshly degassed ice-cold water with magnetic stirring. T o this was added dithiothreitol (20 pmol) to ensure that all of the coenzyme was in the reduced form. The pH was brought to 8.0 to 8.2 with NaOH. Alkylation was performed by addition of monobromoacetone (250 pmol) dissolved in 5 ml of 95% ethanol just prior to use. Disappearance of free ”SH groups was

0 a. c ~ A - s - C - C H ~

0 b. CoA-S-CHZ-C-CHa

FIG. 1. Structural formulas of S-acetyl-coA (a) and S-ace- tonyl-CoA (b).

7858

S-Acetonyl-CoA: A Nonreactive Analog of Acetyl-coA 7859

TABLE I Disappearance of DTNB reactivity on alkylation of CoASH with

bromoacetone Bromoacetone added All:

pmol

0 0.70 50 0.54

100 0.41 150 0.25 200 250

0.11 0

Absorbance was read against a DTNB blank as described in the text. Including dithiothreitol sulfhydryl groups, 240 pmol of "SH groups were present at the initiation of the reaction.

60

40

20

0

L

-0 20 40 60

FRACTION NUMBER

FIG. 2. Sephadex G-15 chromatography of acetonyl-CoA and diacetonyl dithiothreitol. Acetonyl-CoA was prepared as de- scribed in the text, and the lyophilized residue was dissolved in 1.5 ml of distilled water and applied to a column (1.5 X 45 cm) of Sephadex G-15 equilibrated with water. Fractions of 1.5 ml were collected, and the acetonyl-CoA was located by its absorbance a t 260 nm. In a similar reaction, dithiothreitol was reacted with a 2-fold excess of bromoacetone. The mixture was lyophilized, dissolved in 1.5 ml, and applied to the same column. Fractions of 1.5 ml were collected, and the alkylated dithiothreitol was located by following its reaction with KMnOl a t 576 nm. A, acetonyl-CoA; B, diacetonyl dithiothreitol.

monitored by adding 5 p1 of the reaction mixture to 1.00 ml of 0.1 mmol/liter of DTNB' adjusted to pH 8.5 with 50 mmol/liter of Tris buffer. The absorbance a t 412 nm was followed in a spectrophotom- eter. Table I shows that an amount of bromoacetone equal to the total free sulfhydryl content quenched the DTNB reactivity of the starting material. The reaction with each aliquot of bromoacetone was usually complete in 1 min or less. The product, after removal of solvent and any excess bromoacetone by lyophilization, was dissolved in 1.5 ml of water and chromatographed on a column (1.5 X 45 cm) of Sephadex (2-15 equilibrated with water. Eluant fractions containing the highest concentration of material absorbing at 260 nm were pooled. Fig. 2 shows that the Sephadex column used completely separated the S-acetonyl-CoA from diacetonyl dithiothreitol, a by- product of the reaction. Analysis of the product on cellulose thin layer plates developed in 1-butanol/pyridine/acetic acid/water solvent (50: 33:1:40, v/v), Solvent A, showed a single UV-absorbing spot with an RF of 0.41. If oxidized CoA (RF 0.19) was present, it accounted for less than 10% of the total UV-absorbing material. S-Acetonyl-CoA, CoA, and dephospho-CoA were applied to a polyethyleneimine cellulose plate, and the plate was developed in 0.3 M LiCI. CoASH and acetonyl- CoA chromatographed with identical mobilities (RF 0.05) while de- phospho-CoA moved with an RF of 0.37 (Fig. 3). This result shows that acetonyl-CoA contains all the phosphates present in the original CoA. Occasionally, the product was further purified by thin layer chromatography on cellulose using Solvent A, elution with water, and a second gel filtration. No effect of this additional step on the ability of acetonyl-CoA to interact with the enzymes tested could be ob- served. The UV spectra of free CoA and of acetonyl-CoA were identical between 235 and 300 nm. For calculating acetonyl-CoA concentrations we used an extinction coefficient, E = 15.4 x IO' M" cm" a t 260 nm. This is identical with the value for CoASH. Judged

I The abbreviation used is: DTNB, 5,5"dithiobis(2-nitrobenzoic acid).

by reactivity with DTNB, there was less than 0.1% contamination of the product by the free coenzyme.

Elemental analysis (Galbraith Laboratories, Knoxville, TN) showed that N, P, and S were present in the ratio 7:3:0.86 (theoretical ratio, 7:3:1). Bromine analysis showed a ratio of Br to S of less than 1:80, indicating that bromine was substantially removed from the preparation. Yields, as judged by final recovery of material with a

F

FIG. 3. Polyethyleneimine cellulose chromatography of ace- tonyl-CoA, CoA, and dephospho-CoA. Acetonyl-CoA, CoA, and dephospho-CoA were applied to a polyethyleneimine cellulose plate (Baker) and developed with 0.3 M LiCI. After drying, the plate was sprayed with a 0.1% (w/v) solution of fluorescein in methanol. and the compounds were visualized under a UV lamp. A, acetonyl-CoA; B, CoA; C, dephospho-CoA.

1

l I

l t I

Fractlon Number

FIG. 4. Reduction of acetonyl-CoA with tritiated sodium borohydride. Twenty nanomoles of either acetonyl-CoA (--) or CoA (- - -) were reduced with [,"HJNaBH, as described under "Ex- perimental Procedures." The reaction mixtures were applied to a 2.0- ml column of Sephadex G-15 equilibrated in water. The products were eluted with water, and fractions of one drop were collected and analyzed for tritium by standard liquid scintillation counting proce- dures. Under these conditions both CoA and acetonyl-CoA are eluted in the void volume. The elution profiles of the reduction products of each of these compounds are shown here.

7860 S-Acetonyl-CoA: A Nonreactive Analog of Acetyl-coA

maximum absorbance at about 260 nm, were routinely greater than 75%.

Digestion of Acetonyl-CoA with Snake Venom Nucleotide Pyro- phosphatase-Acetonyl-CoA and CoASH were separately digested for 2 h at 30°C with a nucleotide pyrophosphatase-containing extract prepared from Crotalus adamanteus venom (10). Aliquots of each reaction mixture (10 pl) were applied to cellulose thin layer plates and developed in Solvent A described above. Both digests gave single UV-absorbing spots with identical RF values, 0.23. These data indicate that the acetonyl-CoA preparation contained a 3’,5”adenosine di- phosphate moiety, with the 5‘-phosphate still in pyrophosphate link- age to the phosphopantetheine moiety of the molecule.

Reduction of Acetonyl-CoA with [3HJNaBH.4-Formation of ace- tonyl-CoA should introduce a borohydride-reducible keto group not present in the original CoASH. To test this prediction, samples (20 nmol) of either acetonyl-CoA or of CoASH were incubated in 30 pl of 50 mmol/liter of NaHCO:J, pH 8.0, containing 40 nmol of [”H]NaBH4, 250 mCi/mmol, for 2 h a t 0°C. Water was then added (20 pl). Each reaction mixture was next applied to a 2-ml Sephadex G-15 column in a 3-ml disposable syringe. Fraction6 of 1 drop (~0.05 ml) were collected and counted by liquid scintillation spectrometry in Aquasol cocktail (New England Nuclear) (see Fig. 4). A similar sample was applied to a cellulose thin layer plate and developed in Solvent A. The dried chromatogram was used to generate an autoradiogram using tritium-sensitive x-ray film. Most of the radioactivity (90%) moved as a single spot with an RF of 0.43, which coincided with a single UV-absorbing spot. When compared with the parent acetonyl- CoA, on the same plate, the reduced isopropanolyl-CoA migrated slightly but consistently further from the origin. A second labeled species comprising 10% of the applied radioactivity was also observed. This may correspond to the sulfoxide of acetonyl-CoA.

Digestion of [3HJZsopropanolyl-CoA with Nucleotide Pyrophos- phatase-[:‘H]Isopropanolyl-CoA (20 nmol) was digested with nu- cleotide pyrophosphatase as described above. The reaction mixture was then chromatographed on a cellulose thin layer plate, as already

- F

- -x

IC- m

c “ A P 2

- 0

- A- ”

A B FIG. 5. Digestion of the tritiated product of acetonyl-CoA

with snake venom nucleotide pyrophosphatase. [“HIS-Isopro- panolyl-CoA, the product of the reduction of acetonyl-CoA with [’HINaBH,. was then digested with snake venom nucleotide pyro- phosphatase as described under “Experimental Procedures” by the procedure of Wang et al. (IO). The compound before and after digestion was chromatographed on a cellulose thin layer plate in Solvent A. Coenzyme A was used to generate adenosine 3’,5”diphos- phate. Autoradiograms, shown here, were then made using tritium- sensitive x-ray film (LKB). A. before digestion; B, after digestion. 0, origin; F, solvent front; IC, isopropanolyl-CoA, X, the tritiated cleav- age product, presumably S-isopropanolyl4-phosphopantetheine. En- zymatic digestion of isopropanolyl-CoA liberated a UV-absorbing, nonlabeled compound that comigrated with adenosine-3’,5”diphos- phate (AP2). The minor component represents about 104% of the total radioactivity as determined by scanning densitometry. This may be the sulfoxide of the thioether.

I I

I I ‘I

H RNCH,CH,SH

S

I 1 ’

. .. , . . , . , I I I , , . . I . , . ” .. 1 , . . I

200 150 100 5 0 25

FIG. 6. I3C-NMR spectra of acetonyl-CoA and CoASH. P rep- resents the thioethanolamine carbon adjacent to the N, Q the thio- ethanolamine carbon adjacent to the S. Ad-3 represents C:J of the adenine ring. The resonances between 150 and 155 ppm are contrib- uted by adenine carbons 2.4, and 6. S represents the internal standard, dioxane. In the upper curve M represents the methyl carbon and C the carbonyl carbon of the acetonyl moiety. These peaks are absent from the CoASH spectrum. See text for further details.

outlined, and an autoradiogram was prepared. The radiolabel now moved as a non-UV-absorbing spot with an RF of 0.64. An unlabeled but UV-absorbing material which co-migrated with 3’,5”diphos- phoadenosine was also produced (Fig. 5).

’3C-NMR Spectra of CoASH and AC~LO~~L-COA-’~C-NMR spec- tra of acetonyl-CoA and CoASH were taken using 100 mmol/liter- concentrations of each compound in D20 on a Brucker HXWE instrument a t 22.634 MHz. Dioxane was used as an internal standard. For obtaining the spectra, 27,263 transients were ,recorded at 22°C. The spectra are shown in Fig. 6.

The spectrum obtained for CoASH is identical with that previously reported (Ref. 11).* For acetonyl-CoA, a carbonyl carbon peak ap- peared at 210 ppm and a new methyl group at 28.7 ppm. Assignments of these peaks were based on the known ‘“C-NMR spectrum of acetone. The resonance of the carbon atom adjacent to the sulfur of the thioethanolamine group shifted downfield 10 ppm in the acetonyl- CoA spectrum. The resonance of the other thioethanolamine carbon shifted 5 ppm in the upfield direction upon reaction of CoA with bromoacetone. Identical shifts were observed when the spectrum of CoA disulfide was compared with that of CoA.” These results show that alkylation of CoASH by bromoacetone had occurred at the sulfur atom. Small changes (2 to 3 ppm) in three of the adenine carbon resonances were observed upon alkylation. However, we believe, for two reasons, that this shift is due to a conformational change in the molecule rather than ring alkylation. First, the amount of alkylating agent needed to quench the reaction of CoA with DTNB was that theoretically required and not a 2-fold excess. Second, although the positions of the adenine carbon resonance shifted slightly after alkyl- ation, additional adenine resonances, which would indicate partial adenine alkylation, were not observed.

RESULTS

The reaction between free CoASH and monobromoacetone to yield S-acetonyl-CoA (Fig. 1) proceeded smoothly and rapidly. The disappearance of DTNB reactivity indicated that the desired alkylation was occurring at the terminal “SH group, and the fact that cleavage of the product by nucleotide

H. Hogenkamp, unpublished results. H. Hogenkamp, personal communication.

S-Acetonyl-CoA: A Nonreactive Analog of Acetyl-coA 7861

pyrophosphatase liberated 3’,5‘-adenosine diphosphate dem- onstrated that no significant alkylation of the adenine ring occurred. Introduction of tritium into acetonyl-CoA by [‘HI- NaBH, reduction, without a similar labeling when free CoASH was used, demonstrated that a keto group was present in the reaction product resulting from the condensation of CoASH and bromoacetone (see Fig. 4). Furthermore, the tritium was contained in the non-nucleotide moiety, presum- ably S-isopropanolyl-4-phosphopantetheine, adding weight to the argument that the keto group is at the thiol end of the parent molecule. These data, as well as the I3C-NMR spectra of the alkylated CoA and CoASH, indicate that the compound we have synthesized is S-acetonyl-CoA.

Interaction of Acetonyl-CoA a n d Isopropanolyl-CoA with Citrate Synthase-Acetonyl-CoA was found to compete effec- tively with CoA for binding to citrate synthase as shown in Fig. 7. Analysis of the data yielded for acetyl-coA a K , of 7.0 pmol/liter. In comparison, K, for acetonyl-CoA was 8.78pmol/ liter indicating the affinity of the acetonyl analog for the enzyme was approximately equal to that of the natural sub- strate. However, reduction of the carbonyl oxygen with so- dium borohydride (Table 11) caused a 9- to 10-fold increase in the K, value of the analog although the inhibition was still competitive (data not shown). Citrate synthase is thought to act by causing carbanion formation at the methyl carbon of the acetyl moiety of acetyl-coA. Such a reaction is theoreti- cally possible with the acetonyl analog especially in view of its high affinity for the enzyme. When an excess of S-acetonyl- CoA was combined with [‘4C]oxaloacetate in the presence of citrate synthase (see under “Experimental Procedures”), no condensation product was observed. Under identical condi- tions at an equivalent acetyl-coA concentration, all of the oxaloacetate was converted to citrate (data not shown).

Interaction of Acetonyl-CoA with Two Acetyltransfer- ases-As discussed previously, a second class of enzymes utilizing acetyl-coA transfers the acetyl moiety via nucleo- philic attack at the thioester group. For model enzymes we used phosphotransacetylase and carnitine acetyltransferase. Phosphotransacetylase normally catalyzes acetyl-coA for- mation from acetyl phosphate and coenzyme A. The enzyme

0.4

/

1 I I I I I I I I

0 .02 04 .06 .08

l/[ Acetyl CoA 1 (pM-1) FIG. 7. Inhibition of citrate synthase by acetonyl-CoA. The

formation of citrate catalyzed by citrate synthase in the presence of varying amounts of acetonyl-CoA was studied. Assays were carried out as described in the text. Lineweaver-Burk plots of the data are shown. t”., no acetonyl-CoA; M, 10 pmol/liter of acetonyl- CoA; D”-I7, 20 pmol/liter; A-A, 40 pmol/liter.

TABLE I1 Interaction of Acetonyl-CoA with pig heart citrate synthase

Assays were performed as described under “Experimental Proce- dures.’’ Isopropanolyl-CoA was prepared by reduction of acetonyl- CoA by sodium borohydride as described under “Experimental Pro- cedures” except that unlabeled borohydride was used. This data represents the combination of two experiments using separately pre- pared batches of isopropanolyl-CoA.

Compound K , K , pmol/ltler

Acetyl-coA 7 f 0.4 Acetonyl-CoA 8.8 f 0.6 Isopropanolyl-CoA 9of 12

TABLE I11 Kinetic parameters of the interaction of acetonyl-CoA with two

acetyltransferases Enzyme K,, acetyl-coA K, acetonyl-CoA

pmol/liter

E. coli phosphotransacetylase“ 60 f 2 125 f 7 Pigeon breast muscle carnitine 62 f 4 152 f 3

acetyltransferase’ Phosphotransacetylase was assayed as described under “Experi-

mental Procedures.” Acetyl-coA concentrations ranged from 40 to 400 pmol/liter. Acetonyl-CoA concentrations ranged from 0 to 200 pmol/liter.

Carnitine acetyltransferase was assayed as described under “Ex- perimental procedures.” Acetyl-coA concentrations ranged from 20 to 200 pmol/liter. Acetonyl-CoA concentrations ranged from 0 to 200 pmol/liter.

TABLE IV Effect of acetonyl-CoA on E . coli phosphoenolpyruvate carboxylase

Activator Concentration” Rate“ pmol/hter pmol/min

Acetyl-coA 120 0.028 Acetyl-coA 80 0.017 Acetyl-coA 40 0.009 Acetyl-coA 0 0.OOO Acetonyl-CoA 120 0.001 Acetonyl-CoA 100 0.009

+ acetyl-coA 40

” Concentration of acetonyl-CoA was determined spectrally using E = 15.4 X 10J M” cm-l at 260 nm. ’ Rates are expressed as micromoles of NADH oxidized per min in

a coupled assay with malic dehydrogenase. Nonspecific NADH oxi- dase activity, independent of added phosphoenolpyruvate or acetyl- CoA, was subtracted from the observed rates (see under “Experimen- tal Procedures.”).

will, however, catalyze the hydrolysis of acetyl-coA in the presence of arsenate. We studied this reaction.

The formation of acetyl carnitine from carnitine and acetyl- CoA catalyzed by carnitine acetyltransferase is freely revers- ible. We studied this reaction in the direction of acetyl carni- tine synthesis. For both enzymes, acetonyl-CoA was an effec- tive competitive inhibitor with respect to acetyl-coA (Table 111). However, for both these enzymes, the K , of the acetonyl compound was not as close to the K, values for acetyl-coA as in the case of citrate synthase.

Interaction ofAcetony1-CoA with Two Enzymes Requiring Acetyl-coA as a n Obligate Actiuator-Members of this third class of enzymes, Escherichia coli phosphoenolpyruvate car- boxylase and rat liver pyruvate carboxylase, were also tested with the acetonyl analog. Both require activation by acetyl- CoA for activity (3, 12). Pyruvate kinase also catalyzes a slow hydrolysis of acetyl-coA, although protein acetylation is not thought to be a part of the activation process (13). The results with phosphoenolpyruvate carboxylase are shown in Table

7862 S-Acetonyl-CoA: A Nonreactive Analog of Acetyl-coA

TABLE V Effect of acetonyl-CoA on rat liver pyruvate carboxylase

Activator Concentration" Rateh

pmol/liter prnoi/rnin

Acetyl-coA 200 0.017 Acetyl-coA I00 0.013 Acetyl-coA 40 0.003 Acetonyl-CoA 100 O.OO0 Acetonyl-CoA + 500 0.002

Acetonyl-CoA + 500 0.012 acetyl-coA 40

acetyl-coA 100 ___

E = 15.4 X IO' M" cm" at 260 nm. " Concentrations of acetonyl-CoA were determined spectrally using

" Rates are expressed as micromoles of NADH oxidized per min in a coupled assay with malic dehydrogenase. Nonspecific NADH oxi- dase activity, independent of added phosphoenolpyruvate or acetyl- CoA, was subtracted from the observed rates (see under "Experimen- tal Procedures").

IV. Over a concentration range similar to that used for acetyl- CoA, no activation of the enzyme was achieved with S-ace- tonyl-CoA. Furthermore, the activation achieved by 40 pmol/ liter of acetyl-coA in the presence of 100 pmol/liter of ace- tonyl-CoA was the same as that observed in the presence of 40 pmol/liter of acetyl-coA alone. Thus, the acetonyl analog competed very poorly, if at all, with acetyl-coA for the acti- vator site on the enzyme surface. Similar results were obtained with rat liver pyruvate carboxylase (Table V). No activity was obtained with 100 pmol/liter of acetonyl-CoA, and the ace- tonyl-CoA competed poorly, if at all, with acetyl-coA binding to the activation site on the enzyme. The small differences in rates are not significant.

DISCUSSION

We have synthesized a nonreactive acetyl-coA analog, S- acetonyl-CoA, and characterized its structure on the basis of functional group identification and specificity toward enzy- matic digestion. The compound is prepared easily and in high yields.

This compound interacts strongly with pig heart citrate synthase as a competitive inhibitor with respect to acetyl- CoA. Reduction of the keto group of acetonyl-CoA with sodium borohydride to the isopropanolyl analog causes an apparent 9- to lO-fold rise in K, from 10 pmol/liter to about 90 pmol/liter. This decrease in binding may occur as a result of the substitution of a hydroxyl group for a keto group. Likewise, it may reflect the fact that borohydride reduction produces two steric isomers with respect to the hydroxyl carbon, one which binds tightly and one which binds Doorly. Our results show that a thioester is required for carbanion formation on the acetyl methyl residue and subsequent reaction with oxal- oacetate. However, shifting the carbon-oxygen double bond one carbon further away from the sulfur, forming a thioether instead of the original thioester, has little effect on the ability of the acetonyl compound to bind specifically to the enzyme. We also find that with phosphotransacetylase and carnitine acetyltransferase, S-acetonyl-CoA is a good competitive inhib- itor with respect to acetyl-coA.

Our results with phosphoenolpyruvate carboxylase and py- ruvate carboxylase, both of which require acetyl-coA as an allosteric activator, were quite surprising to us. Based on our observations with the acetyltransferases and citrate synthase,

we would have predicted that acetonyl-CoA would either activate these enzymes or at least inhibit activation by acetyl- CoA. This prediction was especially strong for pyruvate car- boxylase where propionyl-CoA, which like acetonyl-CoA has 3 carbon atoms linked to the sulfur, is as effective as is acetyl- CoA (13). We observed for both enzymes, however, that not only was acetonyl-CoA incapable of activation, but also that it would not compete with acetyl-coA for binding at the concentrations tested. These results show the absolute re- quirements for a thioester for both binding to and activation of these enzymes. Scrutton and Utter (13) reported that pyruvate carboxylase catalyzed the slow hydrolysis of acetyl- CoA but suggested that the hydrolysis itself was not required for enzyme activation. This observation coupled with our results with acetonyl-CoA is consistent with the idea that for binding and activation, a specific interaction between the enzyme and the thioester moiety of acetyl-coA may be re- quired. This interaction may involve formation of a covalent bond between the carboxyl carbon of the thioester and the enzyme.

We have also tested acetonyl-CoA as an inhibitor of nascent polypeptide chain acetylation in a rabbit reticulocyte protein translation system. Preliminary results show that when ace- tonyl-CoA is used in conjunction with citrate synthase and oxaloacetate (1) inhibition of protein acetylation is much greater than if citrate synthase and oxaloacetate are used alone. Details of these experiments will be published else- where.

In summary, we have synthesized S-acetonyl-CoA, a non- reactive acetyl-coA analog, and used it successfully as a mechanistic probe for enzymic catalysis and as a tool in in vitro protein acetylation studies. This molecule should prove useful as a tool in studying any process involving the partici- pation of acetyl-coA.

Acknowledgments-William Belias and Aaron Rappaport, partic- ipants in the Summer Science Training Program at the University of Iowa were involved in the initial phases of this research. NMR spectra were recorded by the Chemistry Department NMR Service, Univer- sity of Iowa. We especially thank Dr. L. Cocco for helping with the computer analyses.

REFERENCES

1. Palmiter, R. D. (1977) J. Bioi. Chem. 252, 8781-8783 2. Catch, J . R., Elliott, D. F., Hey, D. H., and Jones, E. R. (1948) J .

3. Scrutton, M., Olmsted, M., and Utter. M. (1969) Methods En-

4. Srere, P. A., Brazil, H., and Gonen, L. (1963) Acta Chem. Scand.

5. Stadtman, E. R. (1952) J . Biol. Chem. 196,527-534 6. Chase, J. F. (1969) Methods Enzymol. 13,387-393 7. Smith, T. (1968) Arch. Biochem. Biophys. 128,611-622 8. McClure, W. R., Lardy, H. A,, and Kneifel, H. P. (1971) J. Biol.

9. Cleland, W. W. (1979) Methods Enzymol. 63, 103-138

Chem. SOC. (Lond.) 272-275

zymol. 13,235-249

17, 5129

Chem. 246, 3569-3578

10. Wang, T., Shuster, L., and Kaplan, N. 0. (1954) J. Biol. Chem.

11. Roeder, S., Master, B., and Stewart, C. (1975) Physiol. Chem.

12. Canovas, J. L., and Komberg, H. L. (1965) Biochim. Biophys.

13. Scrutton, M. C., and Utter, M. F. (1967) J. Biol. Chem. 242, 1723-

206,209-309

Phys. 7, 115-123

Acta 96, 169-172

1735